B-spline curve has plenty of beneficial properties, such as easy local adjusting,simple calculation and implementation. It could be used as the membership function of the fuzzy logic system, thereby form B-spline neurofuzzy network, which is believed to have advantages of both fuzzy logic system and neural network. In this paper, a controller constructed with B-spline neurofuzzy network is proposed. The network is optimized with a special designed generic algorithm, which could optimize almost all the parameters of the network, such as distribution of knots, the order of B-spline curve and the connection weights, thus, find the global optimal solution for the controller.The controller is applied to control the double inverted pendulum, which is an unstable, nonlinear, and multi-variables system. The simulation result withrnMATLAB/SIMULINK is positive, which verify the effectiveness of the controller.
展开▼