关键词：模糊神经网络；非线性系统；控制
摘 要：The integration of fuzzy systems, Wavelet theory, and neural networks has recently become a popular approach in the engineering fields for control of nonlinear systems. Therefore, the application of Fuzzy Wavelet Neural Network controllers is clearly obvious to investigators. A lot of research has been done in the control of nonlinear systems by using the models based on type-1 Fuzzy Logic Systems (FLS). However, they are regularly unable to handle uncertainties in the rules. This chapter develops a novel structure of Type-2 Fuzzy Wavelet Neural Networks (T2FWNN) to control a nonlinear system. This has been performed by invoking some of the specific advantages of wavelets, such as dynamic compatibility, compression, and step parameter adaptation along with a combination of type-2 fuzzy concepts regarding the neural networks abilities. The proposed network is constructed based on a set of TSK fuzzy rules that includes a wavelet function in the consequent part of each rule. This can provide appropriate tools on adaptation of plant output signal to follow a desired one. In this regard, the merits of utilizing wavelets and type-2 FLS simultaneously have been discussed and explored to efficiently handle the uncertainties. It is worth mentioning that the stability of the system is effectively dependent on the learning procedure and the initial values of the network parameters. Here, an adaptive gradient descent strategy is used to adjust the unknown parameters. Furthermore, the performance of the proposed T2FWNN is compared with the type-1 FLS networks. As investigated, this method has gained considerably high levels of accuracy with the reasonable number of parameters. Finally, the efficiency of the proposed approach is demonstrated via the simulation results of two nonlinear case studies.