| 引用本文: | 陈碧云,丁晋,陈绍南.基于关联规则挖掘的电力生产安全事故事件关键诱因筛选[J].电力自动化设备,2018,(4): |
| CHEN Biyun,DING Jin,CHEN Shaonan.Selection of key incentives for power production safety accidents based on association rule mining[J].Electric Power Automation Equipment,2018,(4): |
|
| 摘要: |
| 随着智能电网、通信网络及电力生产安全事故事件分析水平的提高和发展，电力生产安全事故事件数据量快速增长、复杂性不断增大，逐步构成了电力生产安全事故事件大数据。为在先验事故事件大数据的基础上高效、可靠地对事故诱因进行分类和识别，基于关联规则挖掘进行电力生产安全事故事件关键诱因筛选。根据事故事件的特点，建立电力生产安全事故诱因分析体系，对不同类型的事故进行布尔离散化，并基于关联规则挖掘提出事故诱因的诱发度计算方法，运用Apriori算法进行深度关联规则挖掘，并根据强关联规则对关键诱因进行筛选和分析。以某区域近5年的事故实例分析验证了该方法的有效性。 |
| 关键词: 电力生产安全事故事件 诱发度 关联规则 Apriori算法 |
| DOI：10.16081/j.issn.1006-6047.2018.04.010 |
| 分类号:TM732 |
| 基金项目:国家自然科学基金资助项目(51767002) |
|
| Selection of key incentives for power production safety accidents based on association rule mining |
|
CHEN Biyun1, DING Jin1, CHEN Shaonan2
|
|
1.Guangxi key Laboratory of Power System Optimization and Energy Technology, Guangxi University, Nanning 530004, China;2.Guangxi Power Grid Electric Power Research Institute, Nanning 530023, China
|
| Abstract: |
| With the development of smart grid, communication network and power production safety accident analysis level, the data of power production safety accident is growing exponentially with more complexity, which gradually forms the big data of accidents in electric power production. In order to classify and identify the causes of accidents effectively and reliably on the basis of a large amount of prior accident data, the association rule mining is adopted to select the key incentives of power production safety accident. The incentive analysis system of power production safety accident is established according to the characteristics of accidents. Boolean discretization is carried out to different type of accidents, the calculation method for incentive degree of accident causes is proposed based on association rule mining, the Apriori algorithm is adopted for deep association rules mining, and the key causes are selected and analyzed based on the strong association rules. The validity of the proposed method is verified by the analysis of 5-year regional accident cases. |
| Key words: power production safety accidents incentive degree association rules Apriori algorithm |
