基于RBF神经网络的水轮机调节系统辨识

针对水轮机确切数学模型难以建立的问题以及水轮机调节系统非线性动态仿真的复杂性,利用RBF神经网络的局部逼近特性和快速收敛能力,实现水轮机调节系统非线性特性的辨识建模。将该模型应用于水轮机调节系统仿真,能快速准确地得到系统及机组内部各参数的变化规律。仿真结果表明,模型精度高,实用性强,从而为调节系统过渡过程的计算以及高级控制策略的研究提供了有力的支持。

Recognition of Hydraulic Turbine Governing System Based On RBF Neural Network

Aimed at a problem that it is difficult to establish the accurate model of turbines and the complexity of the dynamic simulation of the Hydraulic Turbine Governing System with non-linear properties, a model that contains the non-linear properties of the Hydraulic Turbine Governing System is established by using the local approaching property and quick convergence ability of RBF Neural Network. Applying the model in the simulation of Hydraulic Turbine Governing System, the change rules of both the system and the turbine's internal parameters can be found rapidly and accurately. The analysis and simulation proves that the model has high precision and good practicability. Thereby, it can provide a strong support for the calculation of the transition process of Hydraulic Turbine Governing System and the study on the advanced control strategies.