群体智能(Swarm intelligence):
群体智能这个概念来自对自然界中昆虫群体的观察，群居性生物通过协作表现出的宏观智能行为特征被称为群体智能。
群体智能应该遵循五条基本原则:
(1) 邻近原则( Proximity Principle) ,群体能够进行简单的空间和时间计算;
(2) 品质原则(Quality Principle) ,群体能够响应环境中的品质因子;
(3) 多样性反应原则( Principle of Diverse Re2sponse) ,群体的行动范围不应该太窄;
(4) 稳定性原则(Stability Principle) ,群体不应在每次环境变化时都改变自身的行为;
(5) 适应性原则(Adaptability Principle) ,在所需代价不太高的情况下,群体能够在适当的时候改变自身的行为。
群体智能具有如下特点:
(1) 控制是分布式的,不存在中心控制。因而它更能够适应当前网络环境下的工作状态,并且具有较强的鲁棒性,即不会由于某一个或几个个
体出现故障而影响群体对整个问题的求解。
(2) 群体中的每个个体都能够改变环境,这是个体之间间接通信的一种方式,这种方式被称为“激发工作”(Stigmergy) 。由于群体智能可以通过非直接通信的方式进行信息的传输与合作,因而随着个体数目的增加,通信开销的增幅较小,因此,它具有较好的可扩充性。
(3) 群体中每个个体的能力或遵循的行为规则非常简单,因而群体智能的实现比较方便,具有简单性的特点。
(4) 群体表现出来的复杂行为是通过简单个体的交互过程突现出来的智能( Emergent Intelli2gence) ,因此,群体具有自组织性。
英文版.群体智能
作者:(美国)James Kennedy (美国)Russell C Eberhart,Yuhui Shi(史玉回)
市场价：¥75.00
出版社：人民邮电出版社
页码：512 页
出版日期：2009年
ISBN：9787115195500
版本：1版
装帧：平装
开本：16
英语：英语
丛书名：图灵原版计算机科学系列
内容简介群体智能是通过模拟自然界生物群体行为来实现人工智能的一种方法。《群体智能》综合运用认知科学、社会心理学、人工智能和演化计算等学科知识，提供了一些非常有价值的新见解，并将这些见解加以应用，以解决困难的工程问题。书中首先探讨了基础理论，然后详尽展示如何将这些理论和模型应用于新的计算智能方法(粒子群)中，以适应智能系统的行为，最后描述了应用粒子群优化算法的好处，提供了强有力的优化、学习和问题解决的方法。
《群体智能》主要面向计算机相关学科的高年级本科生或研究生以及相关领域的研究与开发技术人员。
作者简介James Kennedy社会心理学家。自1994年起，他一直致力于粒子群算法的研究工作，并与Russell C．Eberhart共同开发了粒子群优化算法。目前在美国劳工部从事调查方法的研究工作。他在计算机科学和社会科学杂志和学报上发表过许多关于粒子群的论文。
RusselI C．Eberhart普度大学电子与计算机工程系主任。IEEE会士。与JamesKennedy共同提出了粒子群优化算法。曾任IEEE神经网络委员会的主席。除了本书之外，他还著有《计算智能：从概念到实现》（影印版由人民邮电出版社出版）等。
Yuhui Shi (史玉回)国际计算智能领域专家，现任Joumal ofSwarm Intellgence编委，IEEE CIS群体智能任务组主席，西交利物浦大学电子与电气工程系教授。1992年获东南大学博士学位，先后在美国、韩国、澳大利亚等地从事研究工作，曾任美国电子资讯系统公司专家长达9年。他还是《计算智能：从概念到实现》一书的作者之一。
媒体推荐“本书内容丰富，富于启发性和思想性，强烈推荐给所有的演进计算研究人员。”
——Genetic Programming and Evolvable'Machines
“这本书极为出色，不愧为PSO和群体智能的最佳参考书：”
——Konstantions E.Parsopoulos 希腊Palras大学
编辑推荐群体智能是近年来发展迅速的人工智能学科领域。通过研究分散、自组织的动物群体和人类社会的智能行为，学者们提出了许多迥异于传统思路的智能算法，很好地解决了不少原来非常棘手的复杂工程问题。与蚁群算法齐名的粒子群优化（particle swarm optimizatiotl，简称PSO）算法就是其中最受瞩目、应用最为广泛的成果之一。
《群体智能》由粒子群优化算法之父撰写，是该领域毋庸置疑的经典著作。作者提出，人类智能来源于社会环境中个体之间的交互，这种智能模型可以有效地应用到人工智能系统中去。书中首先从社会心理学、认知科学和演化计算等多个角度阐述了这种新方法的基础，然后详细说明了应用这些理论和模型所得出的新的计算智能方法——粒子群优化，进而深入地探讨了如何将粒子群优化应用于广泛的工程问题。
《群体智能》的C及ViSLlaI Basic源代码可以在图灵网站（WWW．turingbook．com）《群体智能》网页免费注册下载。
目录part one Foundations
chapter oneModels and Concepts of Life and Intelligence3
The Mechanics of Life and Thought4
Stochastic Adaptation: Is Anything Ever Really Random?9
The “Two Great Stochastic Systems”12
The Game of Life: Emergence in Complex Systems16
The Game of Life17
Emergence18
Cellular Automata and the Edge of Chaos20
Artificial Life in Computer Programs26
Intelligence: Good Minds in People and Machines30
Intelligence in People: The Boring Criterion30
Intelligence in Machines: The Turing Criterion32
chapter twoSymbols, Connections, and Optimization by Trial and Error35
Symbols in Trees and Networks36
Problem Solving and Optimization48
A Super-Simple Optimization Problem49
Three Spaces of Optimization51
Fitness Landscapes52
High-Dimensional Cognitive Space and Word Meanings55
Two Factors of Complexity: NK Landscapes60
Combinatorial Optimization64
Binary Optimization67
Random and Greedy Searches71
Hill Climbing72
Simulated Annealing73
Binary and Gray Coding74
Step Sizes and Granularity75
Optimizing with Real Numbers77
Summary78
chapter threeOn Our Nonexistence as Entities: The Social Organism81
Views of Evolution82
Gaia: The Living Earth83
Differential Selection86
Our Microscopic Masters?91
Looking for the Right Zoom Angle92
Flocks, Herds, Schools, and Swarms: Social Behavior as Optimization94
Accomplishments of the Social Insects98
Optimizing with Simulated Ants: Computational Swarm Intelligence105
Staying Together but Not Colliding: Flocks, Herds, and Schools109
Robot Societies115
Shallow Understanding125
Agency129
Summary131
chapter fourEvolutionary Computation Theory and Paradigms133
Introduction134
Evolutionary Computation History134
The Four Areas of Evolutionary Computation135
Genetic Algorithms135
Evolutionary Programming139
Evolution Strategies140
Genetic Programming141
Toward Unification141
Evolutionary Computation Overview142
EC Paradigm Attributes142
Implementation143
Genetic Algorithms146
An Overview146
A Simple GA Example Problem147
A Review of GA Operations152
Schemata and the Schema Theorem159
Final Comments on Genetic Algorithms163
Evolutionary Programming164
The Evolutionary Programming Procedure165
Finite State Machine Evolution166
Function Optimization169
Final Comments171
Evolution Strategies172
Mutation172
Recombination174
Selection175
Genetic Programming179
Summary185
chapter fiveHumans—Actual, Imagined, and Implied187
Studying Minds188
The Fall of the Behaviorist Empire193
The Cognitive Revolution195
Bandura’s Social Learning Paradigm197
Social Psychology199
Lewin’s Field Theory200
Norms, Conformity, and Social Influence202
Sociocognition205
Simulating Social Influence206
Paradigm Shifts in Cognitive Science210
The Evolution of Cooperation214
Explanatory Coherence216
Networks in Groups218
Culture in Theory and Practice220
Coordination Games223
The El Farol Problem226
Sugarscape229
Tesfatsion’s ACE232
Picker’s Competing-Norms Model233
Latané’s Dynamic Social Impact Theory235
Boyd and Richerson’s Evolutionary Culture Model240
Memetics245
Memetic Algorithms248
Cultural Algorithms253
Convergence of Basic and Applied Research254
Culture—and Life without It255
Summary258
chapter sixThinking Is Social261
Introduction262
Adaptation on Three Levels263
The Adaptive Culture Model263
Axelrod’s Culture Model265
Experiment One: Similarity in Axelrod’s Model267
Experiment Two: Optimization of an Arbitrary Function268
Experiment Three: A Slightly Harder and More Interesting Function269
Experiment Four: A Hard Function271
Experiment Five: Parallel Constraint Satisfaction273
Experiment Six: Symbol Processing279
Discussion282
Summary284
part twoThe Particle Swarm and Collective Intelligence
chapter sevenThe Particle Swarm287
Sociocognitive Underpinnings: Evaluate, Compare, and Imitate288
Evaluate288
Compare288
Imitate289
A Model of Binary Decision289
Testing the Binary Algorithm with the De Jong Test Suite297
No Free Lunch299
Multimodality302
Minds as Parallel Constraint Satisfaction Networks in Cultures307
The Particle Swarm in Continuous Numbers309
The Particle Swarm in Real-Number Space309
Pseudocode for Particle Swarm Optimization in Continuous Numbers313
Implementation Issues314
An Example: Particle Swarm Optimization of Neural Net Weights314
A Real-World Application318
The Hybrid Particle Swarm319
Science as Collaborative Search320
Emergent Culture, Immergent Intelligence323
Summary324
chapter eightVariations and Comparisons327
Variations of the Particle Swarm Paradigm328
Parameter Selection328
Controlling the Explosion337
Particle Interactions342
Neighborhood Topology343
Substituting Cluster Centers for Previous Bests347
Adding Selection to Particle Swarms353
Comparing Inertia Weights and Constriction Factors354
Asymmetric Initialization357
Some Thoughts on Variations359
Are Particle Swarms Really a Kind of Evolutionary Algorithm?361
Evolution beyond Darwin362
Selection and Self-Organization363
Ergodicity: Where Can It Get from Here?366
Convergence of Evolutionary Computation and Particle Swarms367
Summary368
chapter nineApplications369
Evolving Neural Networks with Particle Swarms370
Review of Previous Work370
Advantages and Disadvantages of Previous Approaches374
The Particle Swarm Optimization Implementation Used Here376
Implementing Neural Network Evolution377
An Example Application379
Conclusions381
Human Tremor Analysis382
Data Acquisition Using Actigraphy383
Data Preprocessing385
Analysis with Particle Swarm Optimization386
Summary389
Other Applications389
Computer Numerically Controlled Milling Optimization389
Ingredient Mix Optimization391
Reactive Power and Voltage Control391
Battery Pack State-of-Charge Estimation391
Summary392
chapter tenImplications and Speculations393
Introduction394
Assertions395
Up from Social Learning: Bandura398
Information and Motivation399
Vicarious versus Direct Experience399
The Spread of Influence400
Machine Adaptation401
Learning or Adaptation?402
Cellular Automata403
Down from Culture405
Soft Computing408
Interaction within Small Groups: Group Polarization409
Informational and Normative Social Influence411
Self-Esteem412
Self-Attribution and Social Illusion414
Summary419
chapter elevenAnd in Conclusion . . .421
Appendix A Statistics for Swarmers429
Appendix B Genetic Algorithm Implementation451
Glossary457
References475
Index497
……
序言At this moment a half．dozen astronauts are assembling a new space station hundreds of miles above the surface of the earth．Thousands of sailors live and work under the sea in submarines．Incas iog through theAndes．Nomads roam the Arabian sands．Homo sapiensliterally,“intelli-gent man” has adapted to nearly every environment on the face of theearth．below it，and as far above it as we can propel ourselves．W_e must bedoing something right．In this book we argue that what we do right is related to our socialit．We will investigate that elusive quality known as intelligence，which isconsidered first of all a trait of humans and second as something thatmight be created in a computer,and our conclusion will be that whatever this“intelligence”is。it arises from interactions among individuals．We humans are the most social of animals：we live together in families，tribes．cities，nations，behaving and thinking according to the rules andnorms of our communities，adopting the customs of our fellows，including the facts they believe and the explanations they use to tie those factstogether．Even when we are alone，we think about other people，andeven when we think about inanimate things，we think using language the medium of interpersonal communication．
