Single Neuron Smith Control of Network Delay Based on Least Square Support Vector Machine

A control strategy composed of Smith predictor based on time-delay prediction and single neuron PID controller was designed to compensate the time variant time-delay.The plant and network were considered as a generalized plant.The time-delay was predicted by Least Square Support Vector Machine(LS-SVM),which improve the accuracy of the prediction model.The Smith predictor was designed to compensate the system.Then a single neuron Smith control method based on time-delay prediction was formed by combining Smith predictive contro1 with single neuron PID control.The simulation results prove that this control strategy can improve the system performance and has good robustness and adaptability in case of model miss matching.