The proportional integral derivative PID controller was improved with the use of a Neural Network and simple hardware implementation, which will improve the control system in our high-turnover industries. The multilayer perception model is the most often used neural network model (MLP). Because it requires a desired output sequence to learn, this sort of neural network is known as a supervised network. However, we present a non-linear control of a stochastic differential equation to Neural Network matching in this paper; the model has been validated, analysed, and compared to other current controllers. The objective is to have control systems in our process industries that can achieve, improve, decrease waste, and are more flexible in terms of conversion level, as well as track set point change and reject load disturbance. This research is a first attempt at designing a simplified neutral network and proportional integral derivative PID control system for a class of non-linear processes, as well as characterising their operational features. Finally, by remodelling the proportional integral derivative PID controller with Neural Network Technique and connecting the plant process control, we were able to achieve a good result where all of the traditional proportional integral derivative PID controller's features were retained and improved using MAT-LAB. The result was great since the process industries' waste and loss were dramatically reduced to a bare minimum.
Author(s) Details:
C. E. Uchegbu,
Department of Electrical and Electronic Engineering, Abia State University
Uturu, Abia State, Nigeria.
I. I. Eneh,
Department of Electrical and Electronic Engineering, Enugu State University
of Science and Technology, Enugu, Nigeria.
M. J. Ekwuribe,
Department of Electrical and Electronic Engineering, Abia State University
Uturu, Abia State, Nigeria.
C. O. Ugwu,
Department of Electrical and Electronic Engineering, Enugu State University
of Science and Technology, Enugu, Nigeria.
No comments:
Post a Comment