Neural network-based state estimation of nonlinear systems: application to fault detection and isolation

'Neural Network-Based State Estimation of Nonlinear Systems' presents efficient, easy to implement neural network schemes for state estimation, system identification, and fault detection and Isolation with mathematical proof of stability, experimental evaluation, and Robustness against unmolded dynamics, external disturbances, and measurement noises. Presents both the Linear-in-Parameter Neural Network based observer and the Nonlinear-in-Parameter Neural Network based observer approaches to nonlinear systems Discusses the neural network structure for fault detection actuators using an application to satellite attitude control systems and robotic manipulators Discusses robust sensor and actuator fault detection and estimation INDICE: Introduction.- Observation (LPNN and NPNN).- Identification (LPNN and NPNN).- Actuator Fault Detection and Isolation: Experiments in Robotic Manipulators.- A Robust Actuator Gain Fault Detection and Isolation Scheme.- A Robust Sensor and Actuator Fault Detection and Estimation: Application to a Satellite's Attitude Control Subsystem.- Preliminary Definitions.- Neural Network Learning Rules for Theorem 6.1.- Stability Conditions of Theorem 6.1-Part 2.

• ISBN: 978-1-4419-1437-8
• Editorial: Springer
• Encuadernacion: Rústica
• Páginas: 175
• Fecha Publicación: 01/11/2009
• Nº Volúmenes: 1
• Idioma: Inglés