Stochastic Neural Adaptive Control for Time Varying Linear Systems based on Newton and Gradient Optimizations

Presented in this paper is a stochastic neural adaptive control algorithm, where the system identification is based on the state space innovations model |15,6,10| and a neural network architecture |10|. Additionally, this identification algorithm is derived using the Newton search optimization. The control law. also based on neural network structure, is derived from a quadratic (one-step-ahead prediction) performance index |10|, which in combination with the neural identification constitutes a unique neural adaptive control algorithm.