dc_888_14 Appraisal and Development of Transportation Systems Using Multiple Criteria Decision Making Methodology D.Sc. Dissertation In partial fulfillment of the requirements for the title of Doctor of the Hungarian Academy of Sciences Farkas, András Budapest 2014 dc_888_14 Table of Contents i Table of Contents Abbreviations . iv Preface .............................................................. 1 Chapter 1 . 2 1 On the Development of Transportation Sysems . 2 1.1 Basic Concepts, Notions and Characteristics of Transportation Systems . 2 1.2 Directions of Transport Policy Developments in the EU and in Hungary . 6 1.3 A Sequential Transport Policy (STP) Model for Sustainable Transport . 7 1.4 Transportation Systems Design and Related Decision-Making . 13 Application 1: An Intelligent GIS-based planning of a metro-rail network . 18 Chapter 2 . 24 2 Multiple Criteria Decision Making (MCDM) Methods . 24 2.1 Taxonomy of MCDM Methods and their Applications to Transportation and Civil Engineering Projects . 24 2.2 The Analytic Hierarchy Process (AHP) with Applications . 25 Application 2: Evaluation and selection of a bridge design using the AHP method ............................................................... 29 Application 3: Effect of prestressing on the appearance of concrete structures . 33 Chapter 3 . 35 3 On the Development of the Analytic Hierarchy Process (AHP) . 35 3.1 Preliminaries . 35 3.2 Transitive and Symmetrically Reciprocal Matrices . 35 3.3 The Spectrum of Certain SR Perturbations of Transitive Matrices . 37 3.4 The Issue of Rank Reversal . 42 Application 4: On the priority ranking problem of the AHP . 45 Application 5: Spectrum of the matrix of economic growth in the dynamic input-output analysis of macroeconomics . 47 Application 6: A vehicle system dynamics problem considering n-axle railway carriages . 47 Chapter 4 . 53 4 Consistency Adjustments to Pairwise Comparison Matrices . 53 4.1 Preliminaries and Problem Statement . 53 4.2 The Linear Approximation . 54 dc_888_14 Table of Contents ii 4.3 The Nonlinear Problem . 57 4.3.1 Numerical Illustration . 59 4.4 On the Non-uniqueness of the Solution to the Nonlinear Problem . 60 4.5 SR Matrices with perturbations of Exponential Type . 63 Application 7: Generating the spectrum of perturbed input spectral density matrices . 65 Chapter 5 . 68 5 Balancing SR Matrices by Transitive Matrices . 68 5.1 A Recursive Rank-one Residue Iteration . 68 5.2 Diagonal Similarity Scaling of Pairwise Comparison Matrices . 70 5.3 Numerical Analysis . 74 Chapter 6 . 78 6 Development of a Combined MOO/MCDA Scaling Method . 78 6.1 Preliminaries . 78 6.2 Formal Description of the Method MAROM . 79 Application 8: Evaluation, ranking and a scenario of alternative-fuel modes of buses . 82 Summary and Theses . 89 Thesis1 ...................................................................... 90 Thesis2 ...................................................................... 90 Thesis3 ...................................................................... 91 Thesis4 ...................................................................... 92 Thesis5 ...................................................................... 93 Thesis6 ...................................................................... 94 Bibliography . 96 References of author’s publications . 96 References of related publications . 98 Appendix A Summary of Multi-Criteria Decision Making (MCDM) Methods with Applications to Transportation/Civil Engineering Problems . .105 A.1 Multi-Objective Optimization (MOO) Methods with Some Applications to Transportation/Civil Engineering Problems . 105 A.1.1 MOO methods with a’priori specification of preferences and applications . 108 A.1.2 MOO methods with a’posteriori specification of preferences and applications . 110 A.1.3 MOO methods with no specification of preferences and applications . 111 dc_888_14 Table of Contents iii A.2 Multi-Criteria Decision Analysis (MCDA) Methods with Some Applications to Transportation/Civil Engineering Problems . 113 A.2.1 Definitions of the basic terms used in the MCDA framework . 114 A.2.2 Non-compensatory MCDA methods with applications . 116 A.2.3 Compensatory MCDA methods with applications . 117 A.2.4 Prioritization methods (Scaling methods) with applications . 120 P Appendix B Derivation of the characteristic polynomial, pn (λ), of the perturbed PCM, Ap .....................................131 Appendix C Development of the principal eigenvector of the simple perturbed PCM, AS .....................................132 Appendix D Deriving the Spectrum of Augmented Pairwise Comparison Matrices . .134 D.1 Simple Perturbed Pairwise Comparison Matrices of Augmented Form . .134 D.2 The Issue of Rank Reversal . 136 D.3 An Illustrative Sample Example . 137 D.4 Theoretical Derivations for Finding the Principal Right Eigenvector of AB . 138 Appendix E Recursive Least-Squares Algorithm for SR matrices . .142 dc_888_14 Abbreviations iv Abbreviations AFV = alternative-fuel vehicle AHP = analytic hierarchy process CBA = cost-benefit analysis CD = conventional diesel engine CI = consistency index CNG = compressed natural gas CR = consistency ratio CTP = Common Transport Policy DM = decision maker DSS = decision support system ECMT = European Conference of Ministers of Transport EV = electric vehicle FCV = fuel-cell vehicle GIS = Geographic Information System GPS = Global Positioning System GSM = Global System for Mobile HEV = hybrid electric vehicle HSPC = high strength prestressed concrete HTP = Hungarian Transport Policy ICEV = internal combustion engine vehicle ILWIS = Integrated Land and Water Information System ITS = Intelligent Transportation System KSIM = Kane simulation LPG = liquefied propane gas LS = least-squares LSR = least-squares recursion MADA = multi-attribute decision analysis MAROM = multi attribute object measurement MCDM = multi-criteria decision making MOO = multi-objective optimization MSU = mean spatial utility NHDP = National Hungarian Development Policy N-K = Newton-Kantorovich PC = prestressed concrete PCM = pairwise comparison matrix PM = particulate matter RC = reinforced concrete SDM = spectral density matrix SMCE = spatial multiple criteria evaluation SPI = symmetric permutation invariant SR = symmetrically reciprocal STP = Sequential Transport Policy TEN-T = Trans-European Network - Transportation TOPSIS = technique for order preference by similarity to ideal solution Triple R-I = recursive rank-one residue iteration UTDS = Unified Transport Development Strategy UTM = Universal Transverse Mercator dc_888_14 Farkas, András Preface 1 Preface This work summarizes the results of the scholarly research I have conducted in the fields of transportation systems engineering and multi-criteria decision making (MCDM) during the last twenty years. My major goals were to reveal and to give exact explanations of a mathematical type on some known shortcomings of existing scaling methods and developing new techniques while applying them to a variety of transportation problems. Central to my interest was to look at these methods with a healthy skepticism. Another objective of this thesis was to stimulate new applications of MCDM by specialists and non-specialists through providing many uses. MCDM methods have become a tool that must be understood not only by the research engineer, operations research analyst and mathematician, but also by the transport planner, geographer, regional expert, civil engineer, public servant and other people without extensive mathematical backgrounds. On the other hand, a pragmatic approach is not enough to support the analysis of such complicated systems. In my view, findings which are presented as a mixture of theoretically based propositions (with proofs) and propositions based on numerical calculations, or only on speculations cannot be sufficient. Therefore, when I was writing the text, I followed solid mathematical rigor, as well as precision in the numerical computations. An implicit goal was to move the practitioners to the computer and to an intensive use of the huge opportunities provided by the internet. Additionally, I tried to motivate the users to apply the ever-growing capabilities of intelligent transportation systems such as, for example, the geographic informa- tion systems (GIS). Similarly, I have placed a large emphasis on the question of sustainable transport, as well as the inclusion of the stakeholders (transport providers, services and users) in the decision making processes and the preparation of new projects using multi-criteria deci- sion analysis (MCDA) methods in such complex areas like transport policy, transport planning, transport design and evaluation of new project plans. Several people have contributed to this work. My distinguished appreciation goes to my dear colleague and friend, the late professor Pál Rózsa, whom I had been working very hard with on these exciting topics over the past twenty years. He has made great achievements concerning some complex mathematical derivations. His intellectual prowess was matched by his genuine character and humanity. I felt privileged to have had this long opportunity to work with him. His valuable remarks have improved the clarity of presentation of our joint papers in a large extent. I wish to express my deepest gratitude to my international collaborators also, to professor Peter Lancaster of University of Calgary, holder of the Hans Schneider Prize, ILAS, who has made significant contributions to the development of both the linear and the non-linear models