Neural Networks: Algorithms and Applications

0.0

Providing a comprehensive introduction to the use of neural networks in mechanical engineering applications this book begins with an overview of different neural network topologies in the first two parts. Functioning of human brain is explained as an analogy with artificial models. Unsupervised models like Hopfield, Bi-directional Associative Memory, fuzzy Associative Memory, Adaptive Resonance Theory, kohonen as well as supervised architectures like Multi-Layer Perceptron, Counter Propagation networks and Radial Basis Function Networks are presented in detail with numerical examples. The third part deals with applications of artificial neural networks for solving of design optimization problems, forward and inverse dynamic analysis applications and system identification and monitoring, as well as motion and vibration control in robotics and structural engineering. Software implementations for neural networks in C/C++ language and necessary optimization techniques in network training are given in Appendices. Key Features • Latest developments in standard neural network architectures like Fuzzy ARTMAP • Network models such as Probabilistic and General Regression Neural Net works (GRNN) • Carefully designed simulation examples Table of Contents Preface PART- 1: OVERVIEW: Introduction Brain as a neural network Basic properties of neurons Artificial neural networks Conclusions Exercises Learning Learning and training Learning rules Stability and plasticity Conclusions Exercises PART- II: TYPES OF NEURAL NETWORKS Hopfield Perceptron and related models Hopfield model Basic models of Hopfield network Cellular neural networks Perceptron Other Associative models Exercises Appendix Adaptive Resonance Theory Network for ART-1 ART-2 Network Conclusions Exercises Self-Organization Maps Kohonen Map Adaptive or learning vector quantization Multilayer self-organizing feature map Exercises Feed-forward back Propagation networks Training of multilayer feed-forward networks by backpropagation Training aspects and variations of backpropagation method Backpropagation as stochastic approximation Conclusions Exercises Hybrid learning neural networks Counter Propagation network Radial Basis Function Networks Exercises Probabilistic models, fuzzy ARTMAP and recurrent networks Probabilistic neural networks General Regression Neural networks Fuzzy ARTMAP Recurrent backpropagation neural networks Exercises PART- III: APPLICATIONS Application of neural networks Design and optimization of systems System identification and Monitoring Patterns Recognition Applications Motion and vibration control applications Summary Exercises Appendix-I Introduction of object oriented programming Appendix-II Optimization schemes used in neural networks Glossary Index