Advances in Evolutionary and Deterministic Methods for Design, Optimization and Control in Engineering and Sciences (eBook)

Edmondo Minisci, Massimiliano Vasile, Jacques Periaux, Nicolas R. Gauger, Kyriakos C. Giannakoglou, Domenico Quagliarella (Herausgeber)

eBook Download: PDF

2018 | 1st ed. 2019
X, 566 Seiten
Springer International Publishing (Verlag)
978-3-319-89988-6 (ISBN)

Infos

This volume presents up-to-date material on the state of the art in evolutionary and deterministic methods for design, optimization and control with applications to industrial and societal problems from Europe, Asia, and America.

EUROGEN 2015 was the 11th of a series of International Conferences devoted to bringing together specialists from universities, research institutions and industries developing or applying evolutionary and deterministic methods in design optimization, with emphasis on solving industrial and societal problems.

The conference was organised around a number of parallel symposia, regular sessions, and keynote lectures focused on surrogate-based optimization in aerodynamic design,  adjoint methods for steady & unsteady optimization,  multi-disciplinary design optimization, holistic optimization in marine design, game strategies combined with evolutionary computation, optimization under uncertainty, topology optimization, optimal planning, shape optimization, and production scheduling.

Inhalt

Preface 6
Contents 8
Keynote 12
Risk, Optimization and Meanfield Type Control 13
1 Introduction 13
2 A General Framework for Meanfield Type Control 14
3 Portfolio Optimization 16
3.1 Polynomial Solution 17
4 Find the Best Production Strategy for an Exhaustible Resource 18
4.1 Notations 19
4.2 Dynamic Programming Solved by a Fixed Point Algorithm 19
4.3 Numerical Implementation 21
5 Numerical Solution of a Systemic Risk Problem 23
References 25
Surrogate-Based Optimization in Aerodynamic Design 27
A Review of Surrogate Modeling Techniques for Aerodynamic Analysis and Optimization: Current Limitations and Future Challenges in Industry 28
1 Introduction 28
2 Design of Experiments (DoE) 29
2.1 Classical DoE 30
2.2 Modern DoE 31
3 Surrogate Models 32
4 Surrogate-Based Optimization (SBO) 34
5 Aircraft Aerodynamic Industrial Applications: Use, Limitations and Future Challenges 35
6 Conclusion 39
References 39
Constrained Single-Point Aerodynamic Shape Optimization of the DPW-W1 Wing Through Evolutionary Programming and Support Vector Machines 43
1 Introduction and Previous Works 44
1.1 Introduction 44
1.2 Recent Research Efforts in SBO for Aerodynamic Shape Design 44
1.3 Garteur AD/AG52 45
2 Definition of the Optimization Problem 45
2.1 Baseline Geometry: DPW-W1 Wing 45
2.2 Parameterization 45
2.3 Aerodynamic Constraints 46
2.4 Geometric Constraints 46
2.5 Design Point and Objective Function 47
2.6 Computational Grids 47
3 Description of the Applied Approach 48
3.1 Adaptive Sampling Focused on Optimization 48
3.2 Support Vector Regression Algorithm as Surrogate Model 48
3.3 Evolutionary Programming 49
3.4 Handling Constrains Within the Optimization Process 50
4 Numerical Results 51
4.1 Inviscid Transonic Flow 52
4.2 Viscous Transonic Flow 52
5 Conclusions 56
References 56
Enabling of Large Scale Aerodynamic Shape Optimization Through POD-Based Reduced-Order Modeling and Free Form Deformation 57
1 Introduction 57
2 Geometry Parametrization 59
2.1 Continuity on Arbitrary Shaped Boundaries 59
2.2 Direct Surface Manipulation 61
2.3 CAMILO 63
3 POD Based Reduced-Order Modeling 64
3.1 Proper Orthogonal Decomposition 64
3.2 Hybrid ROM Based on Domain Decomposition 65
3.3 Automatic Detection of the HFM/ROM Interface 65
4 Sail Optimization 67
5 Conclusions 70
References 70
Application of Surrogate-Based Optimization Techniques to Aerodynamic Design Cases 72
1 Introduction 72
2 Literature Review 73
3 Surrogate Model 75
3.1 Pseudo-continuous Global Representation 77
4 Surrogate Model Sequential In-fill 78
4.1 Factorization Criterion 80
4.2 Expected Improvement-Like Criterion 81
5 Surrogate-Based Optimization 83
6 Application to Aerodynamic Design Cases 84
6.1 RAE 2822 Airfoil Case 84
6.2 Drag Prediction Workshop Wing Case 89
7 Conclusions 98
References 99
Efficient Global Optimization Method for Multipoint Airfoil Design 101
1 Introduction 101
2 Methods 103
3 Metamodel Building: Gaussian Process Approach 104
4 Rating a Metamodel: Cross Validation and Metrics 105
5 Infill Strategy: Expected Improvement 106
5.1 Weighted EI—A Global-Local Approach 107
5.2 Adaptive Approach 108
6 Multipoint Airfoil Shape Optimization 108
6.1 Problem Description 109
7 Results 111
7.1 Validation Results 111
7.2 Optimization Results 114
8 Conclusions 119
References 119
Adjoint Methods for Steady and Unsteady Optimization 121
Checkpointing with Time Gaps for Unsteady Adjoint CFD 122
1 Introduction 122
2 Background 123
2.1 Solving the Flow and Adjoint Equations 123
2.2 Physical Checkpointing 124
3 Checkpointing with Gaps 125
4 Test Case 126
4.1 Primal Solver Setup 126
4.2 Incomplete Checkpointing Setup 127
4.3 Adjoint Solver Setup 127
5 Results 128
5.1 Overall Accuracy: Angle of Attack 129
5.2 Spatial Accuracy: Surface Sensitivity 130
5.3 Temporal Accuracy: Flow Control 131
6 Conclusion, Possible Extensions 134
References 134
Shape Optimization of Wind Turbine Blades Using the Continuous Adjoint Method and Volumetric NURBS on a GPU Cluster 136
1 Introduction 136
2 Navier-Stokes, Adjoint Equations and Sensitivity Derivatives 137
2.1 Flow (Primal) Equations 138
2.2 Continuous Adjoint Formulation 138
2.3 Discretization and Numerical Solution 140
3 Parameterization Through Volumetric NURBS 141
4 Implementation on GPUs 142
5 Optimization of the Wind Turbine Blade 145
6 Conclusions 148
References 148
Aerodynamic Shape Optimization Using the Adjoint-Based Truncated Newton Method 150
1 Introduction to the Truncated Newton Method 150
2 The Continuous Adjoint Method for the Computation of ?F?bi 151
3 Computation of Hessian(F)–Vector Products 153
4 Computation of overlinevi and overlinep 154
5 Computation of overlineui and overlineq 155
6 Computation of overlinexk and overlineoverlinexk,n 156
7 The TN Algorithm—Comments on the CPU Cost 156
8 Applications 157
9 Conclusions 160
References 161
Application of the Adjoint Method for the Reconstruction of the Boundary Condition in Unsteady Shallow Water Flow Simulation 162
1 Introduction 162
2 Mathematical Model 163
2.1 Adjoint System 164
2.2 Numerical Method 165
3 Application 166
3.1 Case Description 166
3.2 Numerical Optimization 168
4 Results and Discussion 170
5 Conclusions 176
References 177
Aerodynamic Optimization of Car Shapes Using the Continuous Adjoint Method and an RBF Morpher 178
1 Introduction 178
2 The Continuous Adjoint Method 180
2.1 Flow Equations 180
2.2 General Objective Function 180
2.3 Continuous Adjoint Formulation 181
3 RBF-based Morphing 183
3.1 RBFs Background 184
3.2 RBF Morph Tool 186
3.3 Coupling of RBF Mesh Morphing with Adjoint Sensitivities 186
4 Optimization Algorithm 187
5 Applications 188
6 Conclusions 191
References 191
Holistic Optimization in Marine Design 193
Upfront CAD—Parametric Modeling Techniques for Shape Optimization 194
1 Introduction 195
2 Overview 196
3 Partially-Parametric Modeling (PPM) 197
3.1 Free-Form Deformation 197
3.2 Shift Transformations 199
3.3 Added Patch Perturbation 200
3.4 Morphing 201
3.5 Radial Basis Functions 202
4 Fully-Parametric Modeling (FPM) 203
4.1 MetaSurface Approach for Fully-Parametric Modeling 204
4.2 Parametric Modeling of Building Patterns 205
4.3 Incorporated Handling of Constraints 208
5 Design Velocities 209
6 Simulation-Driven Design 210
7 Requirements and Comparison 212
8 Conclusions 212
References 213
Simulation-Based Design Optimization by Sequential Multi-criterion Adaptive Sampling and Dynamic Radial Basis Functions 215
1 Introduction 215
2 Optimization Problem Formulation 217
3 Dynamic Radial Basis Function Method for Optimization 217
3.1 Surrogate Model 217
3.2 Multi-criterion Adaptive Sampling 218
3.3 Optimization Procedure 219
4 Deterministic Particle Swarm Optimization 220
5 Optimization Problems 220
5.1 Unconstrained Global Optimization Test Problems 220
5.2 Hull-Form Optimization of the DTMB 5415 221
6 Numerical Results 223
6.1 Unconstrained Global Optimization Test Problems 223
6.2 Hull-Form Optimization of the DTMB 5415 226
7 Conclusions and Future Work 227
References 228
Application of Holistic Ship Optimization in Bulkcarrier Design and Operation 231
1 Introduction 231
2 Overview of the Holistic Methodology 233
2.1 Design and Simulation Environment 233
2.2 Geometric Core 235
2.3 Initial Hydrostatic Properties 235
2.4 Lackenby Variation 235
2.5 Cargo Hold Modelling 235
2.6 Resistance Prediction 237
2.7 Deadweight Analysis 239
2.8 Stability and Loadline Check 239
2.9 Operational Profile Simulation 239
2.10 Energy Efficiency Design Index Calculation 242
3 Design Concept 242
3.1 Large Bulkcarrier Market 242
3.2 Baseline Vessel—208k Newcastlemax 243
3.3 Proposed Design Concept Characteristics 243
4 Optimization Studies 244
4.1 Optimization Target/Goals 244
4.2 Design Variables 245
4.3 Optimization Procedure 245
4.4 Design of Experiment 247
4.5 Global Optimization Studies 248
4.6 Dominant Variant Ranking 250
5 Discussion of the Results—Future Perspectives 250
References 254
Game Strategies Combined with Evolutionary Computation 255
Designing Networks in Cooperation with ACO 256
1 Introduction 256
2 A Network Design Model 258
3 The Modified Network Design Model 261
3.1 Computational Procedure 263
4 Airline Network Design 264
5 Concluding Remarks 267
References 267
Augmented Lagrangian Approach for Constrained Potential Nash Games 269
1 Introduction 269
2 Iterative Scheme 270
3 Augmented Lagrangian 272
4 Optimization Procedure Implementation 274
5 A Test Problem with Duality Gap 275
6 Facility Location Problem 276
6.1 Test Case 1 278
6.2 Test Case 2 280
7 Conclusions 281
References 281
A Diversity Dynamic Territory Nash Strategy in Evolutionary Algorithms: Enhancing Performances in Reconstruction Problems in Structural Engineering 283
1 Introduction 283
2 Nash–Evolutionary Algorithms 284
2.1 Static Nash Territory DD 285
2.2 Dynamic Nash Territory DD 285
2.3 Diversity Dynamic Nash Territory DD 285
3 The Structural Problem 285
4 Test Case 286
5 Results and Discussion 288
5.1 Comparing Static Nash Domain Decomposition 291
5.2 Comparing Dynamic Nash Domain Decomposition 292
5.3 Comparing Diversity Dynamic Nash Domain Decomposition 293
5.4 Comparing Dynamic Nash Domain Decomposition with Diversity Dynamic Nash Domain Decomposition 294
5.5 Final Comparisons 296
5.6 Overall Discussion 297
6 Conclusions 300
References 300
Interactive Inverse Modeling Based Multiobjective Evolutionary Algorithm 302
1 Introduction 302
2 Interactive IM-MOEA Algorithm 303
2.1 IM-MOEA 303
2.2 Incorporation of Decision Maker Preferences Through Adapting Reference Vectors 304
2.3 Decision Making Module 305
3 Numerical Experiments 307
3.1 Example with a Biobjective Problem 308
3.2 Example with a Three Objective Problem 311
4 Conclusions 313
References 314
Multi-disciplinary Design Optimization of Air-Breathing Hypersonic Vehicle Using Pareto Games and Evolutionary Algorithms 315
1 Introduction 316
2 Integrated Forebody Surface/SCRAMJET Inlet Analysis 317
3 Divergent Area Supersonic Combustor Analysis 318
4 Multi Disciplinary Optimization of Airbreathing Vehicle 320
4.1 Introduction 321
4.2 Definition of the Test Case 321
4.3 Software Requirement Requirements 321
4.4 Computational Domain 321
4.5 Baseline Configuration with Prescribed Boundary Conditions on the Fore Body and After Body Geometries 322
4.6 Optimization and Objective Functions 323
4.7 Cooperative Games: Pareto Optimality 323
4.8 Design Point 325
4.9 Design Parameters: Search Space for Shape Optimization 325
4.10 Outputs Results 325
4.11 Some Preliminary Results Obtained with NSGA II Software for the Shape Optimisation Scramjet Problem 326
5 Conclusion 327
References 328
Optimisation Under Uncertainty 330
Innovative Methodologies for Robust Design Optimization with Large Number of Uncertainties Using ModeFRONTIER 331
1 Introduction 331
2 UQ of Large Number of Variables: SS-ANOVA and Stepwise Regression for Sparse Collocation 332
3 UQ Test Case Application 335
4 RDO: Classical Versus MINMAX Approach 339
5 Reliability-Based RDO 344
6 Conclusion 346
References 346
A Novel Method for Inverse Uncertainty Propagation 348
1 Introduction 348
2 Problem Definition 349
3 Enablers 352
3.1 Univariate Reduced Quadrature (URQ) Method 352
3.2 Workflow Reversal 353
4 Method for Inverse Propagation 353
5 Validation 355
5.1 Test Case 1: Linear Functions 355
5.2 Test Case 2: Non-Linear Functions 356
6 Industrial Testcase 357
6.1 Testcase Setup 357
6.2 Forward Propagation 359
6.3 Inverse Propagation 359
7 Conclusion and Future Work 363
References 363
Uncertainty Sources in the Baseline Configuration for Robust Design of a Supersonic Natural Laminar Flow Wing-Body 366
1 Introduction 366
2 Design Problem Description 367
2.1 Geometry and Design Problem Definition 367
2.2 Optimization Problem 367
3 Computational Model 370
4 Definition of Uncertainties 370
4.1 Geometrical Uncertainties 371
4.2 Operational Uncertainties 372
4.3 Model Uncertainties (epistemic) 372
5 Preliminary Parametric Analysis 373
5.1 Effect of Uncertainty of Ncritical Factor 373
5.2 CL and Mach Contrast Effects on Ncritical 374
5.3 NURBS Parameterization to Model Uncertainties in Leading Edge Shape 375
6 Sensitivity Analysis 377
6.1 Variance-Based Decomposition and Sobol Indices 379
6.2 Design Space Sampling 380
6.3 Failure Handling 380
6.4 Sensitivity Analysis Results 381
7 Conclusions and Future Prospects 384
References 384
Robust Airfoil Design in the Context of Multi-objective Optimization 386
1 Introduction 386
2 Robust Multi-objective Optimization 388
2.1 Robust Pareto-Optimal Solutions 388
2.2 Uncertainty Quantification 389
2.3 Multi-objective Optimization Method 391
3 Application and Results 392
3.1 Uncertainties in the Flight Conditions 392
3.2 Geometrical Uncertainties 395
4 Summary and Outlook 397
References 398
An Alternative Formulation for Design Under Uncertainty 399
1 Motivation and Objectives 399
2 Optimization Problem 400
3 Computation of the Minimum 401
4 Results 403
4.1 Computation of the Minimum 404
4.2 Optimization with Minimum 406
5 Conclusions 410
References 411
Polynomial Representation of Model Uncertainty in Dynamical Systems 412
1 Introduction 412
2 Polynomial Expansion of Unmodelled Components 413
2.1 Problem Statement 414
2.2 Treatment of Stochastic Observations 415
2.3 Uncertainty Distance 416
2.4 Solution Through Optimisation 417
3 Examples 417
3.1 Linear Elastic Dynamics with Friction 417
3.2 Orbital Motion with Unmodelled Drag 418
3.3 Chaotic and Hypersensitive Systems 422
4 Reachability Under Model Uncertainty 423
5 Final Remarks 425
References 425
Algorithms and Industrial Applications 426
Improved Archiving and Search Strategies for Multi Agent Collaborative Search 427
1 Introduction 427
2 Problem Formulation 429
2.1 Tchebycheff Scalarisation 429
3 Implementation 430
3.1 Individualistic Actions 431
3.2 Social Actions 432
3.3 The New Archiving Strategy 433
4 Test Cases 438
4.1 CEC 2009 UF Functions 438
4.2 ZDT4 and 3 Impulse Problem 444
5 Conclusions 446
References 446
Comparison of Multi-objective Approaches to the Real-World Production Scheduling 448
1 Introduction 448
2 Related Work 449
3 Implemented Multi-objective Algorithms 450
3.1 Non-dominated Sorting Genetic Algorithm-II 450
3.2 Strength Pareto Evolutionary Algorithm 2 451
3.3 Indicator-Based Evolutionary Algorithm 451
4 Production Scheduling Problem 451
4.1 Production Schedule Encoding 452
4.2 Population Initialization 452
4.3 Reproduction Operators 452
4.4 Fitness Evaluation 453
4.5 Ending Condition 453
5 Memetic Algorithms 453
6 Performance Evaluation 454
6.1 Experimental Environment 454
6.2 Test Cases 454
6.3 Control Parameter Settings 455
6.4 Results 455
7 Conclusion 461
References 462
Elucidation of Influence of Fuels on Hybrid Rocket Using Visualization of Design-Space Structure 463
1 Introduction 464
2 Design Informatics 465
2.1 Optimization Method 465
2.2 Data-Mining Technique 466
3 Problem Definition 467
3.1 Objective Functions 468
3.2 Design Variables 468
3.3 Evaluation Method 468
4 Optimization Results 470
5 Data-Mining Results 472
6 Conclusions 477
References 477
Creating Optimised Employee Travel Plans 479
1 Introduction and Motivation 479
2 Previous Work 480
3 Problem Instances 481
3.1 The Edinburgh Datasets 482
3.2 The London Datasets 482
4 Methodology 483
4.1 The Evolutionary Algorithm 483
4.2 Building Travel Plans 484
4.3 Experiments 484
5 Results 485
6 Conclusions 488
7 Future Work 490
References 492
A New Rich Vehicle Routing Problem Model and Benchmark Resource 493
1 Introduction 493
2 Background 494
2.1 Benchmark Data for VRP 497
3 Model 497
3.1 Scenario Characteristics 498
3.2 Scenario Characteristics 498
4 Problem Generation 502
4.1 Vehicle Fleet 502
4.2 Customer Databases 502
4.3 Generating Problem Instances 503
5 Benchmark Datasets 504
5.1 XML Description 504
5.2 Parsing Problems Using the Supplied Java Parser 506
5.3 Other Formats or Programming Languages 507
6 Conclusion 507
References 507
Genetic Algorithm Applied to Design Knowledge Discovery of Launch Vehicle Using Clustered Hybrid Rocket 509
1 Introduction 510
2 Design Methods 511
2.1 Evaluation of Clustered Hybrid Rocket 511
3 Design Optimization 514
3.1 Non-dominated Sorting Genetic Algorithm-II (NSGA-II) 514
3.2 Self-Organizing Map (SOM) 515
4 Formulation 517
5 Results 521
5.1 Design Exploration Results 521
5.2 Comparison of Designs from Non-dominated Solutions 521
5.3 Visualization of Design Space by SOM 523
6 Conclusions 524
References 524
Topology Optimization of Flow Channels with Heat Transfer Using a Genetic Algorithm Assisted by the Kriging Model 526
1 Introduction 526
2 Computational Methods 527
2.1 Flow Channel Representation 527
2.2 Building-Cube Method 529
2.3 Genetic Algorithm 529
2.4 Kriging Model 530
3 Optimization Problems of Minimizing Pressure Loss 530
3.1 Nozzle Example (Case1) 530
3.2 Double Pipe Example (Case 2) 532
4 Optimization Problems of Maximizing Heat Transfer (Case 3 and 4) 534
4.1 Problem Definition 534
4.2 Results 535
5 Multi-objective Optimization Problems 537
5.1 Problem Definition 537
5.2 Results 538
6 Conclusion 540
References 540
Topology Optimization Using GPGPU 542
1 Introduction 542
2 Previous Work 543
3 Finite Element Formulation 544
3.1 Linear Elasticity Equations 544
3.2 Per Node Equations 545
4 Multigrid Solver 546
4.1 Per Node Equations Assembly for All Levels 546
4.2 Gauss-Seidel Relaxation 546
5 Topology Optimization Formulation 547
6 Imposing Dirichlet Boundary Conditions 548
6.1 Nitsche Terms 549
7 Results 550
7.1 GE Challenge 550
7.2 Bridge Design 550
7.3 Cantilever 553
7.4 Dirichlet Boundary Conditions 554
References 554

Produkt-Details

• Erscheint lt. Verlag: 2.7.2018
• Reihe/Serie: Computational Methods in Applied Sciences
• Reihe/Serie: Computational Methods in Applied Sciences
• Zusatzinfo: X, 566 p. 272 illus.
• Verlagsort: Cham
• Sprache: englisch
• Themenwelt: Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik
• Themenwelt: Mathematik / Informatik ► Mathematik
• Themenwelt: Technik ► Maschinenbau
• Schlagworte: Computational Intelligence • computational methods • Engineering design • Optimization for Societal Problems • Optimization Methods
• ISBN-10: 3-319-89988-0 / 3319899880
• ISBN-13: 978-3-319-89988-6 / 9783319899886