Movable propeller turbine neural network model and nonlinear simulation

Due to the difficulty in describing the nonlinear characteristic of movable propeller turbine and the complex simulation of the movable propeller turbine governing system, this paper takes advantage of the powerful nonlinear approximate ability of the feed forward neural network to put up the integrated characteristic neural network model NZZM firstly. Then put up the runaway characteristic neural network model to obtain the no-load opening and no-load flow being the control point extending to small opening quickly and accuracy and then obtain the integrated characteristic neural network model describing the flow characteristic and efficiency characteristic under big and small opening. At last, the coordinated characteristic neural network NZZC is put up. The Levenberg-Marquardt algorithm is used to train the above neural network models. The convergence speed of the offline training is fast and the accuracy of the model is high. NZZM, NZZC and other models consist of the movable propeller turbine neural network model ZZ587. Matlab and Simulink are used for the nonlinear simulation of the movable propeller turbine neural network model ZZ587. The variability of the different inner parameters of the system and the turbine can be attained quickly and truly. It provides a good base for the research of control policy of the movable propeller turbine governing system.