Estimation and Improvement of Recovery of Low Grade Copper Oxide Using Sulfide Activation Flotation Method Based on GA–BPNN

processes

Article Estimation and Improvement of Recovery of Low Grade Copper Oxide Using Sulfide Activation Flotation Method Based on GA–BPNN

Wenlin Nie 1,2 , Jianjun Fang 1,*, Shuming Wen 1, Qicheng Feng 1,*, Yanbing He 3 and Xiaoyong Yang 4

1 State Key Laboratory of Complex Nonferrous Metal Resources Clean Utilization, Faculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming 650093, China; [email protected] (W.N.); [email protected] (S.W.) 2 Department of Mathematics and Physics, West Yunnan University, Lincang 677000, China 3 Lincang Science and Technology Information Station, Lincang 677000, China; [email protected] 4 School of Biotechnology and Engineering, West Yunnan University, Lincang 677000, China; [email protected] * Correspondence: [email protected] (J.F.); [email protected] (Q.F.); Tel.: +86-13769182741 (J.F.); +86-15987190563 (Q.F.)

Abstract: Copper oxide ore is an important copper ore resource. For a certain copper oxide ore in Yunnan, China, experiments have been conducted on the grinding fineness, collector dosage, sodium sulfide dosage, inhibitor dosage, and activator dosage. The results showed that, by controlling the above conditions, better sulfide flotation indices of copper oxide ore are obtained. Additionally, ammonium bicarbonate and ethylenediamine phosphate enhanced the sulfide flotation of copper oxide ore, whereas the combined activator agent exhibited a better performance than either individual activator. In addition, to optimize all of the conditions in a more reasonable way, a combination of the 5-11-1 genetic algorithm and back propagation neural network (GA–BPNN) was used to set up Citation: Nie, W.; Fang, J.; Wen, S.; a mathematical optimization model. The results of the back propagation neural network (BPNN) Feng, Q.; He, Y.; Yang, X. Estimation 2 and Improvement of Recovery of Low model showed that the R value was 0.998, and the results were in accordance with the requirement Grade Copper Oxide Using Sulfide model. After 4169 iterations, the error in the objective function was 0.001, which met the convergence Activation Flotation Method Based on requirements for the final solution. The genetic algorithm (GA) model was used to optimize the GA–BPNN. Processes 2021, 9, 583. BPNN model. After 100 generations, a copper recovery of 87.62% was achieved under the following https://doi.org/10.3390/pr9040583 conditions: grinding fineness of 0.074 mm, which accounted for 91.7%; collector agent dosage of 487.7 g/t; sodium sulfide dosage of 1157.2 g/t; combined activator agent dosage of 537.8 g/t; inhibitor Academic Editor: Fausto Gallucci dosage of 298.9 g/t. Using the combined amine and ammonium salt to enhance the sulfide activation efficiency, a GA–BPNN model was used to achieve the goal of global optimizations of copper oxide Received: 26 February 2021 ore and good flotation indices were obtained. Accepted: 22 March 2021 Published: 26 March 2021 Keywords: copper oxide; sulfide activation flotation method; BP neural network; genetic algorithm

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- 1. Introduction iations. Copper oxide is an important mineral resource, and studying the recovery of low grade copper oxide is of great importance for solving the problems of the shortage of copper resources and promoting its efficient utilization [1–3]. The recycling methods of copper oxide mainly include flotation, leaching, and dressing-metallurgy. Flotation is the Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. most widely used copper recycling method [4–6]. Because of the hydrophilic nature of This article is an open access article copper oxide minerals, vulcanizing agents are generally added to make the surface of distributed under the terms and sulfide minerals hydrophobic for flotation. However, most copper oxide ores are associated conditions of the Creative Commons with copper sulfide minerals. The vulcanizing agent reacts on the interface of copper oxide Attribution (CC BY) license (https:// and exhibits different degrees of inhibition on copper sulfide minerals [7,8]. Therefore, the creativecommons.org/licenses/by/ development of effective activators is needed to solve these problems. On one hand, the 4.0/). sulfide concentration in the interface of copper oxide minerals should be increased, but on

Processes 2021, 9, 583. https://doi.org/10.3390/pr9040583 https://www.mdpi.com/journal/processes Processes 2021, 9, 583 2 of 14

the other hand, the inhibition of the vulcanizing agent on copper sulfide minerals should be reduced [9]. Recently, some studies have been conducted on the activators used in the recycling of copper oxide minerals [10]. These studies have revealed problems, such as an unsatisfactory activation effect and high consumption of the activator. Some studies have shown that the combination of amine and an ammonium salt activator exhibits a satisfactory effect on the flotation of sulfide in copper oxide minerals [11–13]. One of the works studied copper oxide minerals from Yunnan, China, using the sulfuration flotation method. The activation effect of amine and ammonium on copper floatation indices was examined. Meanwhile, on the basis of the experimental study, an optimization model to reasonably predict the recovery of floatation concentrate was established to improve the flotation effect of copper oxide minerals and provide references for production and further research. The optimization models of flotation have mainly included traditional models, such as the orthogonal design, response surfaces method, and factor design. Furthermore, a neural network model has been developed in recent years [14,15]. Results from some previous studies have shown that the prediction and optimization of experimental data using the neural network model was better than those of the response surface method [16,17]. In particular, the combined model tended to have more advantages than the individual neural network model. In a combined model, the neural network is combined with one of the other models, namely the genetic algorithm, particle swarm algorithm, simulated annealing algorithm, and support vector machine. In particular, the combined model of the genetic algorithm (GA) and back propagation (BP) neural network has most commonly been used in recent years [18–20]. The BP neural network is a kind of multilayer feed-forward neural network, whose transfer function of the neurons is an S type function, and the output is a continuous quantity between 0–1. The model can produce arbitrary nonlinear predictions from input to output [21,22]. The BP neural network algorithm has become one of the most important models of the neural network and has been widely used in chemical, safety, and electrical industries. The genetic algorithm is a random search method that simulates natural selec- tion and the genetic mechanism of Darwinian evolution in biological evolution processes, and using this method helps in searching for the optimal solution. As a discontinuous or differentiable global optimization algorithm to solve the objective function, the BP neural network lacked the ability to optimize and made up for the deficiency of the neural network because it has the advantages of a good global searching ability, potential paral- lelism, ability to compare multiple individuals at the same time, and a simple process. The combined model of the genetic algorithm and BP neural network has been widely used in experimental design and optimization. Furthermore, the application of the combined model has also been reported in flotation experiments. Using the neural network and genetic algorithm, Allahkarami [23] established a mathematical model to improve the copper recovery and grade in the process of industrial flotation. The feed-forward neural networks of 10-10-10-4 with two hidden layers were selected, and studied under various conditions of pH, dosage of reagents, granularity content of below 75 µm, and feeding granularity. The initial weights of the neural network were optimized using the genetic algorithm. The results showed that the model based on the genetic algorithm and neural network was better than that based on the individual neural network. In addition, Dong [24] used the genetic algorithm and neural network to establish a parameter prediction model for gold mine mill-grinding and the flotation process. Furthermore, Tang [25] established a slurry pH prediction model based on the mixed adaptive genetic neural network of bubble visual information. Copper oxide flotation indices have mainly included recovery and grade. The test was conducted mainly to improve the recovery of flotation concentrate. A prediction optimization model of copper oxide flotation concentrate, which was based on the genetic algorithm and back propagation neural network (GA–BPNN), was established considering the single factor experimental data. In addition, a regression analysis model was established for comparing the results. Processes 2021, 9, x FOR PEER REVIEW 3 of 13

considering the single factor experimental data. In addition, a regression analysis model was established for comparing the results.

2. Materials and Methods 2.1. Materials The test core samples were taken from an undisclosed location in Yunnan province, Processes 2021, 9, 583 3 of 14 China. The particle size distribution of the undressed ore mineral was not uniform, and it largely showed a block, conglomeratic, and plate-type distribution. The XRD spectrum of the2. Materials undressed and Methods ore is shown in Figure 1. It can be seen from Figure 1 that the original min- 2.1. Materials eral belonged to high-alkaline minerals. The gangue was mainly composed of dolomite The test core samples were taken from an undisclosed location in Yunnan province, andChina. quartz. The particle The size copper distribution-containing of the undressed minerals ore mineral wasmainly not uniform, consisted and it of malachite and chalcopy- largely showed a block, conglomeratic, and plate-type distribution. The XRD spectrum of rite.the undressed The copper ore is shown phase in Figure analysis1. It can be seenresults from Figureare 1presented that the original in mineral Table 1. It can be seen from Table 1belonged that the to high-alkaline total copper minerals. content, The gangue oxidation was mainly rate, composed and of dolomitecombination and rate were 0.47%, 50.66%, quartz. The copper-containing minerals mainly consisted of malachite and chalcopyrite. andThe copper 12.34%, phase respectively. analysis results are These presented values in Table1 .showed It can be seen that from the Table oxidation1 rate was higher; this belongedthat the total copperto low content,-grade oxidation copper rate, ore, and combination which is rate difficult were 0.47%, to 50.66%,oxidize. Based upon these values, and 12.34%, respectively. These values showed that the oxidation rate was higher; this itbelonged can be to low-gradesaid that copper the ore, sulfide which is difficultflotation to oxidize. method Based uponcan thesebe used values, to effectively recycle copper mineralit can be said resources. that the sulfide flotation method can be used to effectively recycle copper mineral resources.

4500 1 4000

3500 1—Dolomite

3000 2—Quartz

-1 2500

2000

Intensity/s 1500 2 1000 1 1 1 1 500 1 1 2 2 2 2 0 20 30 40 50 60 70 80 2q/deg FigureFigure 1. XRD1. XRD analysis analysis of materials. of materials. Table 1. Results of copper phase analysis.

Table 1. ResultsCopper Phase of copper phase analysis. Content/% Percentage/% Free oxidized copper 0.18 38.30 Conjunction oxidizedCopper copper Phase 0.058Content/% 12.34 Percentage/% Chalcanthite 0.022 4.68 Copper sulﬁdeFree and others oxidized copper 0.21 44.680.18 38.30 ConjunctionTotal copper oxidized copper 0.47 1000.058 12.34

Some commonly usedChalcanthite flotation reagents include sodium sulfide (industrial-grade),0.022 4.68 ammonium bicarbonate (industrial-grade), ethylenediamine phosphate (industrial-grade), isoamyl xanthateCopper (industrial-grade), sulfide sodiumand others hexametaphosphate (industrial-grade),0.21 and 44.68 sodium silicate (industrial-grade).Total copper The equipment used in the current study included0.47 an 100 XMQ67 type ball mill produced by Wuhan prospecting machinery, China, HG101-3 type electrothermal blower produced by Nanjing testing instruments, China, and XFD type small flotationSome machinecommonly produced used by Jilin flotation prospecting machinery,reagents China. include sodium sulfide (industrial-grade), ammonium2.2. Experiments bicarbonate (industrial-grade), ethylenediamine phosphate (industrial- grade),Ore mineral isoamyl samples xanthate were crushed (industrial to −3 mm size- usinggrade), a laboratory sodium crusher, hexametaphosphate and (industrial- crushed samples were preserved for further use. A 500 g ore sample was taken for each test. grade),After a certain and period sodium of grinding silicate in a wet (industrial ball mill, the pulp-grade). was moved The into equipment an XFD-III used in the current study includedflotation tank, an which XMQ67 had a volume type of 1.5 ball L. According mill toproduced the dosing formulation by Wuhan system, prospecting the machinery, China, HG101-3 type electrothermal blower produced by Nanjing testing instruments, China, and XFD type small flotation machine produced by Jilin prospecting machinery, China.

2.2. Experiments Ore mineral samples were crushed to −3 mm size using a laboratory crusher, and crushed samples were preserved for further use. A 500 g ore sample was taken for each test. After a certain period of grinding in a wet ball mill, the pulp was moved into an XFD- Ⅲ flotation tank, which had a volume of 1.5 L. According to the dosing formulation system, the reagent was selected and added to the mixture. As shown in Figure 2, the flotation

Processes 2021, 9, x FOR PEER REVIEW 4 of 13

Processes 2021, 9, 583 4 of 14 process used was the “two roughing-one scavenging”. The flotation reagents were matched with a certain solution concentration and were added directly to the pulp. In this work,reagent sodium was selected sulfide and added was to used the mixture. as the As vulcanizing shown in Figure agent;2, the flotation ethylenediamine process phosphate and ammoniumused was the “two bicarbonate roughing-one and scavenging”. the combination The flotation of reagents two kinds were matched of activators with a were used as the activationcertain solution agent; concentration the collec andtor were was added isoamyl directly xanthate; to the pulp. sodium In this work, hexametaphosphate sodium and so- sulfide was used as the vulcanizing agent; ethylenediamine phosphate and ammonium diumbicarbonate silicate and were the combination selected ofas two inhibitor kinds of activatorsagents; 2# were oil used was as the the activationfoaming agent. The temper- atureagent; thewas collector room was temperature isoamyl xanthate; (298.15 sodium K), hexametaphosphate and ultrapure and water sodium was silicate used as the flotation water.were selected Because as inhibitor of the agents; limitation 2# oil was of theprocess foaming conditions, agent. The temperature the influence was room of pH was not consid- temperature (298.15 K), and ultrapure water was used as the flotation water. Because of ered.the limitation After the of process flotation, conditions, the flotation the influence concentrate of pH was was not considered. filtered using After thevacuum filtration, and thenflotation, the the filtrate flotation was concentrate placed was in filtered a drying using vacuum oven filtration,for drying. and then After the filtrate drying, the filtrate was weighedwas placed and in a drying shrunk. oven Then, for drying. the After sample drying, of thethe filtrate copper was weighedgrade in and the shrunk. concentrate was tested. Then, the sample of the copper grade in the concentrate was tested. Based upon the test Basedresults, upon the recovery the test was calculated.results, the recovery was calculated.

Raw ore

RoughingⅠ 8min Roughing Ⅱ

7min Scavenging

8min

Concentrates Tailings Middlings FigureFigure 2. 2.Flow Flow chart chart of the of two the roughing-one two roughing scavenging-one mineral scavenging processing. mineral processing. 3. Results and Discussion 3. ResultsConsidering and theDiscussion low grade oxidized copper ore in Yunnan province (China), first, a single factor experiment was carried out to examine the changes in various factors that affectConsidering the flotation the indices, low andgrade the oxidized amine activation copper effect ore of in the Yunnan ammonium province salt (China), first, a singlewas demonstrated. factor experiment was carried out to examine the changes in various factors that affect the flotation indices, and the amine activation effect of the ammonium salt was 3.1. Grinding Fineness demonstrated.First, under different grinding conditions, the minerals were examined for changes in flotation concentrate and recovery of copper grade. The ground particles finer than 3.1.0.074 Grinding mm accounted Fineness for 70%, 75%, 80%, 85%, 90%, and 95% of the total ground sample respectively. Under the conditions of a fixed collecting agent dosage of 600 g/t, vulcanizing agentFirst, dosage under of 1000 different g/t, and foaming grinding agent conditions, dosage of 100 the g/t, minerals the experiments were were examined for changes incarried flotation out. The concentrate obtained results and are shownrecovery in Figure of 3copper. grade. The ground particles finer than 0.074It canmm be accounted seen from Figure for3 that70%, the 75%, content 80%, particles 85%, that 90%, are of −and0.074 95% mm accountedof the total ground sample for 90%, and 78.73% of the highest flotation recovery were obtained. When the grinding respectively.fineness was lower Under than the 90%, conditions the flotation indexof a fixed was not collecting ideal, which agent was because dosage of of 600 g/t, vulcaniz- ingthe insufficientagent dosage dissociation of 1000 of copper g/t, and mineral, foaming which couldagent not dos be sufficientlyage of 100 increased. g/t, the experiments were carriedWhen the out. grinding The finenessobtained was results higher than are 90%,shown the flotationin Figure index 3. decreased, which can be attributed to granular slime. Meanwhile, the sulfide ores containing copper was inhibited due to the addition of vulcanizing agent. The flotation concentrate copper grade first increased and then exhibited a decreasing trend, whereas it had the highest value for the grinding fineness of 85%. Therefore, the optimum grinding fineness for −0.074 mm was the one that accounted for 90%.

Figure 3. Influences of the grinding fineness of copper oxide flotation index.

Processes 2021, 9, x FOR PEER REVIEW 4 of 13

process used was the “two roughing-one scavenging”. The flotation reagents were matched with a certain solution concentration and were added directly to the pulp. In this work, sodium sulfide was used as the vulcanizing agent; ethylenediamine phosphate and ammonium bicarbonate and the combination of two kinds of activators were used as the activation agent; the collector was isoamyl xanthate; sodium hexametaphosphate and sodium silicate were selected as inhibitor agents; 2# oil was the foaming agent. The temperature was room temperature (298.15 K), and ultrapure water was used as the flotation water. Because of the limitation of process conditions, the influence of pH was not considered. After the flotation, the flotation concentrate was filtered using vacuum filtration, and then the filtrate was placed in a drying oven for drying. After drying, the filtrate was weighed and shrunk. Then, the sample of the copper grade in the concentrate was tested. Based upon the test results, the recovery was calculated.

Raw ore

RoughingⅠ 8min Roughing Ⅱ

7min Scavenging

8min

Concentrates Tailings Middlings Figure 2. Flow chart of the two roughing-one scavenging mineral processing.

3. Results and Discussion Considering the low grade oxidized copper ore in Yunnan province (China), first, a single factor experiment was carried out to examine the changes in various factors that affect the flotation indices, and the amine activation effect of the ammonium salt was demonstrated.

3.1. Grinding Fineness First, under different grinding conditions, the minerals were examined for changes in flotation concentrate and recovery of copper grade. The ground particles finer than 0.074 mm accounted for 70%, 75%, 80%, 85%, 90%, and 95% of the total ground sample respectively. Under the conditions of a fixed collecting agent dosage of 600 g/t, vulcaniz- Processes 2021, 9, 583 ing agent dosage of 1000 g/t, and foaming agent dosage of 1005 of 14 g/t, the experiments were Processes 2021, 9, x FOR PEER REVIEW 5 of 13 carried out. The obtained results are shown in Figure 3.

It can be seen from Figure 3 that the content particles that are of −0.074 mm accounted for 90%, and 78.73% of the highest flotation recovery were obtained. When the grinding fineness was lower than 90%, the flotation index was not ideal, which was because of the insufficient dissociation of copper mineral, which could not be sufficiently increased. When the grinding fineness was higher than 90%, the flotation index decreased, which can be attributed to granular slime. Meanwhile, the sulfide ores containing copper was inhibited due to the addition of vulcanizing agent. The flotation concentrate copper grade first increased and then exhibited a decreasing trend, whereas it had the highest value for the grinding fineness of 85%. Therefore, the optimum grinding fineness for −0.074 mm was the one that accounted for 90%.

3.2. Dosage of the Collecting Agent Figure 3. Influences of the grinding fineness of copper oxide flotation index. FigureThe 3. dosages Influences of collectingof the grinding agent fineness tested were of copper 200 g/t, oxide 300 flotation g/t, 400 index.g/t, 500 g/t, 600 g/t, and3.2. 700 Dosage g/t, of theand Collecting the grinding Agent fineness was fixed at −0.074 mm accounting for 90% of the The dosages of collecting agent tested were 200 g/t, 300 g/t, 400 g/t, 500 g/t, 600 g/t, totaland ground 700 g/t, and sample the grinding. In addition, fineness was the fixed vulcanizing at −0.074 mm agent accounting dosage for 90% was of set the to be 1000 g/t, and thetotal inhibitor ground sample. agent In dosage addition, was the vulcanizing 300 g/t. The agent foaming dosage was agent set to be dosage 1000 g/t, was and kept the same and hadthe a inhibitor value agentof 100 dosage g/t. was The 300 single g/t. The factor foaming experiment agent dosage wasfor keptthe thecollecting same and agent dosage was had a value of 100 g/t. The single factor experiment for the collecting agent dosage was carried out, and the results are shown in Figure 4. carried out, and the results are shown in Figure4.

FigureFigure 4. 4. InfluencesInfluences of theof collectingthe collecting agent dosage agent of thedosage copper of oxide the flotationcopper index. oxide flotation index. It can be seen from Figure4 that the recovery rate of the oxide copper flotation increasedIt can with be an seen increase from in isoamyl Figure xanthate 4 that dosage, the recovery which was 500rate g/t of later, the and oxide became copper flotation in- creasedunaffected with with an changes increase in isoamyl in isoamyl xanthate xanthate dosage. The dosage, grade of which flotation was concentrate 500 g/t later, and became was gradually reduced with the increase in isoamyl xanthate dosage. Therefore, in the unaffectedfollowing single with factor changes experiments, in isoamyl the dosage xanthate of isoamyl dosage. xanthate was The kept grade to be 500 of g/t.flotation concentrate was gradually reduced with the increase in isoamyl xanthate dosage. Therefore, in the 3.3. Dosage of Vulcanizing Agent following single factor experiments, the dosage of isoamyl xanthate was kept to be 500 The dosage of vulcanizing agent is also an important factor for the flotation study g/t.of copper oxide ore. The experiments were conducted for sodium sulfide dosages of 400,

3.3. Dosage of Vulcanizing Agent The dosage of vulcanizing agent is also an important factor for the flotation study of copper oxide ore. The experiments were conducted for sodium sulfide dosages of 400, 600, 800, 1000, 1200, and 1400 g/t. Furthermore, the collecting agent dosage was 500 g/t, and the foaming agent’s dosage was 100 g/t. Additionally, the grinding fineness was −0.074 mm accounting for 90%. The experimental results are as shown in Figure 5.

Processes 2021, 9, 583 6 of 14

Processes 2021, 9, x FOR PEER REVIEW 6 of 13 600, 800, 1000, 1200, and 1400 g/t. Furthermore, the collecting agent dosage was 500 g/t, and the foaming agent’s dosage was 100 g/t. Additionally, the grinding ﬁneness was −0.074 mm accounting for 90%. The experimental results are as shown in Figure5.

FigureFigure 5. 5.Influences Influences of the sodiumof the sulfide sodium dosage sulfide of the copper dosage oxide of flotation the copper index. oxide flotation index. It can be seen from Figure5 that when the sodium sulfide dosage was 1200 g/t, the recoveryIt can rate wasbe theseen highest from and Figure reached a5 value that of when 79.83%%, the which sodium corresponded sulfide to a dosage was 1200 g/t, the grade of 2.88%. Furthermore, the copper grade increased with an increase in dosage of recoverysodium sulfide. rate Additionally, was the whenhighest the sodium and reached sulfide dosage a value was 1200 of g/t, 79.83%%, the recovery which corresponded to a graderate was of the 2.88%. highest. Furt Therefore,hermore, in the following the copper single factorgrade experiments, increased the sodiumwith an increase in dosage of sulfide dosage was set to be 1200 g/t. sodium sulfide. Additionally, when the sodium sulfide dosage was 1200 g/t, the recovery rate3.4. Dosagewas ofthe the Activationhighest. Agent Therefore, in the following single factor experiments, the sodium On one hand, the addition of sodium sulfide enhanced the interfacial vulcanization of sulfidethe oxidized dosage ore, the was hydrophobicity set to be of 1200 the interface g/t. of the oxidizing ore, and the flotation effect of the oxidized ore. However, on the other hand, it inhibited the sulfide ore, which 3.4.was detrimentalDosage of to the the flotationActivation of sulfide Agent ore. Therefore, the use of an activator was needed; an activator plays an important role in achieving an enhanced oxidation of the ore and the reductionOn one of sulfidehand, for the reducing addition the inhibitory of sodium effect. Insulfide this study, enhanced ethylenediamine the interfacial vulcanization phosphate and ammonium bicarbonate were used individually, and a combination of these ofwere the used oxidized as combined ore, amine the andhydrophobicity ammonium salt to of examine the interface the effect of of flotation the oxidizing on ore, and the flotationcopper effect oxide minerals.of the Ethylenediamineoxidized ore. phosphate However, is as a copperon the oxide other flotation hand, activator, it inhibited the sulfide ore, whichhas a wide was range detrimental of applications into industrial the flotation production, of and sulfide achieves ore. good Therefore, results. First, the use of an activator the activation effects of ammonium bicarbonate and ethylenediamine phosphate were wasexplored. needed; Six different an activator dosage rates plays of ethylenediamine an important phosphate role were in achieving selected for the an enhanced oxidation of theexperiments. ore and These the dosage reduction rates were of 50, sulfide 100, 150, for 200, 250,reducing and 300 g/t.the The inhibitory dosage rates effect. In this study, eth- for ammonium bicarbonate were 200, 300, 400, 500, 600, and 700 g/t. Furthermore, the ylenediaminecollecting agent dosage phosphate and foaming and agent ammonium dosage were 500 bicarbonate g/t and 100 g/t, were respectively. used individually, and a com- binationIn addition, of the these grinding were fineness used was − as0.074 combined mm accounting amine for 90%e, and and ammonium the vulcanizing salt to examine the effect agent dosage was set to be 1200 g/t. The test results are as shown in Figures6 and7. of floIttation can be seen on from copper Figures oxide6 and7 that minerals. both ammonium Ethylenediamine bicarbonate and ethylenedi- phosphate is as a copper oxide flotationamine phosphate activator, had a positive has a effect wide on sulfiderange flotation of applications of copper oxide in ore. industrial Additionally, production, and achieves goodthe dosage results. of ethylenediamine First, the activation phosphate was effects obviously of lowerammonium than that of bicarbonate ammonium and ethylenediamine bicarbonate. As low-dosage activators, the optimal dosages of ethylenediamine phosphate phosphateand ammonium were bicarbonate explored. were 150Six g/t different and 600 g/t, dosage respectively. rates On of the ethylenediamine basis of a phosphate were selectedsingle-factor for activator the experiments. test, a dosage test These was conducted dosage on therates combined were activator 50, 100, (com- 150, 200, 250, and 300 g/t. bination of amine and ammonium salts). The test selected a combination of amine and The dosage rates for ammonium bicarbonate were 200, 300, 400, 500, 600, and 700 g/t. Furthermore, the collecting agent dosage and foaming agent dosage were 500 g/t and 100 g/t, respectively. In addition, the grinding fineness was −0.074 mm accounting for 90%e, and the vulcanizing agent dosage was set to be 1200 g/t. The test results are as shown in Figures 6 and 7.

Figure 6. Influences of the ethylenediamine phosphate dosage of the copper oxide flotation index.

Processes 2021, 9, x FOR PEER REVIEW 6 of 13

Figure 5. Influences of the sodium sulfide dosage of the copper oxide flotation index.

It can be seen from Figure 5 that when the sodium sulfide dosage was 1200 g/t, the recovery rate was the highest and reached a value of 79.83%%, which corresponded to a grade of 2.88%. Furthermore, the copper grade increased with an increase in dosage of sodium sulfide. Additionally, when the sodium sulfide dosage was 1200 g/t, the recovery rate was the highest. Therefore, in the following single factor experiments, the sodium sulfide dosage was set to be 1200 g/t.

3.4. Dosage of the Activation Agent On one hand, the addition of sodium sulfide enhanced the interfacial vulcanization of the oxidized ore, the hydrophobicity of the interface of the oxidizing ore, and the flotation effect of the oxidized ore. However, on the other hand, it inhibited the sulfide ore, which was detrimental to the flotation of sulfide ore. Therefore, the use of an activator was needed; an activator plays an important role in achieving an enhanced oxidation of the ore and the reduction of sulfide for reducing the inhibitory effect. In this study, ethylenediamine phosphate and ammonium bicarbonate were used individually, and a combination of these were used as combined amine and ammonium salt to examine the effect of flotation on copper oxide minerals. Ethylenediamine phosphate is as a copper oxide flotation activator, has a wide range of applications in industrial production, and achieves good results. First, the activation effects of ammonium bicarbonate and ethylenediamine phosphate were explored. Six different dosage rates of ethylenediamine phosphate were Processes 2021, 9, 583 7 of 14 selected for the experiments. These dosage rates were 50, 100, 150, 200, 250, and 300 g/t. The dosage rates for ammonium bicarbonate were 200, 300, 400, 500, 600, and 700 g/t. ammoniumFurthermore, salts as the activator collecting to investigate agent thedosage ﬂotation and effect foaming of copper oxide agent mineral. dosage were 500 g/t and 100 Theg/t, dosage respectively. rates of the activatorIn addition, agent were the 110, grinding 220, 330, 440, fineness 550, and was 660 g/t, −0.074 and the mm accounting for 90%e, ratio between ammonium bicarbonate and ethylenediamine phosphate was 10:1. Further- more,and the collecting vulcanizing agent, vulcanizing agent dosage agent, and was foaming set to agent be dosages1200 g/t. were The 500 g/t, test results are as shown in 1200Figures g/t, and 6 and 100 g/t, 7. respectively. The grinding ﬁneness was −0.074 mm accounting for 90%. The results are shown in Figure8.

Processes 2021, 9, x FOR PEER REVIEW 7 of 13

Figure 6. Inﬂuences6. Influences of the ethylenediamine of the ethylenediamine phosphate dosage phosphate of the copper oxidedosage ﬂotation of index.the copper oxide flotation index.

FigureFigure 7. 7.Influences Influences of the ammoniumof the ammonium bicarbonate dosage bicarbonate of the copper dosage oxide flotation of the index. copper oxide flotation index. The results presented in Figures6–8 show that the dosage for the combination of amineIt and can ammonium be seen salts from was significantlyFigures 6 lessand than 7 thatthat for both the individual ammonium salts (when bicarbonate and ethylene- used as activators). Additionally, the combined activation agent had an obvious effect on diaminethe flotation phosphate recovery of copper had and a positive exhibited more effect stringent on requirementssulfide flotation for the dosage. of copper oxide ore. Addi- tionally,When the dosagethe dosage was low, of the ethylenediamine activation effect was not phosphate obvious. At a highwas dosage obviously rate, an lower than that of am- obvious inhibitory effect was observed on the mineral. Jiang and Mao [12,13] studied the moniumenhancement bicarbonate. of copper recovery As from low pure-do mineralsage malachite activators, through the experiments optimal using dosages the of ethylenediamine phosphatecombined ammonia-ammonium and ammonium activator. bicarbonateThe results showed were that150 the g/t combined and 600 ammonia- g/t, respectively. On the basis ofammonium a single activator-factor (based activator on ethylenediamine test, a dosage phosphate test and was ammonium conducted bicarbonate) on the combined activator exhibited better activation effects along with good synergistic effects. The results were in (combination of amine and ammonium salts). The test selected a combination of amine and ammonium salts as the activator to investigate the flotation effect of copper oxide mineral. The dosage rates of the activator agent were 110, 220, 330, 440, 550, and 660 g/t, and the ratio between ammonium bicarbonate and ethylenediamine phosphate was 10:1. Furthermore, the collecting agent, vulcanizing agent, and foaming agent dosages were 500 g/t, 1200 g/t, and 100 g/t, respectively. The grinding fineness was −0.074 mm accounting for 90%. The results are shown in Figure 8.

Figure 8. Influences of the amine-ammonium combination activating dosage of the copper oxide flotation index.

The results presented in Figures 6–8 show that the dosage for the combination of amine and ammonium salts was significantly less than that for the individual salts (when used as activators). Additionally, the combined activation agent had an obvious effect on the flotation recovery of copper and exhibited more stringent requirements for the dosage. When the dosage was low, the activation effect was not obvious. At a high dosage rate, an obvious inhibitory effect was observed on the mineral. Jiang and Mao [12,13] studied the enhancement of copper recovery from pure mineral malachite through experiments using the combined ammonia-ammonium activator. The results showed that the combined ammonia-ammonium activator (based on ethylenediamine phosphate and ammonium bicarbonate) exhibited better activation effects along with good synergistic effects. The results were in accordance with those obtained in the present study. Therefore, the dosage of 440

Processes 2021, 9, x FOR PEER REVIEW 7 of 13

Figure 7. Influences of the ammonium bicarbonate dosage of the copper oxide flotation index.

It can be seen from Figures 6 and 7 that both ammonium bicarbonate and ethylenediamine phosphate had a positive effect on sulfide flotation of copper oxide ore. Addi- tionally, the dosage of ethylenediamine phosphate was obviously lower than that of ammonium bicarbonate. As low-dosage activators, the optimal dosages of ethylenediamine phosphate and ammonium bicarbonate were 150 g/t and 600 g/t, respectively. On the basis of a single-factor activator test, a dosage test was conducted on the combined activator (combination of amine and ammonium salts). The test selected a combination of amine and ammonium salts as the activator to investigate the flotation effect of copper oxide mineral. The dosage rates of the activator agent were 110, 220, 330, 440, 550, and 660 g/t,

Processes 2021, 9, 583 and the ratio between ammonium bicarbonate and ethylenediamine8 of 14 phosphate was 10:1. Furthermore, the collecting agent, vulcanizing agent, and foaming agent dosages were 500 g/t, 1200 g/t, and 100 g/t, respectively. The grinding fineness was −0.074 mm account- accordance with those obtained in the present study. Therefore, the dosage of 440 g/t was ingpreliminarily for 90%. selected The forresults the combined are shown activator in agent Figure to carry 8. out further investigations.

Processes 2021, 9, x FOR PEER REVIEW 8 of 13

g/t was preliminarily selected for the combined activator agent to carry out further investigations. Figure 8. Influences of the amine-ammonium combination activating dosage of the copper oxide 3.5.Figure Dosage 8. Influencesof the Inhibitor of the Agent amine -ammonium combination activating dosage of the copper oxide flotationflotation index. index. 3.5.On Dosage the of basis the Inhibitor of the Agent above test results, the ratio of sodium hexametaphosphate and sodium silicate was selected to be 1:1. The dosages were 100 g/t, 200 g/t, 300 g/t, 400 g/t, OnThe the results basis of the presented above test results, in Figures the ratio of6– sodium8 show hexametaphosphate that the dosage and for the combination of 500sodium g/t, and silicate 600 was g/t. selected Furthermore, to be 1:1. The the dosages grinding were 100fineness g/t, 200 was g/t, 300kept g/t, the 400 same g/t, as −0.074 mm accountingamine500 g/t, and for 600 ammonium 90%. g/t. Furthermore, The dosages salts the grindingofwas vulcanizing significantly fineness was agent, kept less combined the than same as that− activator0.074 for mm the agent, individual and col- salts (when lectingusedaccounting asagent activators). for were 90%. The1200 dosagesAdditionally, g/t, 440 of vulcanizingg/t, and the 500 agent,combined g/t, combined respectively. activation activator The agent, agent results and had are an shown obvious in effect on collecting agent were 1200 g/t, 440 g/t, and 500 g/t, respectively. The results are shown in FiguretheFigure flotation 9.9. recovery of copper and exhibited more stringent requirements for the dosage. When the dosage was low, the activation effect was not obvious. At a high dosage rate, an obvious inhibitory effect was observed on the mineral. Jiang and Mao [12,13] studied the enhancement of copper recovery from pure mineral malachite through experiments using the combined ammonia-ammonium activator. The results showed that the combined ammonia-ammonium activator (based on ethylenediamine phosphate and ammonium bicarbonate) exhibited better activation effects along with good synergistic effects. The results were in accordance with those obtained in the present study. Therefore, the dosage of 440

FigureFigure 9. 9. InfluencesInfluences of of the the inhibitor inhibitor dosages dosages of the copper of the oxide copper flotation oxide index. flotation index. It can be seen from Figure9 that the flotation recovery of the copper oxide mineral reachedIt can a maximumbe seen fro valuem ofFigure about 9 300 that g/t the for theflotation inhibitor. recovery The addition of the of inhibitor copper oxide mineral reachedcan improve a maximum the flotation value recovery of about of the copper300 g/t oxide for ore;the however,inhibitor. the The effect addition was quite of inhibitor can improvesmall. Furthermore, the flotation low and recovery high dosages of the of the copper inhibitor oxide showed ore; an adverse however, effect on the the effect was quite flotation recovery. small. Furthermore, low and high dosages of the inhibitor showed an adverse effect on the flotation recovery. The results of single-factor tests showed that by controlling the grinding fineness, collecting dosage, sodium sulfide dosage, combined activator agent dosage, and inhibitor dosage, better sulfide flotation indices of copper oxide ore can be achieved. Compared with the individual ammonium bicarbonate and ethylenediamine phosphate salts, the combined activator agent showed a better activating effect, and the dosage required was also lower than that of the individual ammonium bicarbonate. To optimize all of the conditions in a more reasonable way, the BP neural network and genetic algorithm were combined, and the results are discussed in the following section.

3.6. Prediction from the BP Neural Network Five variables, namely, the (a) grinding fineness, (b) combined activator agent dosage, (c) sodium sulfide dosage, (d) collecting agent dosage, and (e) inhibitor dosage were selected as the input layer neurons for the BP neural network. Based upon this, the recovery rate of the oxide copper flotation concentrate was estimated as the output factor. In this way, based on the flotation concentrate recovery rate and operating conditions, the BP neural network prediction model was established. The input node number of the model was 5, and the output node number was 1. It is worth mentioning that when the input layer node was N, the hidden layer nodes were chosen to be 2 N + 1 [26–28]. Under these conditions, the identified single hidden layer BP network could accurately reflect the actual situation and could guarantee the accuracy of the network. Therefore, the hidden layer neurons were set to be 11 [29,30], and the BP neural network structure was set to be 5-11-1, as shown in Figure 10.

Processes 2021, 9, 583 9 of 14

The results of single-factor tests showed that by controlling the grinding fineness, collecting dosage, sodium sulfide dosage, combined activator agent dosage, and inhibitor dosage, better sulfide flotation indices of copper oxide ore can be achieved. Compared with the individual ammonium bicarbonate and ethylenediamine phosphate salts, the combined activator agent showed a better activating effect, and the dosage required was also lower than that of the individual ammonium bicarbonate. To optimize all of the conditions in a more reasonable way, the BP neural network and genetic algorithm were combined, and the results are discussed in the following section.

3.6. Prediction from the BP Neural Network Five variables, namely, the (a) grinding fineness, (b) combined activator agent dosage, (c) sodium sulfide dosage, (d) collecting agent dosage, and (e) inhibitor dosage were selected as the input layer neurons for the BP neural network. Based upon this, the recovery rate of the oxide copper flotation concentrate was estimated as the output factor. In this way, based on the flotation concentrate recovery rate and operating conditions, the BP neural network prediction model was established. The input node number of the model was 5, and the output node number was 1. It is worth mentioning that when the input layer node was N, the hidden layer nodes were chosen to be 2 N + 1 [26–28]. Under these conditions, the identified single hidden layer BP network could accurately reflect the actual situation and could guarantee the accuracy of the network. Therefore, the hidden layer Processes 2021, 9, x FOR PEER REVIEW 9 of 13 neurons were set to be 11 [29,30], and the BP neural network structure was set to be 5-11-1, as shown in Figure 10.

Hidden layer(11)

a: Grinding fineness

s b: I oamyl xanthate Copper recovery c: Sodium sulfide Output layer(1)

d: A-A combination activator

e: Inhibitor

Input layer(5)

W W + + + + b b

FigureFigure 10.10. FlowFlow chartchart ofof thethe backback propagationpropagation (BP)(BP) neuralneural networknetwork model.model.

MATLABMATLAB (MathWorks,(MathWorks, Massachusetts, United United States) States) was was used used to tocarry carry out out the thenu- numericalmerical calculations calculations and and simulations simulations of ofthe the BP BP neural neural network network algorithm. algorithm. Furthermore, Furthermore, C Clanguage language was was used used for for programming programming and and to tocall call the the corresponding corresponding toolkit toolkit function. function. In the In theBP neural BP neural network network algorithm, algorithm, Equation Equation (1) in (1) the in input the input layer layer was used was usedto input to input the origi- the originalnal data datato carry to carry out the out process the process of normalization of normalization [31,32]. [31, 32]. XXX0 '−X = i i minmin XXi i  (1)(1) Xmax − Xmin XXmax min 0 where Xmax, Xmin and X respectively' represent the maximum, minimum, and experimen- where X max , X , andi X respectively represent the maximum, minimum, and ex- tal values of the originalmin samplei data. perimental values of the original sample data. The sigmoid function was used as the transfer function in the network hidden layer. The gradient descent method was used to train, and the “traingd” function was selected as the objective function. As shown in Figures 3–5, 8, 9, 30 groups of experimental data were used as the samples for the BP neural network model. These data samples were randomly sorted. Addi- tionally, 20 groups of data were randomly selected as the optimization samples for the BP neural network, and the remaining 10 groups of data were selected as the test samples used to test the network model. The results of the BP neural network calculations were stored for the next step of the genetic algorithm. The minimum error for the objective function was set to be 0.001 (for convergence), the iteration process was repeated 10,000 times, and the vector value was 0.05. The BP neural network algorithm was run. It was found that the fitting line of the copper recovery, as predicted by the BP neural network model, was consistent with the target line. The coefficients of deter-mination (R2) value was 0.998, as shown in Figure 11, showing good agreement between the experimental and predicted results. Figure 12 shows the values of error for the BP neural network predictions. After 4169 iterations, the optimization target had an error of only 0.001 and met the maximum error requirement. Figure 13 shows the predicted output values of the training performance (in graphical form).

Processes 2021, 9, 583 10 of 14

The sigmoid function was used as the transfer function in the network hidden layer. The gradient descent method was used to train, and the “traingd” function was selected as the objective function. As shown in Figures3–5,8 and9, 30 groups of experimental data were used as the samples for the BP neural network model. These data samples were randomly sorted. Ad- ditionally, 20 groups of data were randomly selected as the optimization samples for the BP neural network, and the remaining 10 groups of data were selected as the test samples used to test the network model. The results of the BP neural network calculations were stored for the next step of the genetic algorithm. The minimum error for the objective function was set to be 0.001 (for convergence), the iteration process was repeated 10,000 times, and the vector value was 0.05. The BP neural network algorithm was run. It was found that the ﬁtting line of the copper recovery, as predicted by the BP neural network model, was consistent with the target line. The coefficients of deter-mination (R2) value was 0.998, as shown in Figure 11, showing good agreement between the experimental and predicted results. Figure 12 shows the values of error for the BP neural network predictions. After 4169 iterations, the optimization target had an error of only 0.001 and met the maximum error requirement. Figure 13 shows the predicted output values of the training performance (in graphical form). To intuitively test the accuracy of predictions made by the BP neural network, the relationship between the copper recovery prediction rate and the actual value is shown in Figure 14. It can be seen from Figure 14 that the predicted copper recovery rate from the BP neural network model was consistent with the experimental value, showing accurate results obtained using the BP neural network model.

3.7. Genetic Algorithm Optimization The genetic algorithm model was used to optimize the data obtained from the BP Processes 2021, 9, x FOR PEER REVIEWneural network model. The input and output values were consistent with those of the 10 of 13 BP neural network, and the parameter design values are presented in Table2. After 100 generations, the ﬁtted results are shown in Figure 15.

Figure 11. 11.Fitting Fitting line of line the copperof the recovery copper predicted recovery by thepredicted BP neural by network the BP model. neural network model.

Figure 12. Mean squared error in the process of the BP neural network training.

Figure 13. Training performance of the back propagation neural network (BPNN) model.

To intuitively test the accuracy of predictions made by the BP neural network, the relationship between the copper recovery prediction rate and the actual value is shown in Figure 14. It can be seen from Figure 14 that the predicted copper recovery rate from the BP neural network model was consistent with the experimental value, showing accurate results obtained using the BP neural network model.

Processes 2021, 9, x FOR PEER REVIEW 10 of 13 Processes 2021, 9, x FOR PEER REVIEW 10 of 13

Processes 2021, 9, 583 11 of 14 Figure 11. Fitting line of the copper recovery predicted by the BP neural network model. Figure 11. Fitting line of the copper recovery predicted by the BP neural network model.

FigureFigure 12. 12.Mean Mean squared squared error in the error process in of the the BPprocess neural network of the training.BP neural network training. Figure 12. Mean squared error in the process of the BP neural network training.

Figure 13. Training performance of the back propagation neural network (BPNN) model. Figure 13. Training performance of the back propagation neural network (BPNN) model. Figure 13. Training performance of the back propagation neural network (BPNN) model. After genetic algorithm optimization, the results obtained were as follows: the maximum predictedTo intuitively output value test was the 87.62, accuracy and the independent of predictions variables a, made b, c, d, and by e hadthe BP neural network, the the valuesTo ofintuitively 91.7, 537.8, 1157.2, test 487.7,the accuracy and 298.9, respectively. of predictions To verify themade reliability by the of BP neural network, the relationshipthe optimization between results, the optimizationthe copper parameters recovery of theprediction selected model rate were and tested. the actual value is shown in relationship between the copper recovery prediction rate and the actual value is shown in FigureThe grinding 14. fineness,It can combinedbe seen activatorfrom Figur agent dosage,e 14 that sodium the sulfide predicted dosage, isoamylcopper recovery rate from the xanthateFigure dosage,14. It andcan inhibitor be seen dosage from were Figur 92%, 550e 14 g/t, that 1150 the g/t, 500predicted g/t, and 300 copper g/t, recovery rate from the BP neural network model was consistent with the experimental value, showing accurate respectively.BP neural A network recovery of 87.35%model was was achieved, consistent which corresponds with the to theexperimental grade of 2.68%. value, showing accurate resultsMoreover, obtained the flotation using recovery the was BP nearly neural 2% higher network than that model. before the optimization. Theresults results obtained showed that using the model the basedBP neural on the geneticnetwork algorithm model. and BP neural network, and when it was used for the cupric oxide flotation, it showed reliable accuracy and optimization ability. As a result, the optimized flotation conditions were accepted.

Processes 2021, 9, x FOR PEER REVIEW 11 of 13

Processes 2021, 9, x FOR PEER REVIEW 11 of 13 Processes 2021, 9, 583 12 of 14

Figure 14. Comparison of predicted values between BPNN and regression analysis.

3.7. Genetic Algorithm Optimization The genetic algorithm model was used to optimize the data obtained from the BP neural network model. The input and output values were consistent with those of the BP neural network, and the parameter design values are presented in Table 2. After 100 generations, the fitted results are shown in Figure 15.

FigureTableFigure 2. 14.14. ValuesComparison Comparison of genetic of predicted algorithmof predicted values betweenparameters values BPNN design. andbetween regression BPNN analysis. and regression analysis.

Table 2. Values of geneticParameters algorithm parameters design. Value 3.7. Genetic AlgorithmPopulation Optimization size 30 Parameters Value Mutation probability 0.2 The geneticPopulation algorithm size model was used to 30 optimize the data obtained from the BP neural networkMutationCrossover model. probability probability The input and output values 0.2 were0.4 consistent with those of the BP NumberCrossover probability of generations 0.4 100 neural network,Number of and generations the parameter design values 100 are presented in Table 2. After 100 gen- InitialInitial population population Random generatedRandom generated erations, the fitted results are shown in Figure 15.

Table 2. Values of genetic algorithm parameters design.

Parameters Value Population size 30 Mutation probability 0.2 Crossover probability 0.4 Number of generations 100 Initial population Random generated

Figure 15. Best fit convergence of genetic algorithm (GA)–BPNN. Figure 15. Best fit convergence of genetic algorithm (GA)–BPNN. 4. Conclusions The results of single-factor experiments showed that the activator agent had a good After genetic algorithm optimization, the results obtained were as follows: the max- activating effect on the sulfide flotation of copper oxide ore. Compared with the individual imumuse of eitherpredicted ammonium output bicarbonate value orwas ethylenediamine 87.62, and phosphate,the independent the combined variables activator a, b, c, d, and e hadagent the had values a better of activating 91.7, 537.8, effect, 1157.2, and the 487.7, required and dosage 298.9, was respectively. also lower than To that verify for the reliability ofthe the individual optimization activator. results, The combined the optimization activator agent parameters had obvious of effects the onselected the copper model were tested. flotation recovery in that it showed stringent dosage requirements. The grindingThe prediction fineness, results combined of the BP neural activator network agent model dosage, showed sodium that the Rsulfide2 value dosage, isoamyl xanthatewas 0.998, dosage, and the predictedand inhibitor results weredosage in accordance were 92%, with 550 the g/t, experimental 1150 g/t, values. 500 g/t, and 300 g/t, respectively. A recovery of 87.35% was achieved, which corresponds to the grade of 2.68%. Moreover, the flotation recovery was nearly 2% higher than that before the optimization. The results showed that the model based on the genetic algorithm and BP neural network,

Figure 15. Best fit convergence of genetic algorithm (GA)–BPNN.

After genetic algorithm optimization, the results obtained were as follows: the maximum predicted output value was 87.62, and the independent variables a, b, c, d, and e had the values of 91.7, 537.8, 1157.2, 487.7, and 298.9, respectively. To verify the reliability of the optimization results, the optimization parameters of the selected model were tested. The grinding fineness, combined activator agent dosage, sodium sulfide dosage, isoamyl xanthate dosage, and inhibitor dosage were 92%, 550 g/t, 1150 g/t, 500 g/t, and 300 g/t, respectively. A recovery of 87.35% was achieved, which corresponds to the grade of 2.68%. Moreover, the flotation recovery was nearly 2% higher than that before the optimization. The results showed that the model based on the genetic algorithm and BP neural network,

Processes 2021, 9, 583 13 of 14

After 4169 iterations, the objective function showed an error of only 0.001, which met the requirements of convergence. The predicted results obtained from the BP neural network model were consistent with the experimental values, and the model achieved high accuracy. The prediction precision was better than that of the regression analysis. The genetic algorithm of the GA–BPNN model was used to optimize the BP neural network model. After 100 generations, a maximum predicted output value of 87.62% was achieved. The optimization parameters of the selected model were tested. The grinding fineness, combined activator agent dosage, sodium sulfide dosage, isoamyl xanthate dosage, and inhibitor dosage were 92%, 550 g/t, 1150 g/t, 500 g/t, and 300 g/t, respectively. The flotation recovery of 87.35% was achieved, and the grade was 2.68%. Moreover, the flotation recovery was nearly 2% higher than that before the optimization. Furthermore, the GA–BPNN model achieved high prediction accuracy. Under the experimental conditions, this study only considered the fineness of grinding, amount of activator, amount of vulcanizing agent, amount of collector, and amount of inhibitor. The effects of pH and temperature shall be considered in a future study. Meanwhile, it is possible to establish a predictive optimization model for copper recovery. This study provides direction for future research in areas such as the particle swarm optimization algorithm and simulated annealing algorithm, which is equivalent to the neural network model.

Author Contributions: W.N. and S.W. conceived and designed the experiments; W.N., J.F., and Q.F. performed the experiments; Y.H. and X.Y. analyzed the data; W.N. wrote the paper; and Q.F. edited the paper. All authors have read and agreed to the published version of the manuscript. Funding: The authors would like to acknowledge the projects funded by Ten Thousand Talent Plans for Young Top-Notch Talents of Yunnan Province (Grant No. YNWR-QNBJ-2018-051) and Basic Research Program of Yunnan (Grant No.202001BA070001-107). Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Wang, Y.J.; Wen, S.M.; Feng, Q.C.; Xian, Y.J.; Liu, D. Leaching characteristics and mechanism of copper flotation tailings in sulfuric acid solution. Russ. J. Non-Ferr. Met. 2015, 56, 127–133. [CrossRef] 2. Feng, Q.C.; Wen, S.M.; Zhao, W.J.; Lv, C.; Bai, X. Leaching of Copper from Malachite with Methane-sulfonic Acid. Solvent. Extr. Res. Dev. 2015, 22, 159–168. [CrossRef] 3. Subrahmanyam, T.V.; Forssberg, K.S.E. Grinding and flotation pulp chemistry of a low grade copper ore. Miner. Eng. 1995, 8, 913–921. [CrossRef] 4. Lee, K.; Archibald, D.; Mclean, J.; Reuter, M.A. Flotation of mixed copper oxide and sulphide minerals with xanthate and hydroxamate collectors. Miner. Eng. 2009, 22, 395–401. [CrossRef] 5. Castro, S.; Goldfarb, J.; Laskowski, J. Sulphidizing reactions in the flotation of oxidized copper minerals, I. Chemical factors in the sulphidization of copper oxide. Int. J. Miner. Process. 1974, 1, 141–149. [CrossRef] 6. Zhang, C.; Jia, X.D.; Lu, Y.L.; Liu, D.W.; Zhang, X.L. Review on New Process and New Reagent of Copper Oxide. Adv. Mater. Res. 2014, 1010–1012, 1618–1621. [CrossRef] 7. Wu, D.D.; Ma, W.H.; Mao, Y.B.; Deng, J.S.; Wen, S.M. Enhanced sulfidation xanthate flotation of malachite using ammonium ions as activator. Sci. Rep. 2017, 7, 2086. [CrossRef][PubMed] 8. Feng, Q.C.; Zhao, W.J.; Wen, S.M.; Cao, Q.B. Copper sulfide species formed on malachite surfaces in relation to flotation. J. Ind. Eng. Chem. 2017, 48, 125–132. [CrossRef] 9. Deng, J.S.; Lei, Y.H.; Wen, S.M.; Chen, Z.X. Modeling interactions between ethyl xanthate and Cu/Fe ions using DFT/B3LYP approach. Int. J. Miner. Process. 2015, 140, 43–49. [CrossRef] 10. Jiang, T.G.; Fang, J.J.; Mao, Y.B.; Li, G.D.; Bi, K.J. Progress of Ammonium (Amine) Salton on Strengthen Sulfide Flotation for Oxidised Copper Ores. Conserv. Util. Miner. Resour. 2015, 3, 65–70. 11. Jiang, T.G.; Fang, J.J.; Zhang, T.M.; Mao, Y.B. Impact of Ammonium (Amine) Salts on the Sulfide Flotation Behavior of Oxidised Copper Ores. Conserv. Util. Miner. Resour. 2014, 1, 15–20. 12. Jiang, T.G.; Fang, J.J.; Mao, Y.B.; Li, G.D.; Bi, K.J. Impact of Ammonium (Amine) Salts on Behavior of Malachite Sulfide Flotation. Conserv. Util. Miner. Resour. 2015, 4, 31–37. 13. Mao, Y.B.; Deng, J.S.; Wen, S.M.; Fang, J.J. Reaction kinetics of malachite in ammonium carbamate solution. Chem. Pap. 2015, 69, 1187–1192. [CrossRef] Processes 2021, 9, 583 14 of 14

14. Koca, S.; Aksoy, D.O.; Cabuk, A.; Celik, P.A.; Sagol, E.; Toptas, Y.; Oluklulu, S.; Koca, H. Evaluation of combined lignite cleaning processes, flotation and microbial treatment, and its modelling by Box Behnken methodology. Fuel 2017, 192, 178–186. [CrossRef] 15. Wang, C.Q.; Wang, H.; Liu, Y.N.; Huang, L.L. Optimization of surface treatment for flotation separation of polyvinyl chloride and polyethylene terephthalate waste plastics using response surface methodology. J. Clean Prod. 2016, 139, 866–872. [CrossRef] 16. Jovanović,I.; Miljanović,I.; Jovanović,T. Soft computing-based modeling of flotation processes–A review. Miner. Eng. 2015, 84, 34–63. [CrossRef] 17. Allahkarami, E.; Nuri, O.S.; Abdollahzadeh, A.; Rezai, B.; Chegini, M. Estimation of Copper and Molybdenum Grades and Recoveries in the Industrial Flotation Plant Using the Artificial Neural Network. Int. J. Nonferr. Metall. 2016, 5, 23–32. [CrossRef] 18. Ai, M.; Xie, Y.; Xie, S.; Zhang, J.; Gui, W. Fuzzy association rule-based set-point adaptive optimization and control for the flotation process. Neural. Comput. Appl. 2020, 32, 14019–14029. [CrossRef] 19. Fu, Y.; Aldrich, C. Flotation froth image recognition with convolutional neural networks. Miner. Eng. 2019, 132, 183–190. [CrossRef] 20. Zhang, H.; Tang, Z.; Xie, Y.; Chen, Q.; Gui, W. FR-R net: A Light-weight deep neural network for performance monitoring in the froth flotation. IEEE Trans. Ind. Infor. 2020. 21. Baliyan, A.; Gaurav, K.; Mishra, S.K. A Review of Short Term Load Forecasting using Artificial Neural Network Models. Procedia. Comput. Sci. 2015, 48, 121–125. [CrossRef] 22. Li, H.Q.; Yuan, D.Y.; Ma, X.D.; Cui, D.Y.; Cao, L. Genetic algorithm for the optimization of features and neural networks in ECG signals classification. Sci. Rep. 2017, 7, 41011. [CrossRef] 23. Allahkarami, E.; Nuri, O.S.; Abdollahzadeh, A.; Rezai, B.; Maghsoudi, B. Improving estimation accuracy of metallurgical performance of industrial flotation process by using hybrid genetic algorithm-artificial neural network (GA-ANN). Physicochem. Probl. Miner. Process. 2016, 53, 366–378. 24. Dong, S.; Wang, B.; Wang, Z.; Hu, X.K.; Song, H.C.; Liu, Q. Comparison of Prediction Models for Power Draw in Grinding and Flotation Processes in a Gold Treatment Plant. J. Chem. Eng. Jpn. 2016, 49, 204–210. [CrossRef] 25. Tang, Z.H.; Liu, J.P.; Chen, Q.; Gui, W.H.; Du, J.F. pH control in flotation process based on prediction model. Control. Theor. Appl. 2013, 30, 885–890. 26. Zhao, M.J.; Fang, J.J.; Zhang, L.; Dai, Z.; Yao, Z.W. Improving the Estimation Accuracy of Copper Oxide Leaching in an Ammonia–Ammonium System Using RSM and GA-BPNN. Russ. J. Non-Ferr. Met. 2017, 58, 591–599. [CrossRef] 27. Zarie, M.; Jahedsaravani, A.; Massinaei, M. Flotation froth image classification using convolutional neural networks. Miner. Eng. 2020, 155, 106443. [CrossRef] 28. Jamróz, D.; Niedoba, T.; Pita, P.; Surowiak, A. The Use of Neural Networks in Combination with Evolutionary Algorithms to Optimise the Copper Flotation Enrichment Process. Appl. Sci. 2020, 10, 3119. [CrossRef] 29. Jorjani, E.; Mesroghli, S.; Chelgani, S.C. Prediction of operational parameters effect on coal flotation using artificial neural network. J. Univ. Sci. Technol. Beijing Miner. Metall. Mater. 2008, 15, 528–533. [CrossRef] 30. Jorjani, E.; Poorali, H.A.; Sam, A.; Chelgani, S.C.; Mesroghli, S.; Shayestehfar, M.R. Prediction of coal response to froth flotation based on coal analysis using regression and artificial neural network. Miner. Eng. 2009, 22, 970–976. [CrossRef] 31. Zhao, M.J.; Fang, J.J.; Zhang, L.; Dai, Z.; Yao, Z.W. Application of Artificial Neural Networks in Flotation Processes. Chin. J. Process. Eng. 2017, 17, 440–446. 32. Yao, Z.W.; Fang, J.J.; Zhao, M.J.; Zhang, L.; Dai, Z.; Zhang, L.; Zheng, R.H. Review on Application of Multivariable Optimization and Decision-making Model in Flotation Process. Bull. Chin. Ceramic. Soc. 2018, 37, 2463–2470.