Authors: Xue-Ling Song; Zhe-Ying Song; Chao-Ying Liu
Addresses: College of Electrical Engineering & Informational Science, Hebei University of Science and Technology, Shijiazhuang, Heibei, China ' College of Electrical Engineering & Informational Science, Hebei University of Science and Technology, Shijiazhuang, Heibei, China ' Polytechnic College, Hebei University of Science and Technology, Shijiazhuang, Heibei, China
Abstract: The principle of heating exchange system is analysed in this paper. According to the strong coupling between the variations and parameters time variable characteristics, it is difficult to set up the arithmetical model and achieve better control effects. To overcome these difficulties, this paper adopts T-S fuzzy model to describe the input-output relationship based on collected data of system. Due to the time variable character, it is hard to control with traditional PID controller. To improve the control effect, combining neural network control algorithm, this paper adopts PIDNN (PID Neural Network) controller. The PIDNN algorithm is not just tuning PID parameters by neural network, and improves the system control ability by defining proportional, integral, derivative characteristics neurons. Basing on the theoretical analysis, simulations are carried out by using MATLAB. The results of simulation experiment show that the designed model is effective and the control system achieves satisfied control effect.
Keywords: heating exchange systems; T-S fuzzy modelling; PIDNN control; neural networks; PID control; time-varying characteristic; fuzzy logic; simulation; heat exchangers; hot water; thermal power plants.
International Journal of Wireless and Mobile Computing, 2014 Vol.7 No.1, pp.94 - 101
Available online: 14 Jan 2014Full-text access for editors Access for subscribers Purchase this article Comment on this article