Shanxi Coking Coal Xishan Coal Electricity (Group) Co., Ltd. Mechanical and Electrical Factory

针对液压支架电液控制系统故障主要采用人工进行排查,故障排查过程中随机及个别故障无法排查到位、故障排查效率低下等问题,构建了基于Hadoop的大数据决策分析服务平台,设计了并行算法为核心的数据故障诊断引擎,基于MapReduce对C4.5决策树分类算法进行改进,对电液控制系统硬件设备进行智能化技术改造。测试结果表明:C4.5决策树分类预测引擎提取电磁先导阀、控制器、压力传感器、行程传感器故障特征曲线研究表明,该故障诊断系统可对电液控制系统各类故障进行有效识别和诊断。该系统在官地矿南六采区12605工作面开采中进行应用,在集控室电脑软件上即可对液压控制系统故障进行有效监控,故障诊断可靠性好、智能化程度高,大幅度提高了液压控制系统故障的维修效率。

In view of the issues of artificial troubleshooting is mainly used by hydraulic support electrohydraulic control system faults, random and individual faults being unable to be troubleshooted in place, low faults troubleshooting efficiency during the process of faults troubleshooting and so on. A big data decision analysis service platform based on Hadoop is constructed, and a data fault diagnosis engine with parallel algorithm as the core is designed. The C4.5 decision tree classification algorithm is improved based on MapReduce, and the hardware equipment of the electrohydraulic control system is technology transformed intelligently. The test results show that the C4.5 decision tree classification prediction engine extracts electromagnetic pilot valves, controllers, pressure sensors, and travel sensors. The research of fault characteristic curves shows that the fault diagnosis system can effectively identify and diagnose various faults in the electrohydraulic control system. This system is used during the 12605 working face mining of the southern sixth mining area of Guandi Mine. The hydraulic control system faults can be effectively monitored through computer software in the centralized control room. The fault diagnosis has good reliability and high intelligent degree, greatly improving the maintenance efficiency of hydraulic control system faults.