Design of a robust neuro-controller for complex dynamic systems

Design of neuro-controller for complex dynamic systems is a big challenge faced by the researchers. In this paper we present a design of a robust neuro-controller for a dynamic system to make the system response fast with no overshoot. Here the control action decided by the controller completely depends on the value of the error at that point of time. The position feedback which controls the bandwidth of the system as well as the dynamic response is a function of the system error. For large error the position feedback is made large increasing the bandwidth of the system, and for small errors the position feedback value is small. Thus, during the dynamic response of the system the bandwidth of the system is controlled by the system error. Similarly, the velocity feedback which controls the damping in the system is kept very small for large errors, and large for small errors. Thus, in the proposed neuro-controller the position feedback K_p(e,t), and velocity feedback K_v(e,t) are made as a function of error which yields a very fast response with no or very little overshoot.