Efficiency Improvement of Regenerative Energy for an EV

Takuya Yabe 1, Kan Akatsu 1, Nobunori Okui2 , Tetsuya Niikuni2 and Terunao Kawai2 1 Shibaura Institute of Technology 2 National Traffic Safety and Environment Laboratory

Abstract Electric Vehicles (EVs) and various Hybrid Electric Vehicles (HEVs) have been attracting a lot of attention for environmental issues and energy crisis. One of advantages of using foregoing vehicles is charging energy by the regenerative brake. The running distance by one electric charge is increased a lot by the regenerative brake. However, the absorbed capacity of the regenerative energy is limited because of the motor capacity and the current limit of the battery. As a result not only the regenerative electric brake but also the mechanical brake must be used. This becomes serious issue in the heavy weight vehicle such as the bus and the truck, the effectiveness of EV/HEV is not obtained. To increase the regenerative energy, the large motor and the battery are requested, however, it is very difficult because of the cost and the limit of the inverter capacity. In this paper, it is verified that the regenerative energy is increased by improving a braking method, averaging the deceleration, without changing the power train system. The proposed method is experimentally evaluated by i-MiEV on the dynamo system, and increases the regenerative energy to 18%. .

Keywords: EV (electric vehicle), HEV (hybrid electric vehicle), regenerative brake,

and the current limit of the battery. As a result not 1 Introduction only the regenerative electric brake but also the mechanical brake must be used. The trend Recent warning of the exhausted fuel source becomes serious for the heavy weight vehicles, the and the global warming are serious problems in bus and the truck and so on. the world. EVs and HEVs are expected to reduce However, in the research area of the brake system CO2 and almost every car manufactures are for EVs and HEVs, studies have been mostly developing them. Although EVs are attractive to focused on controlling the braking force to achieve reduce CO2, it should increase the running the optimal braking performance [1]-[3]. Recent distance by one electric charge to expand its some studies have dealt with the optimal design of market. To solve this problem, recent the brake system for EVs and HEVs [4]-[6]. commercial EVs can absorb the regenerative However, few publications have experimentally energy as much as they can. verified an improvement of the regenerative The brake system of EVs and HEVs is shown in energy by the brake system. Figure 1. A regenerative brake is achieved by using the traction motor which generates the negative torque in vehicle deceleration mode, it battery can convert the kinetic energy into the electric inverter energy and it is charged to a battery, however, a general mechanical brake ingenerates the brake signal regenerative motor brake stopping force by overbearing brake disk to brake split model machine hydraulic brake pad, the braking energy is converted into foot brake brake pressure the heat and it is not reused. Then, the regenerative brake has little detriment and the running distance by one electric charge is Figure 1: Brake system increased by the regenerative brake. However, the possible absorbed energy of the regenerative In this paper, the reason why the regenerative energy is limited because of the motor capacity energy is limited is firstly shown, and an improve

speed command method is proposed. The proposed method Tref slip ratio r + + 1 ω ωr-v speed r slip increases the regenerative energy by improving a controller - - Js ωr breaking method. It is experimentally evaluated B by actual i-MiEV on the dynamo system. The r + 1 proposed method is possible to use to improve - Ms the regenerative energy in the heavy weight Cv 2 vehicles. μ + x Mg Cu e Ca slip -

Cb slip 2 Simulation model First, a vehicle simulation model is created by Figure 2: Vehicle model Matlab/Simulink which is shown in Figure 2 to make clear the effect of the motor capacity and Demand torque Regenerative the battery current to the regenerating energy. In brake this paper, a simple model that the vehicle Vehicle Brake split model model Mechanical differential equation is connected with the motor Demand power differential equation by the slip is made to brake understand the basis of phenomenon. Where, it assumes the ideal condition, that is, a bobbing Figure 3: Brake model motion, a pitching movement and a rolling movement are not taken into account and it is The brake model is shown in Figure 3. The brake also assumed that the vehicle has one motor force is divided into the mechanical brake and the connected with the gear which ratio is selectable. regenerative brake by the brake split model. The The motor differential equation is defined as mechanical brake force is defined as a value follow. subtracts the regenerative brake force from the d demand brake command. J T Fr  B (1) dt e Where J is the inertia force,  is motor angular 3 Limitation of the regenerative speed, F is driving force, r is wheel radius and B brake is friction factor. The regenerative brake torque is calculated by the The vehicle motion differential equation is vehicle and the brake models. To make sure of the defined as follow. limitation of the regenerative brake, a heavy HEV dv model is calculated, and its parameter is shown in M  FF r (2) dt Table I. Where M is the vehicle mass, v is vehicle speed, Fr is friction force, for example air friction and Table I: Vehicle parameter the rolling resistance. The friction factor between Vehicle mass (t) 6 the road and the wheel associates equation (1) Motor max torque (Nm) 350 with (2). It is defined as equation (3) by the slip ratio. The slip ratio is defined as equation (4). Motor max power (kW) 35

 u SCC la exp()(exp(  SC lb )) (3) final gear ratio 5.428   vr S  (4) l  r Figure 4 shows a calculated result of the torque distribution between the mechanical brake and the Where, Cu, Ca and Cb are coefficients to determine the friction factor. The driving force is electric regenerative brake when the heavy HEV is defined as follow. decelerated from 100 km/h by 0.1G constant F  Mg (5) deceleration. The electric regenerative brake is limited because of the motor capacity, 35kW, at The simple vehicle simulation model is made the high speed range, adding that even in the low from the equation (1)-(5). The demand torque speed range it is limited because of the input and power are calculated by inputting a speed current limit of the battery. As a result, less than reference to the speed controller and the torque half of the braking energy is absorbed by the reference is outputted to the vehicle model in regenerative brake. which the output torque is equal to the reference.

battery. At the high SOC condition the current 120 9000 input density is decreased. As a result the Vehicle speed Demand torque 8000 regenerative brake is limited by the maximum 100 Regenerative torque 7000 battery current. Mechanical torque 80 6000 Especially for the heavy EVs, HEVs like a bus

5000 and a truck, it is very hard to increase the battery 60 4000 capacity because of a trade off between the weight Motor power 35kW 40 3000 and the power. Also the motor capacity is hard to Brake torque [Nm] torque Brake Vehicle speed [km/h] speed Vehicle 2000 increase because of the cost and the limit of the 20 1000 inverter capacity.

0 0 0 5 10 15 20 25 30 Time [s] 4 A proposed method to increase

Figure 4: Distribution of the mechanical and the the amount of the regenerative regenerative brake energy Regenerative torque The regenerative brake of EVs, HEVs is limited 2500 Regenerative torque 10kW Limitation of battery input current 100[A] because of the motor capacity and the current limit Limitation of battery input current 50[A] 2000 of the battery as described above section. As a result not only the regenerative electric brake but 1500 also the mechanical brake must be used. To increase the regenerative energy, the motor 1000 capacity and the battery capacity are requested to Brake torque [Nm] torque Brake be large, however it is very difficult because of the 500 cost and the limit of the inverter capacity. Therefore, in this paper, the regenerative energy 0 0 5 10 15 20 25 30 is increased by improving a breaking method Time [s] Figure 5: Regenerative brake torque of each capacity of without changing the power train system. the motor and the maximum input current into battery

4.1 Averaging the deceleration method Output As an example JC08 mode which is used for the fuel consumption measurement in Japan is used as shown in Figure 7. The proposed method is Input averaging the deceleration when the car is decelerated. Figure 8 shows a comparison of the deceleration between the JC08 model and the

Power density [W/kg] density Power proposed model. The deceleration of the proposed model is smaller than the JC08 model by averaging 0 50 100 the deceleration. SOC [%] Figure 9, 10 and 11 show the comparison results Figure 6: Input / Output property of battery of the vehicle speed, the deceleration and the demand brake power, respectively. Figure 10 Figure 5 shows the study of the motor capacity shows the deceleration of the proposed model is and the battery input current limit. From this half of the maximum value of the original model figure, it is observed that not only the motor by averaging the deceleration. As a result not only capacity but also the battery input current effect the deceleration but also the max demand power is to the electrical brake limit. reduced two-thirds value. As a result the The maximum battery current is defined as regenerative energy of the proposed model is follow. larger than the JC08 model even when the MP bb regenerative brake is limited by the input current I .max bat  (6) Vbat limit of the battery. Averaging the deceleration Where, Imax,bat is the maximum battery current, method can be achieved by using a driving Pb is battery current density, Mb is battery mass assistance tool, for example, the deceleration per one cell and Vbat is battery voltage per one signal is shown on a display with the navigation sell. Figure 6 shows Input / Output property of which gives the advice to the driver on the optimal

30 vehicle operation [7], or controlling a brake pedal JC08 by applying a brake system like an active 20 averaging the deceleration accelerator pedal, which indicates the ratio of 10 acceleration of the driver [8]. 0 1010 1060 1110 1160 1210 -10

90 -20 Demand power [kW] power Demand 80 -30 70 -40 60 Time[s] 50 Figure 11: The comparison of the demand brake 40 power 30 Vehicle speed [km/h] speed Vehicle 20 10 0 4.2 Experimental equipments 0 200 400 600 800 1000 1200 Time[s] In the experiment to evaluate the proposed

Figure 7: JC08mode method, “i-MiEV” is used to measure the regenerative electric power at the real condition. -0.01 0 200 400 600 800 1000 1200 Table II shows the vehicle parameter, and both -0.03 Figure 12 and 13 show the experimental -0.05 equipments. The proposed method is evaluated by -0.07 running i-MiEV on the dynamo system. The -0.09

Deceleration [G] Deceleration experiment by the dynamo system is taken account -0.11 of the air resistance and the rolling friction -0.13 JC08 averaging the deceleration resistance by pre-measured. -0.15 Time [s]

Figure 8: Deceleration Table II: Vehicle parameter 100 JC08 Vehicle mass (kg) 1110 90 averaging the deceleration 80 Width (mm) 1475 70 Height (mm) 1610 60 50 Battery voltage (V) 330 40 30 Motor max power (kW) 47 Vehicle speed [km/h] speed Vehicle 20 Motor max torque (Nm) 180 10 0 Final gear ratio 6.066 1010 1060 1110 1160 1210 Time [s] Wheel radius (m) 0.2865 Figure 9: The comparison of the vehicle speed

0.20 JC08 0.15 averaging the deceleration 0.10

0.05

0.00 1010 1060 1110 1160 1210

Deceleration [G] Deceleration -0.05

-0.10

-0.15 Time [s]

Figure 10: The comparison of the deceleration

Figure 12: “i-MiEV” on the dynamo system

140 JC08 120 proposed method 100

20 Motor torque [Nm] torque Motor

0 0 200 400 600 800 1000 1200 -20

-40 Time [s]

Figure 16: Motor torque Figure 13: Battery of “i-MiEV” 370 365 4.3 Experimental results 360 4.3.1 Experimental result 355 Figure 15 shows the motor current, the 350 345 accelerating current is larger than the [V] voltage Battery 340 decelerating one because the limitation of the JC08 335 regenerative torque. It is confirmed that the proposed method 330 regenerative torque is limited to about 20 Nm in 0 200 400 600 800 1000 1200 Time [s] Figure 16. The detail is shown in the next section. Figure 17: Battery voltage Figure 17 shows the battery voltage of both methods. The voltage in the proposed model is 25 smaller than it in the JC08 model because of the 20 influence of the SOC condition. 15 10 5 90 0 JC08 80 0 200 400 600 800 1000 1200

Motor power [kW] power Motor -5 proposed method 70 -10 60 JC08 -15 proposed method 50 -20 Time [s] 40 30 Figure 18: Motor power Vehicle speed [km/h] speed Vehicle 20

10 4.3.2 Limitation of the regenerative brake 0 0 200 400 600 800 1000 1200 The experimented result of the motor torque with Time [s] decelerating is shown in Figure 19. The Figure 14: Vehicle speed regenerative torque is limited to 24 Nm. This is because “i-MiEV” does not control the 70 regeneration brake associated with the mechanical

50 brake, that results difficulty of the precise braking control, it means the regeneration brake is worked 30 every when the driver release the acceleration pedal. Then to keep the safety also to avoid much 10 de-acceleration the regenerative torque is limited. 0 200 400 600 800 1000 1200 -10

Motor current [A] current Motor Figure 20 shows the comparison between the JC08 demand power and the regenerative power. Less prpposed method -30 than half of the braking energy is absorbed by the -50 regenerative brake. As a result the mechanical Time [s] brake is much used. Figure 15: Motor current

0 4.3.4 Comparison of the regenerative energy 0 1000 2000 3000 4000 5000 The motor power is compared in each running -5 modes. Figure 22 shows an expansion of the motor -10 input/output power comparison between the proposed model and the original model. The -15 regenerative power in the proposal model is much -20 absorbed than in JC08 mode especially when the deceleration is changed with smooth de- Regenerative torque [Nm] torque Regenerative -25 acceleration. -30 Figure 23 shows a comparison of the regenerative Motor speed [rpm] energy. Three times experimental results are Figure 19: Regenerative braking torque of motor averaged and compared with the original model. The regenerative energy is increased 18% by the 0 0 500 1000 proposed averaging deceleration method. -5

-10 Motor power JC08 30 Motor power proposal model 90 Vehicle speed JC08 -15 Vehicle speed proposal model 80 20 -20 70 10 60

Brake power [kW] power Brake -25 50 0 -30 demand brake power 1010 1060 1110 1160 40 regenerative power -10 30

-35 power[kW] Motor Time [s] speed[km/h] Vehicle 20 Figure 20: Comparison of demand power and -20 10 regenerative power -30 0 Time[s]

Figure 22: Comparison of motor power 4.3.3 Regenerative limitation in low speed range 850 Figure 21 shows the regenerative limit in the 18% low speed range. From this figure it is confirmed 800 that the regeneration is completely terminated , increase the current becomes zero under the speed is 750 bellow 16km/h. As described above, that is because the regeneration control in the low speed 700 range is sensitive, adding that the motor voltage 650 becomes small in the low speed region. Regenerative Regenerative energy [kWs] 600 60 365 JC08 Proposal model speed current 363 Figure 23: Comparison of the amounts of the absorbed 40 voltage 361 regenerative electric power 16km/h 359 20 357 0 355 1160 1180 1200 1220 353 5 Conclusion -20 Battery voltage [V] voltage Battery Motor current [A] current Motor Limitation of 351 Vehicle speed [km/h] speed Vehicle In this paper, the reason why the regenerative regenerative 349 -40 energy is limited is firstly shown, and an improve 347 method to absorb the regenerative energy is -60 345 Time [s] proposed. The proposed method is experimentally evaluated by i-MiEV on the dynamo system, and

Figure 21: Limitation of regenerative brake in low speed increases the regenerative energy 18%. The range proposed method is realized by the driving operation, the regenerative energy is much increased, for example it is effective to adapt to the

public bus because a bus lane allows the optimal deceleration to absorb the regenerative energy.

References [1] Chunting Mi, Hui Lin and Yi Zhang, "Iterative Learning Control of Antilock Barking of Electric and Hybrid Vehicles", IEEE Transaction on Vehicular Technology, Vol. 54, No. 2, Sep. 2005 [2] Nobuyoshi Mutoh, Hiromichi Yahagi, "Electrical Braking Control Methods for Electric vehicles with Independently Driven Front and Rear Wheel Structure", Industrial Electronics, IEEE transaction on, Vol. 54, Issue 2. April 2007 [3] Li Yu-shan, Zeng Qing-liang, Wang Cheng- long, Wang Liang, "Research on Control strategy for Regenerative Braking of a Plug in Hybrid Electric City Public Bus", IEEE Intelligent Computation Technology and Automation, Oct. 2009 [4] Yimin Gao, Liang Chu, Mehrdad Ehsani, "Design and Control Principles of Hybrid Braking System for EV, HEV and FCV", IEEE Vehicle Power and Propulsion Conference, Sept. 2007, [5] Liang Chu, Wanfeng Sun, Liang Yao, Yongsheng Zhang, Yang Ou, Wenruo Wei, Minghui Liu, Jun Li, "Integrative control strategy of regenerative and hydraulic braking for hybrid electric car ", IEEE Vehicle Power and Propulsion Conference, Oct. 2009, [6] S. R. Clikanek, K. E. Bailey, "Regenerative Braking System for A Hybrid Electric Vehicle", Proceedings of the American Control Conference, Anchorage, AK May, 2002, [7] M. van der Voort, M. Dougherty, "Reducing fuel consumption by using a new fuel-efficiency support tool", IEEE Africon, Sep.1999, [8] A. Varhelyi, M. Hjalmdahl, C. Hyden, and M. Draskoczy, "Effects of an active accelerator pedal on driver behaviour and trafﬁc safety after long-term use in urban areas", Accident Analysis and Prevention, vol. 36 no. 5, pp. 729–737, 2004,

EVS26 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium 7