基于原子和人工生态系统的新智能优化算法及应用

《基于原子和人工生态系统的新智能优化算法及应用（英文版）》基于原子动力学、人工生态系统和经济学提出了三种元启发优化算法，并把这三种新颖的智能优化算法应用于实际工程中。《基于原子和人工生态系统的新智能优化算法及应用（英文版）》共分为7章：第1章介绍了智能优化算法的发展及其优缺点；第2章介绍了原子搜索优化算法的过程和基本原理，并对算法进行了测试；第3章将原子搜索优化算法应用于解决工程问题；第4章介绍了人工生态系统优化算法的提出过程和基本原理，并对算法进行了测试；第5章将基于人工生态系统的优化算法应用于工程实际中；第6章介绍了基于经济学的智能优化算法的启发和基本原理；第7章给出了基于经济学的智能优化算法在不同工程实例中的应用。

Contents
Preface vii
Acknowledgments ix
1. Introduction 1
1.1 Optimization algorithms 1
1.2 A short outline of optimization algorithms 2
1.3 Organization of this book 6
References 7
2. Atom search optimization algorithm 13
2.1 Introduction 13
2.2 Basic molecular dynamics 15
2.3 Atom search optimization 18
2.4 Experimental results 26
2.5 Conclusions 43
References 44
3. Engineering applications of atom search optimization algorithm 47
3.1 Introduction 47
3.2 Parameter estimation for chaotic system 47
3.3 Circular antenna array design problem 51
3.4 Spread spectrum radar polyphase code design 52
3.5 Conclusions 56
References 57
4. Artificial ecosystem-based optimization algorithm 59
4.1 Introduction 59
4.2 Artificial ecosystem-based optimization 61
4.3 Results and discussion 68
4.4 Conclusions 88
References 88
5. Engineering applications of artificial ecosystem-based optimization 93
5.1 Engineering optimization using the AEO algorithm 93
5.2 Static economic load dispatch problem 103
5.3 Hydrothermal scheduling problem 111
5.4 Conclusions 119
References 120
6. Supply-demand-based optimization 123
6.1 Introduction 123
6.2 Supply-demand-based optimization 124
6.3 Experimental results and discussion 134
6.4 Conclusions 141
References 141
7. Engineering applications of supply-demand-based optimization 143
7.1 Introduction 143
7.2 Three-bar truss design 143
7.3 Cantilever beam design 146
7.4 Rolling element bearing design 147
7.5 Gear train design 150
7.6 Conclusions 151
References 152
Appendix 153
Index 163

CHAPTER 1 Introduction
　　Contents
　　1.1 Optimization algorithms 1
　　1.2 A short outline of optimization algorithms 2
　　1.3 Organization of this book 6
　　References 7
　　1.1 Optimization algorithms
　　Optimization algorithms are increasingly popular in intelligent computing and are widely applied to a large number of real-world engineering prob-lems. Their popularity derives from the following aspects. First， all of these optimization techniques have some fundamental theories and mathematical models that have been proved to be reasonable， which come from the real world and are inspired by all types of physical phenomena or biological behaviors (Kirkpatrick et al.， 1983; Kennedy and Eberhart， 1995). The the-ories about these optimization algorithms are simple and easy to understand. Second， these optimization algorithms can be thought of as a black box. This means that given a set of inputs， these algorithms can easily provide a set of outputs for any optimization problem. They are very flexible and versatile because one can change the structures and parameters of the algorithms to obtain better solutions. Third， metaheuristic algorithms can effectively avoid local optima， which is very valuable for addressing engineering problems which are typically considered as multimodal functions. In addition， one can develop their variants by absorbing the merits of other algorithms to improve the accuracy of solutions within a reasonable time. Fourth， the metaheuristic optimization algorithms can tackle different types of problems including， but not limited to， single-objective and multiobjective problems， low-dimensional and high-dimensional problems， unimodal and multi-modal problems， and discrete and continuous problems (Liu et al.， 2017; Li et al.， 2018; Duan et al.， 2018; Zhao et al.， 2019).
　　1.2 A short outline of optimization algorithms
　　Since the 1970s， many optimization algorithms have been developed and applied to different optimization problems. These algorithms， which mimic natural or physical phenomena， have provided effective and robust tech-niques for solving complex optimization problems in a wide spectrum of disciplines. Many metaheuristic algorithms with different inspirations have been proposed and successfully used in a variety of fields， which are roughly classified into four categories (Hare et al.， 2013): evolution-inspired，uhlenbein al.， al.，(Muet 1988; Gong et 2013， 2014)， physics-inspired (Geem et al.， 2001)， swarm-inspired (Krause et al.， 2013)， and human-inspired (Montiel et al.， 2007).
　　Evolution-inspired algorithms are a stochastic， population-based approach. Protecting a population’s diversity is very important for the sus-tainable development of the algorithms iteratively. Many evolution-inspired algorithms maintain a population’s diversity by mimicking basic genetic rules， including reproduction， mutation， selection， chemotaxis， elimination， and migration (Passino， 2002; Falco et al.， 2012). These algorithms ran-domly initialize a population evolved from subsequent iterations and eval-uate the individual quality using a fitness function. Genetic algorithm (GA)， originally presented by Holland (1975)， is a well-known classic evolutionary algorithm (EA). As GA can generally obtain high-quality solutions using mutation， crossover， and selection steps， the original version and its variants are widely applied to many real-world problems (Gong et al.， 2018). Since its emergence， a series of schemes aiming to enhance GA have been devel-oped. With increasing popularity of GA， a number of other evolution-based algorithms in the literature， including evolutionary strategies (ES) (Beyer and Schwefel， 2002)， differential evolution (DE) (Rocca et al.， 2011)， evo-lutionary programming (EP)(Juste et al.， 1999)， and memetic algorithms (MA) (Moscato et al.， 2007)， have been proposed. Additionally， many types of new EAs have been proposed recently， such as bacterial foraging optimi-zation (BFO) (Passino， 2002)， bat algorithm (BA) (Yang and Hossein， 2012)， fruit fly optimization algorithm (FOA) (Pan， 2012)， monkey king evolu-tionary (MKE) (Meng and Pan， 2016)， artificial algae algorithm (AAA) (Uymaz et al.， 2015)， biogeography-based optimization (BBO) (Simon， 2009)， yin-yang-pair optimization (YYPO) (Punnathanam and Kotecha， 2016)， invasive weed optimization (IWO) (Mehrabian and Lucas， 2006)， and dynamic virtual bats algorithm (DVBA) (Topal and Altun， 2016).
　　Physics-inspired algorithms simulate physical laws in the universe， among which， simulated annealing (SA) (Kirkpatrick et al.， 1983) is one of the most well-known algorithms. SA is inspired from the annealing pro-cess used in physical material in which a heated metal cools and freezes into a crystal texture with a minimum amount of energy. Recently， many novel physics-inspired algorithms have been proposed， such as gravitational search algorithm (GSA) (Rashedi et al.， 2009)， electromagnetism-like mechanism (EM)