【摘要】 随着数字化技术在电力系统中的广泛应用及对电力系统运行可靠性要求的不断提高,不良数据的辨识显得越来越重要。目前广泛应用的状态估计法,存在残差污染等缺点。论文研究了建立在神经网络和聚类分析基础上的GSA不良数据辨识算法,该算法运用神经网络完成对测量数据的预处理,然后由GSA算法对聚类分析后的结果进行判断,完成不良数据的辨识。论文借助M atlab及C语言对GSA算法进行了仿真,通过一个具体的网络不良数据辨识将此算法与状态估计算法进行了比较,验证了该算法的有效性及实用性,有效地避免了不良数据的漏检、误检。更多还原

【Abstract】 With the wide application of numeric technology in power system and increasingly higher requirements of the power system operation reliability,the recognition of the bad data seems more and more important.The state estimating algorithm(be widely used at present) has disadvantage of residual pollution.This paper researches the GSA bad data algorithm based on neural network and clustering analysis.The algorithm accomplishes the pretreatment of surveying data via the neural network,and then judges the result of clustering analysis so as to finish the recognition of the bad data.The paper employs Matlab and C Language to simulate the GSA algorithm and compares the GSA algorithm with the state estimating algorithm via a concrete network,which shows that GSA algorithm can effectively remove residual pollution so as to accomplish the recognition of the bad data.更多还原