Multi-input and multi-output modeling method based on T-S fuzzy neural network and its application

Abstract

With the advancement of science and technology, more and more complex systems require the model to have the ability to output multiple parameters simultaneously. Fuzzy neural network (FNN) is widely used in complex system modeling because of its combination of the nonlinear analysis ability of artificial neural network (ANN) and the fuzzy inference ability of fuzzy system. Therefore, this paper constructs a multi-input and multi-output (MIMO) model based on T-S (Takagi-Sugeno) FNN. First, according to the construction mechanism of TS-FNN, the MIMO network structure is designed. Then, a multi-output parameter update algorithm is designed, which takes into account the global performance and local performance of the network. Finally, simulation experiments are designed through benchmark experiments and modeling problems in an industrial process, which proves the feasibility and effectiveness of the neural network model.