动车组运维效率关联规则挖掘优化算法
发布时间：2018-08-22 18:21
【摘要】：随着动车组运营时间和运营里程的增长,动车组运维系统积累了大量的数据.利用高效的关联规则挖掘算法从动车组运维数据中快速发现有用的信息,对于提高动车组关键部件运维效率具有重要意义.针对动车组运维数据的数据量巨大、价值密度低的特点,设计一种基于近似最小完美Hash函数的AMPHP(approximate minimum perfect hashing and pruning)算法,相较于传统的直接Hash和修剪(direct hashing and pruning,DHP)算法,它可以过滤掉所有的非频繁项集,无需额外的数据库扫描.为了突破单机算法的性能限制,借鉴SON算法思想对AMPHP算法进行并行化改进,提出AMPHPSON算法,进一步提高算法性能.使用实际的动车组牵引电机运维数据进行测试分析,实验结果表明,AMPHP-SON算法具有很好的时间性能,且挖掘出的规则可以有效地指导动车组修程修制优化,从而达到提高动车组运维效率的目的.
[Abstract]:With the increase of EMU operating time and mileage, EMU operation and maintenance system has accumulated a lot of data. Using efficient association rule mining algorithm to quickly find useful information from EMU operation and maintenance data is of great significance to improve the efficiency of EMU key components. In view of the large amount of data and low value density of EMU operation and maintenance data, a AMPHP (approximate minimum perfect hashing and pruning) algorithm based on approximate minimum perfect Hash function is designed, which is compared with the traditional direct Hash and pruning (direct hashing and pruning DHP algorithms. It filters out all infrequent itemsets without additional database scans. In order to break through the performance limitation of single machine algorithm, using the idea of SON algorithm to improve the parallelization of AMPHP algorithm, a AMPHPSON algorithm is proposed to further improve the performance of the algorithm. The test results show that the AMPHP-son algorithm has good time performance, and the rules can effectively guide the EMU maintenance system optimization. In order to improve the efficiency of EMU operation and maintenance.
【作者单位】： 北京交通大学计算机与信息技术学院;高速铁路网络管理教育部工程研究中心(北京交通大学);
【基金】：国家“八六三”高技术研究发展计划基金项目(2015AA043701)~~
【分类号】：TP311.13;U269
，
本文编号：2197923
[Abstract]:With the increase of EMU operating time and mileage, EMU operation and maintenance system has accumulated a lot of data. Using efficient association rule mining algorithm to quickly find useful information from EMU operation and maintenance data is of great significance to improve the efficiency of EMU key components. In view of the large amount of data and low value density of EMU operation and maintenance data, a AMPHP (approximate minimum perfect hashing and pruning) algorithm based on approximate minimum perfect Hash function is designed, which is compared with the traditional direct Hash and pruning (direct hashing and pruning DHP algorithms. It filters out all infrequent itemsets without additional database scans. In order to break through the performance limitation of single machine algorithm, using the idea of SON algorithm to improve the parallelization of AMPHP algorithm, a AMPHPSON algorithm is proposed to further improve the performance of the algorithm. The test results show that the AMPHP-son algorithm has good time performance, and the rules can effectively guide the EMU maintenance system optimization. In order to improve the efficiency of EMU operation and maintenance.
【作者单位】： 北京交通大学计算机与信息技术学院;高速铁路网络管理教育部工程研究中心(北京交通大学);
【基金】：国家“八六三”高技术研究发展计划基金项目(2015AA043701)~~
【分类号】：TP311.13;U269
，
本文编号：2197923
本文链接：https://www.wllwen.com/kejilunwen/ruanjiangongchenglunwen/2197923.html
