This book investigates two types of static multi-fidelity surrogates modeling approaches, sequential multi-fidelity surrogates modeling approaches, the multi-fidelity surrogates-assisted efficient global optimization, reliability analysis, robust design optimization, and evolutionary optimization. Multi-fidelity surrogates have attracted a significant amount of attention in simulation-based design and optimization in recent years. Some real-life engineering design problems, such as prediction of angular distortion in the laser welding, optimization design of micro-aerial vehicle fuselage, and optimization design of metamaterial vibration isolator, are also provided to illustrate the ability and merits of multi-fidelity surrogates in support of engineering design. Specifically, lots of illustrative examples are adopted throughout the book to help explain the approaches in a more “hands-on” manner. This book is a useful reference for postgraduates and researchers of mechanical engineering, as well as engineers of enterprises in related fields.
Author(s): Qi Zhou, Min Zhao, Jiexiang Hu, Mengying Ma
Series: Engineering Applications of Computational Methods, 12
Publisher: Springer
Year: 2022
Language: English
Pages: 460
City: Singapore
Contents
1 Introduction
1.1 Concept of Multi-fidelity Surrogates
1.2 Review of Multi-fidelity Surrogates in Engineering Design
1.2.1 Classification of Multi-fidelity Surrogate Modeling Methods
1.2.2 Research Progress on Multi-fidelity Surrogate Modeling Methods
1.3 Sampling Method for Multi-fidelity Surrogates
1.3.1 One-Shot Sampling Method
1.3.2 Sequential Sampling Method
1.4 Optimization Design Based on MF Surrogates
1.4.1 MF Surrogate Management
1.4.2 Heuristic Optimization Algorithms Based on MF Surrogates
1.4.3 Uncertainty Optimization Design Based on MF Surrogates
1.5 Application of MF Surrogates in the Optimization Design of Complex Equipment
References
2 Hierarchical Multi-fidelity Surrogate Modeling
2.1 Difference Mapping Based on Ensembles of Surrogates for Multi-fidelity Surrogate Modeling
2.1.1 Tuning the LF Surrogate
2.1.2 Difference Mapping Based on Ensembles of Surrogates
2.1.3 Process Flow of the Proposed DMF-EM
2.1.4 Examples and Results
2.2 Bumpiness of the Scaling-Function Reduction Method for Multi-fidelity Surrogate Modeling
2.2.1 Introduction of the BR-SF Model
2.2.2 Numerical Tests
2.3 Space-Mapping Method for Multi-fidelity Surrogate Modeling
2.3.1 Introduction of the SM-MF
2.3.2 Introduction of the SM-RBF
2.3.3 Numerical Tests
2.4 Generalized Hierarchical Cokriging Model for Multi-fidelity Surrogate Modeling
2.4.1 Introduction of the GCK Model
2.4.2 Numerical Tests
References
3 Nonhierarchical Multi-fidelity Surrogate Modeling
3.1 Variance-Weighted-Sum Method for Multi-fidelity Surrogate Modeling
3.1.1 Introduction of the VWS-NMF Model
3.1.2 Numerical Tests of VWS-MFS
3.2 Derivative of the Scaling Function Reduction Method for Multi-fidelity Surrogate Modeling
3.2.1 Introduction of the DR-SF Surrogate
3.2.2 Numerical Tests of NHLF-Cokriging Surrogate
3.3 Multi-Output Gaussian Process Model for Multi-fidelity Surrogate Modeling
3.3.1 Introduction of the NH-MOMF Surrogate
3.3.2 Numerical Tests of the NH-MOMF Surrogate
References
4 Sequential Multi-fidelity Surrogate Modeling
4.1 Difference Mapping Method for Multi-fidelity Surrogate Modeling
4.1.1 Introduction of the DM-EM Model
4.1.2 Numerical Tests
4.2 Weighted Cumulative-Error-Based Sequential Multi-fidelity Surrogate Modeling
4.2.1 Introduction of the AL-SMF Method
4.2.2 Numerical Tests of the AL-VFM Method
4.3 Predicted-Improvement-Level-Based Sequential Multi-fidelity Surrogate Modeling
4.3.1 Introduction of the SMF Method
4.3.2 Numerical Tests of the SMF Method
4.4 Bootstrap-Estimator-Based Sequential Multi-fidelity Surrogate Modeling
4.4.1 Introduction of the BB-SMF Method
4.4.2 Numerical Tests
References
5 Multi-fidelity Surrogate Assisted Efficient Global Optimization
5.1 Lower Confidence Bounding Method for Efficient Multi-fidelity Global Optimization
5.1.1 Introduction of the LCB-MFO Method
5.1.2 Numerical Tests
5.2 Probability of Improvement Method for Efficient Multi-fidelity Global Optimization
5.2.1 Introduction of the MF-PI Method
5.2.2 Numerical Tests
5.3 Space Preselection Method for Efficient Multi-fidelity Global Optimization
5.3.1 Introduction of the SP-MFO Method
5.3.2 Numerical Tests
References
6 Multi-fidelity Surrogate Assisted Reliability Design Optimization
6.1 Augmented Expected Feasibility Function-Based Method for Multi-fidelity Surrogate Assisted Reliability Design Optimization
6.1.1 Introduction of the EGRA-MF Method
6.1.2 Numerical Tests of EGRA-MF
6.2 Contour Prediction Method for Multi-fidelity Surrogate Assisted Reliability Design Optimization
6.2.1 Introduction of the EEI Method
6.2.2 Numerical Tests
References
7 Multi-fidelity Surrogate Assisted Robust Design Optimization
7.1 Multi-fidelity Surrogate Assisted Six-Sigma Robust Optimization
7.1.1 Introduction of the MF-RO Method
7.1.2 Numerical Tests of the MF-RO Method
7.2 Multi-fidelity Surrogate Assisted Sequential Robust Optimization
7.2.1 Introduction of the MF-SRO Method
7.2.2 Numerical Tests of the MF-SRO Method
7.3 Conservative Multi-fidelity Surrogate Assisted Robust Optimization
7.3.1 Introduction of the CMF-RO Method
7.3.2 Numerical Tests of the CMF-RO Method
References
8 Multi-fidelity Surrogate Assisted Evolutional Optimization
8.1 Multi-fidelity Surrogate Assisted Multi-objective Genetic Algorithm
8.1.1 Introduction of the AMFS-MOGA Method
8.1.2 Numerical Tests of AMFS-MOGA
8.2 Multilevel Multi-fidelity Surrogate Assisted Multi-objective Genetic Algorithm
8.2.1 Introduction of the TSMA-MOGA Method
8.2.2 Numerical Tests of TSMA-MOGA
8.3 Online Multi-fidelity Surrogate Assisted Multi-objective Genetic Algorithm
8.3.1 Introduction of the OLMF-MOGA Method
8.3.2 Numerical Tests of OLMF-MOGA
References
9 Engineering Applications
9.1 Prediction of Angular Distortion in Laser Welding
9.1.1 Laser Welding Experiment
9.1.2 Finite Element Simulation
9.1.3 Results and Discussion
9.2 Optimization of Laser Beam Welding Parameters
9.2.1 Two Models with Different Fidelity Levels
9.2.2 Proposed Approach
9.2.3 Results and Discussion
9.3 Optimization of Metamaterial Vibration Isolator Design
9.3.1 Establishment of the FEM of the Honeycomb Structure Vibration Isolator
9.3.2 Experimentation and Validation of the FE Model
9.3.3 Optimization Design of the MI300-Type Honeycomb Structure Vibration Isolator
9.4 Optimization Design of a Stiffened Cylindrical Shell with Variable Ribs
References
10 Concluding Remarks
10.1 Conclusions
10.2 Remaining Challenges
