基于马尔科夫模型的配电网空间负荷智能预测方法  

Markov Model-based Intelligent Prediction Method for Spatial Load of Distribution Network

作　　者：赵青宇 ZHAO Qingyu(China Energy Engineering Group Guangdong Electric Power Design Institute Co.,Ltd.,Guangzhou 510000,China)

机构地区：[1]中国能源建设集团广东省电力设计研究院有限公司,广东广州510000

出　　处：《电工技术》2023年第24期60-62,65,共4页Electric Engineering

摘　　要：常规的配电网空间负荷智能预测流程一般是独立的,预测的范围较小,导致预测负荷相对误差不断增加,为此提出基于马尔科夫模型的配电网空间负荷智能预测方法的设计与验证分析。根据当前实际的电网空间负荷智能预测需求及标准,先对电网空间负荷指标进行求取,采用多目标的方式,扩大负荷预测的范围,设计多目标智能预测流程。以此为基础,构建马尔科夫体系电网空间负荷预测模型,采用多维模糊修正实现负荷智能预测处理。最终的测试结果表明:对比于传统并行数据计算配电网空间负荷智能预测小组、传统模糊概率配电网空间负荷智能预测小组,此次所设计的马尔科夫模型配电网空间负荷智能预测小组最终得出的预测负荷相对误差被较好地控制在3%以下,说明该预测方法的针对性与稳定性更高,预测的效率及质量更加稳定,误差可控,具有实际的应用价值。The conventional spatial load intelligent prediction process for distribution networks is generally independent,with a small prediction range,resulting in an increasing relative error of predicted loads.Therefore,this paper proposes the design and validation analysis of a distribution network spatial load intelligent prediction method based on Markov models.Based on the current actual demand and standards for intelligent prediction of power grid spatial load,the spatial load indicators of power grid are first calculated,and a multi-objective approach is adopted to expand the scope of load prediction and design a multi-objective intelligent prediction process.Based on this,a Markov system power grid spatial load prediction model is constructed,and multi-dimensional fuzzy correction is used to achieve intelligent load prediction processing.The final test results show that compared with traditional parallel data calculation spatial load intelligent prediction group and traditional fuzzy probability spatial load intelligent prediction group,the proposed Markov model distribution network spatial load intelligent prediction group can achieve satisfactory control of relative error of predicted load at less than 3%,demonstrating higher specificity and stability,more stable efficiency and quality,controllable errors,and practical application value of this method.

关 键 词：马尔科夫模型 配电网空间 空间负荷 智能预测 预测方法 电网控制 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN99[自动化与计算机技术—计算机科学与技术]