Approximation and prediction using a TSK-type neural fuzzy controller with an improved imperialist competitive algorithm

This study proposes an improved imperialist competitive algorithm (IICA) for a TSK-type neural fuzzy controller (TNFC) in order to solve prediction problems and nonlinear control system problems. This study proposes the IICA in order to reduce the use of the control parameters. Therefore, the proposed IICA adopts a Gaussian distribution to improve the assimilation policy and the exploration ability of the global solution. Finally, the experimental results show that the proposed IICA better approximates the global optimal solution and that it has faster convergence compared with other methods.