基于改进遗传算法的非线性发电机励磁系统参数辨识

Parameters identification of nonlinear generator excitation system based on an improved genetic algorithm

为了确保电力系统建模的精确性和安全稳定分析的可靠性,进行发电机励磁系统参数辨识测试是一项重要的工作。提出采用改进的GA遗传算法应用于发电机励磁系统的非线性参数辨识,利用GA较强的全局寻优能力和BP梯度法较强的局部搜索能力,较快同时又较好地辨识出发电机励磁系统参数估计值。实际发电机励磁系统参数测试试验结果表明,基于改进遗传算法的励磁系统参数辨识方法计算速度快,精度高,鲁棒性强,为非线性发电机励磁系统的参数辨识提供了一种有效的新方法。

To ensure the accuracy of power system modeling and reliability of stabilization analysis, it is essential to test the parameters of generator excitation system. An improved genetic algorithm is applied to the parameters identification of the nonlinear generator excitation system. Combining the global optimal search of GA with the local optimal search of BP network based on Gradient method, the parameters evaluation of excitation system can be identified quickly and accurately. The effectiveness of the proposed approach is demonstrated via experiments by applying it to a generator excitation system. The improved GA has the advantages of fast computation, high precision and better robustness, and has provided a new way for parameters identification of nonlinear generator excitation system.