UKF-based key-parameters compensation control for abnormal firing in PR model

摘要

Abnormal firing of neurons is often caused by the abnormality of one or more parameters. These parameters can't be measured in actual physical experiment, but can be estimated with estimator. Brain is a complex nonlinear system with a large amount of noise. UKF (Unscented Kalman Filter) is a good choice to estimate the parameters of nonlinear system. In this paper we design a compensation control based on UKF to recover the morbid system. Take the deviation between the estimated and the standard parameters as control signal and use this signal to control the firing state of the neurons. It is found that our control can eliminate the abnormal firing caused by two or more abnormal parameters.