Variable universe fuzzy closed-loop control of tremor predominant Parkinsonian state based on parameter estimation

Abstract

A variable universe fuzzy closed-loop control method based on parameter estimation is proposed for controlling tremor predominant Parkinsonian state. First, the computational model of thalamocortical relay neuron is established to characterize the cortico-basal ganglia-thalamocortical loop behavior in relation to Parkinsonian state. Then, in order to estimate critical parameter which exhibits the different levels of the tremor state, unscented Kalman filter is presented. Finally, an efficient control strategy on Parkinsonian state is designed by using variable universe fuzzy control theory. In the whole strategy, the slow variable that is easily reconstructed from the measured membrane potential is regarded as feedback variable, being vital to excellent control performance and low energy of the control signals. By comparing with simple proportional–integral control algorithm and ordinary fuzzy control method, it can be demonstrated that variable universe fuzzy control can avoid the repeated determinations of the controller׳s parameters, quicken the convergence speed, and improve the robustness of the controlled system, which may become a universal and valid method to alleviate any levels of the tremor state, and its control signals may apply to the current deep brain stimulation.