Evolutionary Computation, a broad field that includes Genetic Algorithms, Evolution Strategies, and Evolutionary Programming, has proven to offer well-suited techniques for industrial and management tasks - therefore receiving considerable attention fom scientists and engineers during the last decade. This monograph develops and analyzes evolutionary algorithms that can be successfully applied to real-world problems such as robotic control. Although of particular interest to robotic control engineers, "Evolutionary Computations" also may interest the large audience of researchers, engineers, designers and graduate students confronted with complicated optimization tasks.

Klappentext

Evolutionary computation, a broad field that includes genetic algorithms, evolution strategies, and evolutionary programming, has proven to offer well-suited techniques for industrial and management tasks - therefore receiving considerable attention from scientists and engineers during the last decade. This monograph develops and analyzes evolutionary algorithms that can be successfully applied to real-world problems such as robotic control. Although of particular interest to robotic control engineers, Evolutionary Computations also may interest the large audience of researchers, engineers, designers and graduate students confronted with complicated optimization tasks.

Inhalt

1. Evolutionary Algorithms: Revisited.- 1.1 Introduction.- 1.2 Stochastic Optimization Algorithms.- 1.2.1 Monte Carlo Algorithm.- 1.2.2 Hill Climbing Algorithm.- 1.2.3 Simulated Annealing Algorithm.- 1.2.4 Evolutionary Algorithms.- 1.3 Properties of Stochastic Optimization Algorithms.- 1.4 Variants of Evolutionary Algorithms.- 1.4.1 Genetic Algorithms.- 1.4.2 Evolution Strategies.- 1.4.3 Evolutionary Programming.- 1.4.4 Genetic Programming.- 1.5 Basic Mechanisms of Evolutionary Algorithms.- 1.5.1 Crossover Mechanisms.- 1.5.2 Mutation Mechanisms.- 1.5.3 Selection Mechanisms.- 1.6 Similarities and Differences of Evolutionary Algorithms.- 1.7 Merits and Demerits of Evolutionary Algorithms.- 1.7.1 Merits.- 1.7.2 Demerits.- 1.8 Summary.- 2. A Novel Evolution Strategy Algorithm.- 2.1 Introduction.- 2.2 Development of New Variation Operators.- 2.2.1 Subpopulations-Based Max-mean Arithmetical Crossover.- 2.2.2 Time-Variant Mutation.- 2.3 Proposed Novel Evolution Strategy.- 2.3.1 Initial Population.- 2.3.2 Crossover.- 2.3.3 Mutation.- 2.3.4 Evaluation.- 2.3.5 Alternation of Generation.- 2.4 Proposed NES: How Does It Work?.- 2.5 Performance of the Proposed Evolution Strategy.- 2.5.1 Test Functions.- 2.5.2 Implementation and Results.- 2.6 Empirical Investigations for Exogenous Parameters.- 2.6.1 Investigation for Optimal Subpopulation Number.- 2.6.2 Investigation for Optimal Degree of Dependency.- 2.7 Summary.- 3. Evolutionary Optimization of Constrained Problems.- 3.1 Introduction.- 3.2 Constrained Optimization Problem.- 3.3 Constraint-Handling in Evolutionary Algorithms.- 3.4 Characteristics of the NES Algorithm.- 3.4.1 Characteristics of the SBMAC Operator.- 3.4.2 Characteristics of the TVM Operator.- 3.4.3 Effects of the Elitist Selection.- 3.5 Construction of the Constrained Fitness Function.- 3.6 Test Problems.- 3.7 Implementation, Results and Discussions.- 3.7.1 Implementation.- 3.7.2 Results and Discussions.- 3.8 Summary.- 4. An Incest Prevented Evolution Strategy Algorithm.- 4.1 Introduction.- 4.2 Incest Prevention: A Natural Phenomena.- 4.3 Proposed Incest Prevented Evolution Strategy.- 4.3.1 Impact of Incest Effect on Variation Operators.- 4.3.2 Population Diversity and Similarity.- 4.3.3 Incest Prevention Method.- 4.4 Performance of the Proposed Incest Prevented Evolution Strategy.- 4.4.1 Case I: Test Functions for Comparison with GA, EP, ESs and NES.- 4.4.2 Case II: Test Functions for Comparison Between the NES and IPES Algorithms.- 4.5 Implementation and Experimental Results.- 4.5.1 Case I: Implementation and Results.- 4.5.2 Case II: Implementation and Results.- 4.6 Summary.- 5. Evolutionary Solution of Optimal Control Problems.- 5.1 Introduction.- 5.2 Conventional Variation Operators.- 5.2.1 Arithmetical Crossover/Intermediate Crossover.- 5.2.2 Uniform Mutation.- 5.3 Optimal Control Problems.- 5.3.1 Linear-Quadratic Control Problem.- 5.3.2 Push-Cart Control Problem.- 5.4 Simulation Examples.- 5.4.1 Simulation Example I: ESs with TVM and UM Operators.- 5.4.2 Simulation Example II: ESs with SBMAC and Conventional Methods.- 5.4.3 Implementation Details.- 5.5 Results and Discussions.- 5.5.1 Results for Example I.- 5.5.2 Results for Example II.- 5.5.3 Results from the Evolutionary Solution.- 5.6 Summary.- 6. Evolutionary Design of Robot Controllers.- 6.1 Introduction.- 6.2 A Mobile Robot with Two Independent Driving Wheels.- 6.3 Optimal Servocontroller Design for the Robot.- 6.3.1 Type-1 Optimal Servocontroller Design.- 6.3.2 Type-2 Optimal Servocontroller Design.- 6.4 Construction of the Fitness Function for the Controllers.- 6.4.1 Basic Notion.- 6.4.2 Method.- 6.5 Considerations for Design and Simulations.- 6.6 Results and Discussions.- 6.6.1 Design Results for Type-1 Controller.- 6.6.2 Design Results for Type-2 Controller.- 6.7 Summary.- 7. Evolutionary Behavior-Based Control of Mobile Robots.- 7.1 Introduction.- 7.2 An Evolution Strategy Using Statistical Information of Subgroups.- 7.2.1 Group Division.- 7.2.2 Max-mean Arithmetical Crossover.- 7.2.3 Mutation with Directly Calculated Standard Deviation.- 7.3 Omnidirectional Mobile Robot.- 7.3.1 Dynamical Mode of the Robot.- 7.3.2 Jacobian Matrix.- 7.4 Fuzzy Behavior-Based Control System.- 7.5 Acquisition of Control System.- 7.5.1 Parameter Setting.- 7.5.2 Learning Result.- 7.6 Summary.- 8. Evolutionary Trajectory Planning of Autonomous Robots.- 8.1 Introduction.- 8.2 Fundamentals of Evolutionary Trajectory Planning.- 8.3 Formulation of the Problem for Trajectory Planning.- 8.4 Polygonal Obstacle Sensing and Its Representation.- 8.4.1 Obstacle Sensing and Representation as Circles.- 8.4.2 Some Practical Considerations.- 8.5 Special Representations of Evolutionary Components.- 8.5.1 Representation of Individuals.- 8.5.2 Representation of SBMAC.- 8.5.3 Representations of Additional Operators.- 8.6 Construction of the Fitness Function.- 8.7 Bounds for Evolutionary Parameters.- 8.7.1 Bounds for Terminal Sampling Instant.- 8.7.2 Bounds for Steering Angle.- 8.8 Proposed Evolutionary Trajectory Planning Algorithm.- 8.9 Considerations and Simulations.- 8.9.1 Simulation Example I: Local Trajectory Planning.- 8.9.2 Simulation Example II: Global Trajectory Planning.- 8.10 Results and Discussions.- 8.11 Summary.- A. Definitions from Probability Theory and Statistics.- A.1 Random Variables, Distributions and Density Functions.- A.2 Characteristics Values of Probability Distributions.- A.2.1 One Dimensional Distributions:.- A.2.2 Multidimensional Distributions.- A.3 Special Distributions.- A.3.1 The Normal or Gaussian Distribution.- A.3.4 The Cauchy Distribution.- B. C-Language Source Code of the NES Algorithm.- C. Convergence Behavior of Evolution Strategies.- C.1 Convergence Reliability.- C.2 Convergence Velocity.- References.