Improvement of PID Controller Based on Expert System (Neural Network)

The proportional integral derivative PID controller improved using Neural Network and easy hard ware implementation, which will improve the control system in our industries with a high turnover. The most common neural network model is the multilayer perception (MLP). This type of neural network is known as a supervised network because it requires a desired output order to learn. However, in this work, we propose a non-linear control of stochastic differential equation to Neural Network matching; the model has been validated, evaluated and compared with other existing controllers. The idea is to have control systems that will be able to achieve, improve, reduce waste and that is more flexible in the level of conversion, to be able to track set point change and reject load disturbance in our process industries. This paper represents a preliminary effort to design a simplified neutral network and proportional integral derivative PID control scheme, and modeling, their operational characteristics for a class of non-linear process. At the end we were able to achieve a good result by remodeling the proportional integral derivative PID controller with Neural Network Technique, and connected the plant process control where all the features of the traditional proportional integral derivative PID controller were retained and as well improved using MAT-LAB. The output was fantastic since the waste and loss encored by the process industries was drastically reduced to minimum.