Feedback control of minimum‐time optimal control problems using neural networks

Abstract

SUMMARYThis paper presents an optimal feedback controller capable of driving a non-linear control system from an arbitrary initial state to a fixed final state in minimum time. The controller is based on a feedforward multilayer neural network trained repeatedly using open-loop optimal control data which densely span the field of extremals of the non-linear system. The effectiveness of the controller is clearly demonstrated by a simulation on a two-link robot manipulator. The effect of sensorlactuator noise and parameter variation is also included to confirm the robustness of the controller.

Synthesis of minimum-time feedback laws for dynamic systems using neural networks

“…4) and, finally, to one which used several unconventional error moments to adapt the gains of a Kalman filter (called EKF A in Ref. 5). The order in which the trackers are listed is indicative of both the tracking performance and the complexity of the algorithm.…”

Fault-Tolerant Recursive Least-Squares Computations on a Mesh-Connected Parallel Processor

“…The robot arm model used in this case study is a two-link model [25]. This two-link robot arm was chosen for this work due to the simple nature of the parameter equations for the model so that the derivation of these equations will not distract from the main idea of this work.…”

A neuro‐optimal approach for thrust‐insensitive trajectory planning

“…Lewis and Abu‐Khalaf (2004) have proposed nearly optimal state feedback control for constrained non‐linear systems through HJB equation using neural networks. Lee and Symth (1993) and Goh et al (1996) have also used neural networks in the solution of minimum time optimal control problem. Pourtakdoust et al (2005a) have developed a time optimal closed‐loop control for non‐linear lunar‐lander problem using fuzzy networks.…”