Deterministic Learning Theory for Identification, Recognition, and Control

Deterministic Learning Theory for Identification, Recognition, and Control presents a unified conceptual framework for knowledge acquisition, representation, and knowledge utilization in uncertain dynamic environments. It provides systematic design approaches for identification, recognition, and control of linear uncertain systems. Unlike many books currently available that focus on statistical principles, this book stresses learning through closed-loop neural control, effective representation and recognition of temporal patterns in a deterministic way. A Deterministic View of Learning in Dynamic Environments The authors begin with an introduction to the concepts of deterministic learning theory, followed by a discussion of the persistent excitation property of RBF networks. They describe the elements of deterministic learning, and address dynamical pattern recognition and pattern-based control processes. The results are applicable to areas such as detection and isolation of oscillation faults, ECG/EEG pattern recognition, robot learning and control, and security analysis and control of power systems. A New Model of Information Processing This book elucidates a learning theory which is developed using concepts and tools from the discipline of systems and control. Fundamental knowledge about system dynamics is obtained from dynamical processes, and is then utilized to achieve rapid recognition of dynamical patterns and pattern-based closed-loop control via the so-called internal and dynamical matching of system dynamics. This actually represents a new model of information processing, i.e. a model of dynamical parallel distributed processing (DPDP).

Introduction Learning Issues in Feedback Control Learning Issues in Temporal Pattern Recognition Preview of the Main Topics RBF Network Approximation and Persistence of Excitation RBF Approximation and RBF Networks Persistence of Excitation and Exponential Stability PE Property for RBF Networks The Dgeterministic Learning Mechanism Problem Formulation Locally-Accurate Identification of Systems Dynamics Comparison with System Identification Numerical Experiments Summary Deterministic Learning From Closed-Loop Control Introduction Learning from Adaptive NN Control Learning from Direct Adaptive NN Control of Strict-Feedback Systems Learning From Direct Adaptive NN Control of Nonlinear Systems in Brunovsky Form Summary Dynamical Pattern Recognition Introduction Time-Invariant Representation A Fundamental Similarity Measure Rapid Recognition of Dynamical Patterns Dynamical Pattern Classification Summary Pattern-Based Learning Control Introduction Pattern-Based Control Learning Control Using Experiences Simulation Studies Summary Deterministic Learning with Output Measurements Introduction Learning from State Observation Non-High-Gain Observer Design Rapid Recognition of Single-Variable Dynamical Patterns Simulation Studies Summary Toward Human-Like Learning and Control Knowledge Acquisition Representation and Similarity Knowledge Utilization Toward Human-Like Learning and Control Cognition and Computation Comparison with Statistical Learning Applications of the Deterministic Learning Theory References Index

Cong Wang, David J. Hill