王彬　王莹　魏联滨　李楠　李朝阳

摘 要：配電网的效益具有差异性，为了提高配电网的效益预测水平，提出了基于大数据分析的差异化投资增量的配电网效益预测方法。建立了差异化投资增量配电网效益评价的指标分析模型，采用大数据分析方法进行差异化投资增量配电网效益分析。结合特征空间重构技术进行差异化投资增量配电网效益的分布式特征序列重组，提取差异化投资增量配电网效益分布特征量。采用关联规则挖掘方法对分布特征量进行效益优先性评价，利用模糊相关检测方法进行配电网效益分布大数据的抗干扰处理。对提取的差异化投资增量配电网效益分布特征量采用神经网络学习方法进行配电网的效益预测，构建效益分布的差异化因子，实现基于大数据分析的差异化投资增量配电网的效益预测。仿真测试结果表明，采用该方法进行差异化投资下的增量配电网效益预测准确性与灵敏度较高，可靠性较好。

关键词：大数据分析;差异化投资;增量配电网;效益预测

中图分类号：TM73;F272      文献标识码：A

Benefit Prediction of differential Investment Incremental

Distribution Network Based on Big Data Analysis

WANG Bin1， WANG Ying1，WEI Lian-bin1，LI Nan2，LI Zhao-yang1

（ 1. Development Planning Department，State Grid Tianjin Electric Power Company Limited，Tianjin 300000，China;

2. Chengnan Power Supply Branch， State Grid Tianjin Electric Power Company Limited，Tianjin 300000，China）

Abstract：In order to improve the efficiency prediction level of distribution network， this paper proposes a method of distribution network efficiency prediction based on big data analysis and differential investment increment， and establishes the index analysis model of the benefit evaluation of the differential investment incremental distribution network， analyzes the benefit of the differential investment incremental distribution network by using the big data analysis method， recombines the distributed characteristic sequence of the benefit of the differential investment incremental distribution network by combining the characteristic space reconstruction technology， extracts the distribution characteristic quantity of the benefit of the differential investment incremental distribution network， and uses the association rule mining method The distribution characteristic quantity is used for benefit priority evaluation， the fuzzy correlation detection method is used for anti-interference processing of big data of distribution network benefit distribution， the neural network learning method is used for benefit prediction of distribution network for the extracted differential investment increment distribution network benefit distribution characteristic quantity， the differential factor of benefit distribution is constructed， and the differential investment increment based on big data analysis is realized Benefit prediction of distribution network. The simulation test results show that the accuracy， sensitivity and reliability of the incremental distribution network benefit prediction based on this method are high.

Key words：big data analysis; differentiation investment; incremental distribution network; benefit prediction

随着高压输电和配电网建设的不断推进，对增量配电网的建设投资要求越来越大，在差异化投资规模下，需要进行增量配电网的投资效益分析和预测，建立增量配电网的投资效益预测模型，对差异化投资增量配电网的效益预测进行设计，结合差异化投资下增量配电网效益大数据分析和统计分析……