基于PSO-LSSVM的电力造价灵敏度分析  

Sensitivity Analysis of Power Cost Based on PSO-LSSVM

作　　者：彭光金[1] 王富平[1] 么远[1] 朱辉[2] 刘瑜[2] 邢晓蕊 

机构地区：[1]重庆大学输配电装备及系统安全与新技术国家重点实验室,重庆400030 [2]重庆市电力公司,重庆400016

出　　处：《电工技术学报》2013年第S2期391-394,共4页Transactions of China Electrotechnical Society

摘　　要：电力工程造价的影响因素非常复杂,造价管控难度较大,因此有必要应用人工智能技术对造价因素进行分析。灵敏度分析是数据挖掘的一种重要技术方法,论文引入粒子群算法和支持向量机算法建立灵敏度分析模型,对历史工程造价数据进行深入分析,找出影响造价的主要因素,最后根据这些造价主要影响因素建立电力造价估算的数学模型,仿真结果表明可以大大提高电力造价审查的效率。It is very difficult to control power cost,because there are too many impact factors,so it seems to be necessary to introduce artificial intelligent algorithms.The sensitivity analysis is an important method.PSO and SVM were introduced to set up sensitivity analysis model,which can find main impact factors for power cost.At last,these factors were used for power cost estimate model,simulation result showed that such model can improve efficiency of power cost evaluation.

关 键 词：电力造价 灵敏度分析 粒子群算法 支持向量机 

分 类 号：TM73[电气工程—电力系统及自动化]