基于改进型反向传播网络预测电压稳定裕度的计算

Forecast of Calculation on Voltage Stability Margin Based on Improved back-propagation Networks

针对系统电压稳定裕度问题,提出一种基于反向传播(back-propagation,BP)网络算法的新算法——自适应学习速率-附加动量法和弹性BP算法相结合的改进型BP算法,该方法利用了BP算法的非线性处理性,模拟出负荷节点的P-U曲线,并根据曲线计算出电压稳定裕度和电压稳定薄弱点。改进型BP算法加快了BP算法的收敛速度,提高了算法的计算速度,减少了计算时间。最后以IEEE5节点系统为例,准确计算出电压稳定裕度和电压稳定薄弱节点,表明了基于改进型BP网络预测电压稳定裕度计算法的准确性。

Aiming at voltage stability margin, the paper brings forward a new algorithm--the improved BP algorithm, which is based on BP algorithm and integrates self-adaptive learning rate-additional momentum method with resilient BP algorithm. By exploiting non-linear process of BP algorithm, P-U curve of load bus is simulated, voltage stability margin and weak points are worked out according to the curve. The improved BP algorithm quickens convergence rate of BP algorithm, promotes calculation speed and reduces the time for calculation. In the end, IEEE 5-bus system is taken as an example, voltage stability margin and weak bus of voltage stability are accurately figured out, which demonstrates forecast accuracy of algorithm for voltage stability margin based on improved BP networks.