Prediction of MSWI furnace temperature based on TS fuzzy neural network
Haijun He, Xi Meng, Jian Tang, Junfei Qiao, Zihao Guo
Abstract
Furnace temperature is an important indicator to control and optimize the Municipal Solid Wastes Incineration (MSWI) process. However, limited by the environment and instruments, it is difficult to measure the furnace temperature online and accurately. In this paper, a TS fuzzy neural network is utilized to design the prediction model in MSWI process, trying to obtain the real-time and accurate measurement of the furnace temperature. First, the mechanism of the MSWI process is introduced in brief. Then, the structure and training method of the TS-fuzzy-neural-network-based prediction model is introduced in details, which helps to build the nonlinear relationship between the furnace temperature and other process variables. Finally, the designed prediction model is applied to a real MSWI plant, and simulation results demonstrate the effectiveness and outperformance of the proposed methodology.