基于多维关联规则的电网脆弱性识别研究  

RESEARCH ON IDENTIFYING VULNERABILITY OF POWER GRID BASED ON MULTIDIMENSIONAL ASSOCIATION RULES

作　　者：颜宏文[1] 谢启龙[1] 

机构地区：[1]长沙理工大学计算机与通信工程学院,湖南长沙410114

出　　处：《计算机应用与软件》2015年第11期36-40,共5页Computer Applications and Software

基　　金：国家自然科学基金项目(51277015)

摘　　要：随着电网线路故障多样性、多重性、不确定性等因素的积累导致大面积停电事故时有发生,能否挖掘出潜在的线路隐患并制定相关的应对措施,对政府部门和电力企业进行决策起着重要作用。针对传统关联规则挖掘表示形式单一、多维度展现的不足、效率不高的缺点,介绍一种有效地将数据立方体技术和FP-Growth算法相结合的线路故障快速预警方法,提出一种从多维角度分析故障的模式,通过可视化输出的判别规则来为电力系统故障预测和预警提供可靠的决策依据。最后以某省电网线路故障数据为例,验证了方法的有效性和实用性。With the accumulation of diversity, multiplicity and uncertainty of line malfunction in power grid, occur frequently. Whether or not the potential lines hazards can be dug out and the correlated large-scale power outages can be developed play an important role in decision-making by the governmental departments and electric power corporations. Since traditional association rules mining has single representation form and the multi-dimensional show is inadequate and inefficient, for those shortcomings, this paper introduces a fast early warning method for line malfunctions, which effectively combines the ＂data cube＂ technology with FP-Growth algorithm, and puts forward a pattern which analyses malfunctions from multidimensional perspective. It provides reliable decision-making basis for malfunctions forecast and early warning of power systems through discriminant rule of visualised outputs. Finally we verify the effectiveness and practicality of the method by taking the line malfunction of power grid in a certain province as the example.

关 键 词：数据挖掘 多维关联规则 数据立方体FP—Growth 故障预测 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]