New Metaheuristic Schemes: Mechanisms and Applications

Preface
Contents
1 Introduction to Metaheuristic Schemes: Characteristics, Properties, and Importance in Solving Optimization Problems
	1.1 Introduction
	1.2 Classical Methods for Optimization
	1.3 Metaheuristic Algorithms
		1.3.1 Typical Scheme of a Metaheuristic Algorithm
	References
2 Exploring the Potential of Agent Systems for Metaheuristics
	2.1 Introduction
	2.2 Models Through Agent-Based Elements
	2.3 Example of Agent-Based Model “Heroes and Cowards”
	2.4 Metaheuristic Methods as Agent-Based Models
		2.4.1 Problem Setting
		2.4.2 A New Metaheuristic Method Based on the Heroes and Cowards Model
		2.4.3 Computational Structure
	2.5 Experiments
		2.5.1 Evaluation of the Performance Considering Its Own Parameters
		2.5.2 Comparison with Other Similar Methods
		2.5.3 Study of Convergence
		2.5.4 Applications
		2.5.5 The Design Problem of Three-Bar Truss
		2.5.6 Tension/Compression of a Spring as Design Problem
	2.6 Conclusions
	Appendix 1 Functions Used in the Experiments
	Appendix 2 Engineering Design Problems
	References
3 Dynamic Multimodal Function Optimization: An Evolutionary-Mean Shift Approach
	3.1 Introduction
	3.2 The Method of Mean-Shift
		3.2.1 Preliminary Concepts
		3.2.2 Attractors
	3.3 Optimization Approach
		3.3.1 Adaptation of the Mean Shift
		3.3.2 Process of Memory
		3.3.3 Dynamical Procedure
		3.3.4 The Dynamical Optimization Method (EMDO)
	3.4 Experiments and Results
		3.4.1 Benchmark Generator (GDBG)
		3.4.2 Performance Metrics
		3.4.3 Configuration
		3.4.4 Outcomes
		3.4.5 Computational Complexity
	3.5 Conclusions
	References
4 Trajectory-Driven Metaheuristic Approach Using a Second-Order Model
	4.1 Introduction
	4.2 Description of Second-Order Models
		4.2.1 Underdamped Behavior ( 0<ζ<1)
		4.2.2 Critically Damped Behavior (ζ=1)
		4.2.3 Overdamped Behavior ( ζ>1)
	4.3 Movement Patterns as Search Strategies
	4.4 Movement Pattern Through Second-Order Models
	4.5 Analysis of Exploration and Exploitation
	4.6 Movement Pattern
		4.6.1 Initialization Stage
		4.6.2 Generation of the Trajectory
		4.6.3 Reset of Agents with Bad Quality
		4.6.4 Mechanism for Avoiding Premature Convergence
	4.7 Results
		4.7.1 Multimodal Experiments
		4.7.2 Unimodal Experiments
		4.7.3 Hybrid Experiments
		4.7.4 Study of Convergence
	4.8 Conclusions
	Appendix 1 List of Benchmark Functions
	References
5 Collaborative Hybrid Grey Wolf Optimizer: Uniting Synchrony and Asynchrony
	5.1 Introduction
	5.2 The Original Grey Wolf Optimizer Method
		5.2.1 Computational Algorithm
	5.3 Improved Versions of the GWO
		5.3.1 The Modified Grey Wolf Optimizer (mGWO)
		5.3.2 The Proportional-Based Grey Wolf Optimizer (PGWO)
		5.3.3 The Tournament-Based Grey Wolf Optimizer (TGWO)
		5.3.4 The Weighted Distance Grey Wolf Optimizer (wdGWO)
		5.3.5 Complex-Valued Encoding Grey Wolf Optimizer (CGWO)
		5.3.6 Evolutionary Population Dynamics Grey Wolf Optimizer (EPD-GWO)
	5.4 The Synchronous-Asynchronous GWO
		5.4.1 The Synchronous-Asynchronous Mechanism
		5.4.2 The Impact and Effect of the Parameter a
		5.4.3 Control Procedure
		5.4.4 Initialization
		5.4.5 Processing Stage
		5.4.6 Diversity Change
		5.4.7 Computational Procedure
		5.4.8 Computational Complexity
	5.5 Experiments
		5.5.1 Functions from CEC2017 with 30 Dimensions
		5.5.2 Functions from CEC2017 with 50 Dimensions
		5.5.3 Functions from CEC2017 with 100 Dimensions
		5.5.4 Analysis of Convergence
		5.5.5 Analysis of Diversity
		5.5.6 Computational Cost
		5.5.7 Application in Engineering Problems
	5.6 Conclusions
	Appendix 1
		Pressure Vessel Design Problem
		Gear Train Design Problem
		Tension/Compression Spring Design Problem
		The Three-Bar Truss Design Problem
		The Welded Beam Design Problem
		Parameter Estimation for FM Synthesizers
	References
6 Efficient Image Contrast Enhancement by Using the Moth Swarm Algorithm
	6.1 Introduction
	6.2 Mean-Shift Scheme
		6.2.1 Estimation of the Probability Density Function
		6.2.2 Attractors
	6.3 Image Contrast Enhancement Considering as an Optimization Problem
	6.4 Moth Swarm Algorithm (MSA)
	6.5 Image Contrast Enhancement Through MSA
		6.5.1 Elimination of Irrelevant Data
		6.5.2 Objective Function
		6.5.3 Penalty Function
	6.6 Experiments
		6.6.1 Comparison Over Grayscale Images
		6.6.2 Performance Comparison Over Distinct Objective Functions
		6.6.3 Comparison Over Grayscale Images Considering Other Metaheuristic Techniques
		6.6.4 Performance in Color Images
	6.7 Conclusions
	References
7 Multi-objective Optimization of Anisotropic Diffusion Parameters for Enhanced Image Denoising
	7.1 Introduction
	7.2 Main Concepts of Anisotropic Diffusion
	7.3 NSGA-III and Multi-objective Optimization
		7.3.1 Multi-objective Optimization
		7.3.2 The NSGA-III
	7.4 Function Cost
		7.4.1 Estimation of Noise
		7.4.2 Improvement of Contrast
	7.5 Multi-objective Methodology
		7.5.1 Process of NSGA-III
		7.5.2 Analysis of the Optimal Pareto Front
	7.6 Experiments
		7.6.1 Configuration of the Tests
		7.6.2 Indexes of Evaluation
		7.6.3 Results of the Comparison
		7.6.4 Discussion of the Results
	7.7 Conclusions
	References