运用PSO-LSSVM模型的城市供电可靠性预测  

Reliability Prediction of Urban Power Network Based on PSO-LSSVM Model

作　　者：董红 石连生[2] 赵鹏程[2] 严俊[2] 

机构地区：[1]广州供电局,广州510620 [2]天津天大求实电力新技术股份有限公司,天津300384

出　　处：《电力系统及其自动化学报》2014年第7期82-86,共5页Proceedings of the CSU-EPSA

摘　　要：传统的可靠性预测方法需要配电网结构和元件可靠性指标历史数据十分准确,难以实现对城市配电网规划供电可靠性指标的预测。为此,提出一种将PSO-LSSVM(基于粒子群优化的最小二乘支持向量机)模型应用到城市电网供电可靠性预测的方法。首先通过分析影响城市供电可靠性的因素得出主要特征量;然后将这些特征量的历史数据作为输入样本,利用粒子群优化的最小二乘支持向量机方法进行建模;最后利用建立好的模型预测规划目标年城市电网供电可靠性指标。对某省多个城市电网的应用结果表明,该方法是可行且有效的。The traditional prediction of power supply reliability is on the basis of true structure of distribution network and historical data of element reliability. So it is hard to predict the planning power supply reliability of complex urban distribution network. For this reason, a least square support vector machine based on particle swarm optimization （PSO-LSSVM） model to predict power supply reliability of urban power network is proposed. Firstly, several principal characteristic quantities are received by analysing the factors of impacting power supply reliability. Then, taking his- torical data of these principal quantities as input samples, the particle swarm optimization（PSO）-least square support vector machine （LSSYM） is trained. Finally, by using the trained PSO-LSSYM, the power supply reliability indices of urban power network in target year can be predicted. The results of applying the proposed method to several urban pow- er networks show that the proposed method is effective.

关 键 词：供电可靠性 城市电网 指标预测 粒子群优化 最小二乘支持向量机 

分 类 号：TM7[电气工程—电力系统及自动化]