An Improved Single Neuron Adaptive PID Controller System Based on Additional Error of an Inversed-Control Signal

This paper presents a novel single neuron adaptive PID controller system that developed by utilizing an additional error of the inversed-control signal from the plant output on its learning mechanism, in order to update the neuron connection weights. This additional error is calculated from the difference between an inversed-control signal through an inverse neural network and the input control signal of the plant. Actually, this additional error is derived mathematically through a new quadratic error performance function that is utilized for learning the single neuron adaptive PID controller. By using this new quadratic error performance function of the controller system, the convergence speed of the adaptive neuron is increased and the control performance of the novel single neuron adaptive PID controller system is greatly improved. The control system parameters such as the rise time, settling time, and the overshoot of the system of the proposed algorithm have been compared with that of the classical single neuron adaptive PID controller system, and the simulation results show that the control effect of this novel controller system outperformed the classical single neuron adaptive PID controller, especially in its strong robustness and better self-adaptation characteritics.