基于Lemke优化算法的负荷静态模型参数辨识

在分析传统优化算法不足的基础上,提出了基于二次规划的Lemke优化算法的负荷静态模型参数辨识新方法。详细描述了最小二乘法、拉格朗日乘数法及Lemke 3种算法的原理。通过对电力负荷元件进行静态故障模拟,在系统辨识原理的基础上,运用最优化理论的Lemke辨识算法,对实验故障数据进行负荷建模和参数辨识,并列出了日光灯、电风扇、空调及三者组合的静态特征参数。分析结果表明:由于电器设备的制造工艺不同,不能照搬国外软件提供的参数,而应采用实测参数;基于Lemke的优化算法拟合效果较理想,避免了局部最优,且辨识时间少;基于Lemke的优化算法避免了辨识参数的分散性,参数结果与初值选择无关。

Parameter identification of static load model based on Lemke optimal algorithm

Lemke optimal algorithm is used to identify parameters of static load model to improve the performance of traditional optimization algorithms.Principles of the least square method,Lagrange method and Lemke method are described.Static faults of electric power load elements are simulated.With fault data,the load modeling and parameter identification are performed based on Lemke optimal algorithm.Static characteristic parameters of the lamp,the fanner,the air conditioner and their combination are listed.The result analysis shows that:adopt measured parameters and never copy those supplied by foreign software because manufacturing techniques are different;fitting results by Lemke optimization algorithm are better,avoiding local optimum and costing less time;Lemke optimization algorithm avoids decentralization of identification parameters and makes identified parameters independent of initial values.