The establishment of the fuzzy control rule library determines the performance of the fuzzy control system. And in the fields of energy production and robot control, higher and higher control-accuracy requirements make the establishment method of conventional fuzzy control rule library sometimes no longer applicable. Therefore, we proposed an improved bacterial foraging optimization (IBFO) algorithm on the basis of swarm intelligence algorithm to improve the establishment of rule library. Firstly, the imperfectness of fuzzy rule library based on artificial experience induction was analyzed. Secondly, the improved fuzzy control system was described. Finally, the Gaussian membership function parameters of the improved TSK fuzzy system (C-ATSKFS,constant-ameliorative TSK fuzzy system) rule library were optimized. Compared with the existing method, the improved algorithm can effectively increase the recognition accuracy of the fuzzy control system. MATLAB simulation results show that the proposed novel foraging algorithm has a high practical value for the establishment of fuzzy control rule library.