Performance Analysis of an Improved Single Neuron Adaptive PID Control

PID control was commonly used for control system which often depend on accurate mathematical models, but in some complicated situations, the model are hard to obtain and the plant parameters are subject to perturbations, the application of PID control are limited and adaptive schemes should be taken. BP neural networks with good model identification and can be used to satisfied PID parameters through self learning, with its adaptive learning strategy the need for a computationally intensive process was eliminated and robust control can realized. The structure of single neuron adaptive PID control system was designed and adaptive algorithm was proposed, the system was simulated by the algorithm and improved algorithm, the simulation result show the improved adaptive algorithm is feasible and the control characteristic realized.