Dissolved oxygen control system based on the T-S fuzzy neural network

In this paper, a novel kind of the dissolved oxygen (DO) concentration control system was proposed based on the T-S fuzzy neural network. The proposed T-S fuzzy neural network controller was used to control the DO concentration in the Benchmark Simulation Model No.1 (BSM1) wastewater treatment platform. The parameters of the neural network were adjusted online through the error back propagation algorithm to get the minimum error. By adjusting the learning rate online, the convergence speed of the system was accelerated, and then the DO concentration in the wastewater treatment system was controlled fast and efficiently in real-time. Compared with BP and PID controllers through the digital simulation, the results showed that the control effect of the DO concentration based on T-S fuzzy neural network control system was better. Besides, the test results under three kinds of weather condition showed that better adaptability and robustness were also gained in this control system.