基于改进网格化算法的智能化配电网精细分析模型设计

Design of fine analysis model for intelligent distribution network based on improved gridding algorithm

针对配电网运行水平影响指标众多、分析不精准的问题,文中采用网格化和Elastic-net算法相结合的方式,设计了一种智能化配电网精细分析模型。该模型将配网网格化后再进行评级,并确立了可改造或规划的区域。同时进一步结合Elastic-net算法从所有影响指标中选出对目标指标影响程度较大的指标,且将其按重要度排序,从而实现了对配网的智能化精细分析。以某配网历史数据进行的实例验证结果表明,与未经筛选便确定相关性的灰色关联算法相比,Elastic-net算法能够在筛选指标时排除弱相关指标。此外,Elastic-net算法还对筛选出的指标进行了重要度排序,进而提升了分析结果的精确性,因此可为精细的配网数据分析与精准的投资方案制定提供支撑。

Aiming at the problems of numerous indicators affecting the operation level of distribution network and inaccurate analysis,an intelligent distribution network fine analysis model is designed by combining grid and Elastic-net algorithm.The model establishes the area that can be transformed or planned by grading the distribution network after grid.Further,it combines the Elastic-net algorithm to select the indicators that have a great impact on the target indicators from all the impact indicators,and sort them according to the importance,so as to realize the intelligent and fine analysis of the distribution network.The example verification results based on the historical data of a distribution network show that,compared with the grey correlation algorithm that determines the correlation without screening,using the Elastic-net algorithm to screen the indicators can eliminate the weak correlation indicators.At the same time,the Elastic-net algorithm also sorts the importance of the screened indicators,improves the accuracy of the analysis results,and provides support for the fine distribution network data analysis and accurate investment scheme formulation.