Abstract—In this paper, Cerebellar Model Articulation Controller (CMAC) is used in conjunction with feedback control law to control a second order unstable plant. The memory of CMAC is updated according to the feedback error which makes the learning rate of CMAC sensitive to feedback error and hence responsible for the convergence of error. The paper starts with the comparison of unit step response of second order plant for different learning rates which show that the proposed variation in CMAC is sensitive to learning rates. Performance of the response is measured in terms in terms of key characteristics like rise time, peak overshoot, settling time and integral square error. Then discontinuous and continuous variable learning rate schemes are proposed to ameliorate the response of the plant in terms of peak overshoot, settling time and integral square error which is well supported by the simulation results done in MATLAB Simulink.

 

Index Terms—Cerebellar model articulation controller, discontinuous variable learning rate CMAC, continuous variable learning rate CMAC

Cite: A. Rawat and M. J. Nigam, "Feedback Error Based Discontinuous and Continuous Variable Learning Rate CMAC," International Journal of Electronics and Electrical Engineering, Vol. 3, No. 4, pp. 305-310, August 2015. doi: 10.12720/ijeee.3.4.305-310