Wind generating set system identification method based on radial basis function (RBF) neural network technique

Abstract

The invention provides a wind generating set system identification method based on the radial basis function (RBF) neural network technique. The wind generating set system identification method comprises the following steps that 1, data required by system identification are obtained, specifically, the input data and the output data which are required by identification are obtained according to the characteristics of a wind generating set system, the sampling time selects the system internal sampling time, an input signal is the generator torque Tg during torque loop identification and is the paddle pitch angle beta during propeller pitch loop identification, and the output data are the generator rotation speed omega; and 2, system identification is conducted based on the RBF technique, specifically, the wind generating set system is described, a torque loop or a propeller pitch loop is set as a nonlinear SISO system, a nonlinear extension autoregressive East China average model NARMAX is adopted for conducting describing, and the RBF neural network training process comprises the following steps that when a signal is forwards propagated, RBF neural network output is calculated, and when an error is reversely propagated, the weights among various layers of an RBF network are adjusted by adopting the delta learning algorithm. The wind generating set system identification method based on the RBF neural network technique has good operation speed, low calculation amount and good stability.

Description

technical field [0001] The invention relates to the technical fields of wind power generation, nonlinear system identification and intelligent control, in particular to a system identification method based on RBF neural network technology. Background technique [0002] The wind turbine system is a complex nonlinear time-varying system, and it is difficult to obtain an accurate mathematical model of the system. When the working conditions of the system change, the control effect of the conventional linear model control system will be reduced, and even affect the normal operation of the entire system. Therefore, it is of great significance to establish an accurate model of the system to provide a theoretical basis for optimizing wind turbines. [0003] To establish a wind turbine system model, the solution is to use a mechanism modeling method and a nonlinear system identification method. The mechanism modeling method is used, that is, various mathematical equations are used...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the deficiencies of the existing wind turbine system identification method, such as complicated operation, large amount of calculation, slow operation speed, and poor stability, the present invention provides a method that simplifies operation, has good operation speed and low calculation amount , better stability wind turbine system identification method based on RBF (Radical Basis Function) neural network technology

Images