Applications Of Multi-Objective Evolutionary Algorithms (Advances in Natural Computation)

9812561064 ......Page 1
FOREWORD......Page 6
PREFACE......Page 8
CONTENTS......Page 12
1.1. Introduction......Page 30
1.2. Basic Concepts......Page 32
1.3. Basic Operation of a MOEA......Page 33
1.4.1. Aggregating Functions......Page 35
1.4.2. Population-Based Approaches......Page 36
1.4.3. Pareto-Based Approaches......Page 37
1.5. MOEA Performance Measures......Page 40
1.6. Design of MOEA Experiments......Page 43
1.6.1. Reporting MOEA Computational Results......Page 44
1.7.1. Part I: Engineering Applications......Page 45
1.7.2. Part II: Scientific Applications......Page 48
1.7.3. Part III: Industrial Applications......Page 49
1.7.4. Part IV: Miscellaneous Applications......Page 50
1.8. General Comments......Page 51
References......Page 52
2.1. Introduction......Page 58
2.2. Multi-Objective Evolutionary Algorithm......Page 60
2.2.1. Algorithms......Page 62
2.3.1. Design of a Welded Beam......Page 65
2.3.2. Preliminary Design of Bulk Carrier......Page 69
2.3.3. Design of Robust Airfoil......Page 75
2.4. Summary and Conclusions......Page 79
References......Page 81
3.1. Introduction......Page 82
3.2. The Algorithms......Page 83
3.2.1. Non-Dominated Sorting Evolution Strategy Algorithm (NSESA)......Page 84
3.3. Case Studies......Page 90
3.3.1. Shape Design of a Shielded Reactor......Page 91
3.3.2. Shape Design of an Inductor for Transverse-Flux-Heating of a Non-Ferromagnetic Strip......Page 98
References......Page 104
4.1. Introduction......Page 108
4.2. Prior Work......Page 110
4.3.2. Problem Formulation......Page 112
4.4. Overview of the -NSGA-II Approach......Page 113
4.4.1. Searching with the NSGA-II......Page 115
4.4.2. Archive Update......Page 116
4.4.3. Injection and Termination......Page 118
4.5. Results......Page 120
4.7. Conclusions......Page 126
References......Page 127
5.1. Introduction......Page 130
5.2. Problem Statement......Page 131
5.3. Our Proposed Approach......Page 133
5.4. Use of a Multi-Objective Approach......Page 136
5.5.1. Example 1......Page 138
5.5.2. Example 2......Page 139
5.5.3. Example 3......Page 141
5.5.4. Example 4......Page 143
5.5.5. Example 5......Page 146
5.5.6. Example 6......Page 147
5.6. Conclusions and Future Work......Page 149
References......Page 151
CHAPTER 6 APPLICATION OF MULTI-OBJECTIVE EVOLUTIONARY ALGORITHMS IN AUTONOMOUS VEHICLES NAVIGATION......Page 154
6.1. Introduction......Page 155
6.2.1. Experimental Setup......Page 156
6.2.2. Vehicle Model......Page 157
6.2.3. Relative Sensor Models......Page 158
6.2.4. Absolute Sensor Models......Page 159
6.2.6. Prediction of the Vehicle State......Page 160
6.3.1. Problem Formulation......Page 162
6.3.2. A General Framework for Searching Pareto-Optimal Solutions......Page 163
6.3.3. Selection of a Single Solution by CoGM......Page 165
6.4.1. Evaluation of Functions......Page 167
6.4.2. Search Methods......Page 168
6.5. Application of Parameter Identification of an Autonomous Vehicle......Page 170
6.6. Conclusions......Page 177
References......Page 180
7.1. Introduction......Page 184
7.2. Performance Based Design Unification and Automation......Page 187
7.2.1. The Overall Design Architecture......Page 188
7.2.2. Control System Formulation......Page 189
7.2.3. Performance Specifications......Page 190
7.3. An Evolutionary ULTIC Design Application......Page 194
7.4. Conclusions......Page 201
References......Page 203
8.1. Introduction......Page 206
8.2.1. Single Screw Extrusion......Page 207
8.2.2. Co-Rotating Twin-Screw Extrusion......Page 208
8.2.3. Optimization Characteristics......Page 212
8.3.1. Multi-Objective Optimization......Page 213
8.3.2. Reduced Pareto Set Genetic Algorithm with Elitism (RPSGAe)......Page 215
8.3.3. Travelling Salesman Problem......Page 216
8.4.1. Single Screw Extrusion......Page 218
8.4.2. Twin-Screw Extrusion......Page 223
8.5. Conclusions......Page 225
References......Page 226
9.1. Introduction......Page 230
9.2. Related Work......Page 231
9.3. ISPAES Algorithm......Page 233
9.3.2. Shrinking the Objective Space......Page 236
9.4.1. Optimization of a 49-bar Plane Truss......Page 241
9.4.2. Optimization of a 10-bar Plane Truss......Page 244
9.4.3. Optimization of a 72-bar 3D Structure......Page 246
9.5. Final Remarks and Future Work......Page 251
References......Page 252
10.1. The Traditional Approach......Page 256
10.2. The MOEA Approach......Page 258
10.3. City Planning: Provo and Orem......Page 260
10.4. Regional Planning: The WFMR......Page 264
10.5. Coordinating Regional and City Planning......Page 267
10.6. Conclusions......Page 268
References......Page 269
CHAPTER 11 A MULTI-OBJECTIVE EVOLUTIONARY ALGORITHM FOR THE COVERING TOUR PROBLEM......Page 276
11.1. Introduction......Page 277
11.2.1. The Mono-Objective Covering Tour Problem......Page 280
11.2.2. The Bi-Objective Covering Tour Problem......Page 281
11.2.3. Optimization Methods......Page 282
11.3.1. General Framework......Page 284
11.3.2. Solution Coding......Page 285
11.3.3. Genetic Operators......Page 286
11.4. Computational Results......Page 287
11.5. Conclusions and Outlooks......Page 289
References......Page 290
12.1. Introduction......Page 298
12.2. Packet Processor Design......Page 300
12.2.1. Design Space Exploration......Page 301
12.2.2. Basic Models and Methods......Page 303
12.3.1. General Considerations......Page 310
12.3.2. Interface Description......Page 312
12.4.1. Problem Instances......Page 313
12.4.2. Simulation Results......Page 315
12.5. Summary......Page 318
References......Page 321
13.1. Introduction......Page 324
13.2. Adaptive Range Multiobjective Genetic Algorithms......Page 326
13.3.1. Furmulation of Optimization......Page 329
13.3.2. CFD Evaluation......Page 331
13.3.3. Overview of Non-Dominated Solutions......Page 332
13.4.1. Neural Network and SOM......Page 334
13.4.2. Cluster Analysis......Page 336
13.4.3. Visualization of Design Tradeoffs: SOM of Tradeoffs......Page 337
13.4.4. Data Mining of Design Space: SOM of Design Variables......Page 339
13.5. Conclusions......Page 340
Acknowledgments......Page 341
References......Page 342
14.1. Introduction......Page 346
14.2. Physical Problem......Page 348
14.3. Genetic Algorithm......Page 349
14.4. Problem Formulation......Page 351
Index......Page 366
References......Page 367
CHAPTER 15 MULTI-OBJECTIVE SPECTROSCOPIC DATA ANALYSIS OF INERTIAL CONFINEMENT FUSION IMPLOSION CORES: PLASMA GRADIENTƒ......Page 370
15.1. Introduction......Page 371
15.2. Self-Consistent Analysis of Data from X-ray Images and Line Spectra......Page 373
15.3. A Niched Pareto Genetic Algorithm for Multi-Objective Spectroscopic Data Analysis......Page 376
15.4. Test Cases......Page 378
15.5. Application to Direct-Drive Implosions at GEKKO XII......Page 383
15.6. Application to Indirect-Drive Implosions at OMEGA......Page 386
15.7. Conclusions......Page 388
References......Page 390
16.1. Introduction......Page 394
16.2. Medical Image Processing......Page 395
16.2.1. Medical Image Reconstruction......Page 396
16.3. Computer Aided Diagnosis......Page 398
16.3.2. Rules-Based Atrial Disease Diagnosis......Page 399
16.4. Treatment Planning......Page 401
16.4.1. Brachytherapy......Page 402
16.4.2. External Beam Radiotherapy......Page 405
16.4.3. Cancer Chemotherapy......Page 410
16.5. Data Mining......Page 411
16.5.1. Partial Classification......Page 412
16.5.2. Identification of Multiple Gene Subsets......Page 414
16.6. Conclusions......Page 415
References......Page 416
17.1. Introduction......Page 422
17.2.1. Machine Learning......Page 425
17.2.2. Generalization......Page 427
17.2.3. Multiobjective Evolutionary Algorithms (MOEA) & Real-World Applications (RWA)......Page 430
17.3. Problem Formulation......Page 435
17.4. MOEA for Partitioning......Page 439
17.4.1. The Algorithm......Page 440
17.4.4. Constraints & Heuristics......Page 441
17.4.5. Convergence......Page 442
17.5. Results and Discussion......Page 444
17.6. Summary & Future Work......Page 448
References......Page 450
CHAPTER 18 GENERALIZED ANALYSIS OF PROMOTERS: A METHOD FOR DNA SEQUENCE DESCRIPTION......Page 456
18.1. Introduction......Page 457
18.2. Generalized Clustering......Page 458
18.3. Problem: Discovering Promoters in DNA Sequences......Page 461
18.4. Biological Sequence Description Methods......Page 463
18.5. Experimental Algorithm Evaluation......Page 467
18.6. Concluding Remarks......Page 471
References......Page 472
19.1. Introduction......Page 480
19.2. Combinatorial MOP Functions......Page 481
19.3.1. Multi-Objective Quadratic Assignment Problem......Page 482
19.3.2. MOEA mQAP Results and Analysis......Page 488
19.3.3. Modified Multi-Objective Knapsack Problem (MMOKP)......Page 494
19.3.4. MOEA MMOKP Testing and Analysis......Page 500
19.4. MOEA BB Conjectures for NPC Problems......Page 505
References......Page 507
20.1. Introduction......Page 512
20.2. The Multi-Objective Optimization Problem......Page 514
20.3. Multi-Objective Genetic Algorithms......Page 515
20.3.1. The Multi-Objective Struggle GA......Page 516
20.3.3. Similarity Measures......Page 518
20.3.4. Crossover Operators......Page 522
20.4. Fluid Power System Design......Page 523
20.4.1. Optimization Results......Page 525
20.5. Mixed Variable Design Problem......Page 527
20.5.2. Optimization Results......Page 528
20.6. Discussion and Conclusions......Page 529
References......Page 531
CHAPTER 21 ELIMINATION OF EXCEPTIONAL ELEMENTS IN CELLULAR MANUFACTURING SYSTEMS USING MULTI-OBJECTIVE GENETIC ALGORITHMS......Page 534
21.1. Introduction......Page 535
21.2. Multiple Objective Optimization......Page 539
21.3.3. Problem Formulation......Page 540
21.3.4. A Numerical Example......Page 544
21.4. The Proposed MOGA......Page 546
21.4.1. Pseudocode for the Proposed MOGA......Page 547
21.4.2. Fitness Calculation......Page 548
21.4.6. Stopping Criteria......Page 549
21.5. Parameter Setting......Page 550
21.6. Experimentation......Page 551
21.7. Conclusion......Page 554
References......Page 555
22.1. Introduction......Page 558
22.2. Permutation Flowshop Scheduling Problems......Page 560
22.3.1. Implementation of Genetic Algorithms......Page 561
22.3.2. Comparison of Various Genetic Operations......Page 564
22.3.3. Performance Evaluation of Genetic Algorithms......Page 568
22.4. Multi-Objective Genetic Algorithms......Page 570
22.4.1. NSGA-II Algorithm......Page 571
22.4.2. Performance Evaluation of the NSGA-II Algorithm......Page 573
22.4.3. Extensions to Multi-Objective Genetic Algorithms......Page 577
22.5. Conclusions......Page 580
References......Page 581
CHAPTER 23 EVOLUTIONARY OPERATORS BASED ON ELITE SOLUTIONS FOR BI-OBJECTIVE COMBINATORIAL OPTIMIZATION......Page 584
23.1. Introduction......Page 585
23.2. MOCO Problems and Solution Sets......Page 586
23.3.1. Overview of the Heuristic......Page 588
23.3.2. The Initial Population......Page 590
23.3.3. Bound Sets and Admissible Areas......Page 591
23.3.4. The Genetic Map......Page 592
23.3.5. The Crossover Operator......Page 593
23.3.6. The Path-Relinking Operator......Page 594
23.3.7. The Local Search Operator......Page 595
23.4.1. Problem Formulation......Page 596
23.4.2. Experimental Protocol......Page 597
23.5.1. Minimal Complete Solution Sets and Initial Elite Solution Set......Page 598
23.5.2. Our Results Compared with Those Existing in the Literature......Page 600
23.6. Numerical Experiments with the Bi-Objective Knapsack Problem......Page 602
23.6.1. Minimal Complete Solution Sets and the Initial Elite Solution Set......Page 603
23.7. Conclusion and Perspectives......Page 604
References......Page 606
CHAPTER 24 MULTI-OBJECTIVE RECTANGULAR PACKING PROBLEM......Page 610
24.1. Introduction......Page 611
24.2.2. Multi-Objective RP......Page 612
24.3. Genetic Layout Optimization......Page 613
24.3.1. Representations......Page 614
24.3.2. GA Operators......Page 616
24.4.1. Multi-Objective Optimization Problems and Genetic Algorithm......Page 618
24.4.2. Neighborhood Cultivation Genetic Algorithm......Page 620
24.5. Numerical Examples......Page 622
24.5.2. Evaluation Methods......Page 623
24.5.3. Results......Page 624
References......Page 629
25.1. Introduction......Page 632
25.2. Attribute Selection......Page 634
25.3. Multi-Objective Optimization......Page 635
25.4. The Proposed Multi-Objective Methods for Attribute Selection......Page 637
25.4.1. The Multi-Objective Genetic Algorithm (MOGA)......Page 638
25.4.2. The Multi-Objective Forward Sequential Selection Method (MOFSS)......Page 640
25.5. Computational Results......Page 641
25.5.1. Results for the Return All Non-Dominated SolutionsŽ Approach......Page 644
25.5.2. Results for the Return the Best Non-Dominated SolutionŽ Approach......Page 645
25.5.3. On the Effectiveness of the Criterion to Choose the BestŽ Solution......Page 649
25.6. Conclusions and Future Work......Page 652
References......Page 653
26.1. Introduction......Page 656
26.2. A Justification for MOEAs in Financial Applications......Page 657
26.3.1. Portfolio Selection Problems......Page 660
26.3.2. Vederajan et al.......Page 662
26.3.3. Lin et al.......Page 665
26.3.4. Fieldsend & Singh......Page 668
26.3.5. Schlottmann & Seese......Page 671
26.4. Conclusion and Future Research Directions......Page 675
References......Page 678
27.1. Introduction......Page 682
27.2.1. Parameter and Structure Representation of the Network......Page 684
27.2.2. Objectives in Network Optimization......Page 685
27.2.3. Mutation and Learning......Page 687
27.3. Selecting Ensemble Members......Page 688
27.4.2. Results on the Ackley Function......Page 690
27.4.3. Results on the Macky-Glass Function......Page 695
27.5. Discussions and Conclusions......Page 698
References......Page 701
28.1. Introduction......Page 704
28.2. Artificial Neural Networks......Page 706
28.3.2. Weight Decay Regularization and Summed Penalty Terms......Page 710
28.3.4. Problems with These Methods......Page 711
28.4.1. Pareto Optimality......Page 713
28.4.2. Extent, Resolution and Density of Estimated Pareto Set......Page 714
28.4.3. The Use of EMOO......Page 716
28.4.4. A General Model......Page 718
28.4.5. Implementation and Generalization......Page 721
28.5. Empirical Validation......Page 722
28.5.2. Model Parameters......Page 723
28.6. Results......Page 724
Acknowledgements......Page 726
References......Page 727
29.1.1. Introduction......Page 730
29.1.3. Results and Discussion......Page 732
29.2.1. Introduction......Page 737
29.2.2. Methodology......Page 738
29.2.3. Results and Discussion......Page 739
29.2.4. Conclusion......Page 741
29.3.3. Results......Page 744
29.3.4. Conclusion......Page 751
29.4.3. Results and Discussion......Page 752
References......Page 753
30.1. Introduction......Page 756
30.2. Diversity in Multi-Objective Optimization......Page 758
30.3. Maintaining Diversity in Multi-Objective Optimization......Page 759
30.3.1. Weighted Vectors......Page 760
30.3.3. Crowding/Clustering Methods......Page 761
30.3.5. Relaxed Forms of Dominance......Page 762
30.3.7. Objective Oriented Heuristic Selection......Page 764
30.4. The Two-Objective Space Allocation Problem......Page 765
30.4.1. Problem Description......Page 766
30.4.2. Measuring Diversity of Non-Dominated Sets......Page 768
30.5.1. Diversity as a Helper Objective......Page 769
30.5.2. Diversity to Control Exploration and Exploitation......Page 770
30.5.3. The Population-Based Hybrid Annealing Algorithm......Page 771
30.6.1. Experimental Setting......Page 773
30.6.2. Discussion of Obtained Results......Page 774
30.7. Summary......Page 776
References......Page 777
INDEX......Page 782