Dynamic programming is an exact method of determining optimal control for a discretized system. Unfortunately, for nonlinear systems the computations necessary with this method become prohibitive. This study investigates the use of adaptive neural networks that utilize dynamic programming methodology to develop near optimal control laws. First, a one dimensional infinite horizon problem is examined. Problems involving cost functions with final state constraints are considered for one dimensional linear and nonlinear systems. A two dimensional linear problem is also investigated. In addition to these examples, an example of the corrective capabilities of critics is shown. Synthesis of the networks in this study needs no external training; they do not need any apriori knowledge of the functional form of control. Comparison with specific optimal control techniques show that the networks yield optimal control over the entire range of training
S. N. Balakrishnan and V. Biega, "Adaptive Critic Based Neural Networks for Control (Low Order System Applications)," Proceedings of the 1995 American Control Conference, Institute of Electrical and Electronics Engineers (IEEE), Jan 1995.
The definitive version is available at http://dx.doi.org/10.1109/ACC.1995.529265
1995 American Control Conference
Mechanical and Aerospace Engineering
Keywords and Phrases
1D Infinite Horizon Problem; 2D Linear Problem; Adaptive Critic Based Neural Networks; Adaptive Neural Networks; Control Synthesis; Control System Synthesis; Corrective Capabilities; Cost Functions; Discretized System; Dynamic Programming; Final State Constraints; Low-Order System; Near-Optimal Control Laws; Neurocontrollers; Nonlinear Control Systems; Nonlinear Systems; One Dimensional Infinite Horizon Problem; Suboptimal Control; Two Dimensional Linear Problem
Article - Conference proceedings
© 1995 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.