基于差分进化极限学习机的电力系统暂态稳定评估方法

电力系统暂态稳定性的破坏可以对电力系统的安全稳定运行产生严重冲击,准确、快速地暂稳评估方法能够提高电力系统的安全防御能力。极限学习机由于其速度快、泛化性能好被应用到电力系统暂态稳定评估中。为了提高极限学习机的评估性能,利用基于差分进化算法的优化方法和序列浮动后向特征选择算法对极限学习机暂态稳定评估性能进行提升。首先对输入特征通过主元分析降维并利用序列浮动后向算法进行特征选择,再将最优特征集输入差分进化极限学习机进行暂态稳定评估,最后在新英格兰10机39节点系统中进行验证分析,结果表明,所提模型与其他极限学习机模型相比,大大提升了其在暂态稳定分类评估中的性能。

Power System Transient Stability Assessment Based on the Differential Evolution Extreme Learning Machine

The damage of power system transient stability can cause serious accident of power grid, so the accurate and fast transient stability assessment method is of great significance to the system safe and stable operation. This paper used a method based on extreme learning machine optimized by the differential evolution algorithm (DE-ELM) and the sequence floating backward feature selection (SFBS). Firstly, the dimension of input features was reduced by principal component analysis and feature selection was performed by sequential floating backward algorithm. Then, the optimal feature set is input into DE limit learning machine for transient stability assessment. Finally, simulation was carried out in IEEE39 standard node system. The results show that the proposed model have good performance in its classification and assessment of transient stability compared with other limit learning machine models.