组合优化的灰色模型在异常电能表查找中的应用

Combinational optimization of gray model in the application of searching abnormal watt-hour meter

电力计量装置是供电企业正常作业的基本条件,是电力系统中不可缺少的组成部分,其计量准确与否直接关系到用户和电力企业的利益。若电力计量装置出现异常,尤其是负误差超差,将直接导致供电公司利益受损,严重的可能导致电力系统无法正常运行。文章利用组合优化后的灰色模型建立用户负荷预测模型:首先利用GM(1,1)建立用户负荷预测模型,然后利用马尔科夫链对GM(1,1)的预测结果进行优化,接着利用蚁群算法对优化后的灰色模型进行再度优化,建立组合优化的灰色模型。最后建立负荷阈值模型,将台区所测得用户负荷值与负荷阈值进行对比,若该用户负荷值在负荷阈值范围内,则认为电能表正常;反之则认为电能表异常,列为"潜在异常对象";实例证明组合优化的灰色模型在异常电能表查找中具有良好的工程实用性和有效性。

Electric power metering device is the basic condition of power supply enterprise normal operation, and an integral part of power system. Its measurement is accurate or not directly related to the interests of the users and the electric power enterprise. If abnormal electric power metering device, especially the abnormal negative error, will directly damage the power supply company interests, it seriously can lead to the abnormal operation of the power system. This article combined the optimized gray model is used to establish the user load forecasting model： firstly, the user load forecasting is established based on GM （1,1） model, and then using the Markov Chain of GM（1,1） prediction results is optimized, and then using Ant Colony Optimization to optimize gray model of optimization, combinational optimization of the gray model is established. At last, the load threshold model is set up, comparing the user load value with the load threshold value measured by the area. If the user load value is in the range of the load threshold, watt-hour meter is normal. On the contrary, watt-hour meter is unusual, regarded as a “potential exception object”. Examples prove that it is practicable and valid to use combinational optimization of gray model in the abnormal watt-hour meter to find good engineering.