The purpose of this monograph is to give the broad aspects of nonlinear identification and control using neural networks. It uses a number of simulated and industrial examples throughout, to demonstrate the operation of nonlinear identification and control techniques using neural networks.
1. Neural Networks.- 1.1 Introduction.- 1.2 Model of a Neuron.- 1.3 Architectures of Neural Networks.- 1.3.1 Single Layer Networks.- 1.3.2 Multilayer Networks.- 1.3.3 Recurrent Networks.- 1.3.4 Lattice Networks.- 1.4 Various Neural Networks.- 1.4.1 Radial Basis Function Networks.- 1.4.2 Gaussian RBF Networks.- 1.4.3 Polynomial Basis Function Networks.- 1.4.4 Fuzzy Neural Networks.- 1.4.5 Wavelet Neural Networks.- 1.4.6 General Form of Neural Networks.- 1.5 Learning and Approximation.- 1.5.1 Background to Function Approximation.- 1.5.2 Universal Approximation.- 1.5.3 Capacity of Neural Networks.- 1.5.4 Generalisation of Neural Networks.- 1.5.5 Error Back Propagation Algorithm.- 1.5.6 Recursive Learning Algorithms.- 1.5.7 Least Mean Square Algorithm.- 1.6 Applications of Neural Networks.- 1.6.1 Classification.- 1.6.2 Filtering.- 1.6.3 Modelling and Prediction.- 1.6.4 Control.- 1.6.5 Hardware Implementation.- 1.7 Mathematical Preliminaries.- 1.8 Summary.- 2. Sequential Nonlinear Identification.- 2.1 Introduction.- 2.2 Variable Neural Networks.- 2.2.1 Variable Grids.- 2.2.2 Variable Networks.- 2.2.3 Selection of Basis Functions.- 2.3 Dynamical System Modelling by Neural Networks.- 2.4 Stable Nonlinear Identification.- 2.5 Sequential Nonlinear Identification.- 2.6 Sequential Identification of Multivariable Systems.- 2.7 An Example.- 2.8 Summary.- 3. Recursive Nonlinear Identification.- 3.1 Introduction.- 3.2 Nonlinear Modelling by VPBF Networks.- 3.3 Structure Selection of Neural Networks.- 3.3.1 Off-line Structure Selection.- 3.3.2 On-line Structure Selection.- 3.4 Recursive Learning of Neural Networks.- 3.5 Examples.- 3.6 Summary.- 4. Multiobjective Nonlinear Identification.- 4.1 Introduction.- 4.2 Multiobjective Modelling with Neural Networks.- 4.3 Model Selection by Genetic Algorithms.- 4.3.1 Genetic Algorithms.- 4.3.2 Model Selection.- 4.4 Multiobjective Identification Algorithm.- 4.5 Examples.- 4.6 Summary.- 5. Wavelet Based Nonlinear Identification.- 5.1 Introduction.- 5.2 Wavelet Networks.- 5.2.1 One-dimensional Wavelets.- 5.2.2 Multi-dimensional Wavelets.- 5.2.3 Wavelet Networks.- 5.3 Identification Using Fixed Wavelet Networks.- 5.4 Identification Using Variable Wavelet Networks.- 5.4.1 Variable Wavelet Networks.- 5.4.2 Parameter Estimation.- 5.5 Identification Using B-spline Wavelets.- 5.5.1 One-dimensional B-spline Wavelets.- 5.5.2 n-dimensional B-spline Wavelets.- 5.6 An Example.- 5.7 Summary.- 6. Nonlinear Adaptive Neural Control.- 6.1 Introduction.- 6.2 Adaptive Control.- 6.3 Adaptive Neural Control.- 6.4 Adaptation Algorithm with Variable Networks.- 6.5 Examples.- 6.6 Summary.- 7. Nonlinear Predictive Neural Control.- 7.1 Introduction.- 7.2 Predictive Control.- 7.3 Nonlinear Neural Predictors.- 7.4 Predictive Neural Control.- 7.5 On-line Learning of Neural Predictors.- 7.6 Sequential Predictive Neural Control.- 7.7 An Example.- 7.8 Summary.- 8. Variable Structure Neural Control.- 8.1 Introduction.- 8.2 Variable Structure Control.- 8.3 Variable Structure Neural Control.- 8.4 Generalised Variable Structure Neural Control.- 8.5 Recursive Learning for Variable Structure Control.- 8.6 An Example.- 8.7 Summary.- 9. Neural Control Application to Combustion Processes.- 9.1 Introduction.- 9.2 Model of Combustion Dynamics.- 9.3 Neural Network Based Mode Observer.- 9.4 Output Predictor and Controller.- 9.5 Active Control of a Simulated Combustor.- 9.6 Active Control of an Experimental Combustor.- 9.7 Summary.
