MRAC Based PID Controller Design with Genetic Algorithm for a Single Joint Robot Arm

Abstract

The Nonlinearity of the actuators in dynamical systems leads to bad performance of the conventional feedback controllers in online execution. Furthermore, the closed loop controllers may be subjected to the changes in the environmental conditions and fluctuation of disturbances. In order to tackle this issue, this paper depicts a PID controller design based on a Model as a Reference with Genetic Adaptive Control, which is presently named (MRGAC) arrangement via employing the MIT rule in order to implement the adaption mechanism. In this abovementioned rule (MIT) the cost function is predefined as an optimization function that indicates the inaccuracy between the outputs of the working plant and the reference model used. The constraints of the controller are adjusted in a way so that this cost function is reduced to the minimum. The goal behind this work is to increase the stability of the DC motor angle that positions a single joint robot arm in free space by defining a reference model to mimic and track its response performance parameters with acceptable values. MIT rule is used for the Model Reference that is already stable and has known parameters. The contribution of this work is to adopt the gradient (MIT) with genetic algorithm (GA) that is auto tune the controller’s factors. Adaption of PID using of the optimization tactics, which is Genetic algorithm, will improve the system performance to reach steady state rapidly.