Research On Equipment Operation State Prediction Method Based On Data Mining

With the continuous promotion of coal technology,The safe and reliable operation of coal mine electromechanical equipment has been the central issue?Because the law of the operating state of electromechanical equipment is often a"metaphor"in a large amount of complex data,Coal mining enterprises need to thoroughly study the intelligent representation of the state information of coal mine electromechanical equipment,and extract hidden valuable information to express the historical operation status of electromechanical equipment.According to these status,the future of the coal mine electrical and mechanical equipment running status can be predicated and evaluated properly,the corresponding preventive measures,which has became the basic model of health maintenance for mechanical and electrical equipment will be applied.Therefore,this paper deeply studies the prediction method of coal mine electromechanical equipment operation state based on data mining.Firstly,for low prediction accuracy and poor applicability of single forecasting model for coal mine electromechanical equipment running state,it is ainalyzes that the structure principle of ARIMA prediction model,gray GM(1,1)prediction model and BP neural network prediction model in the paper,and proposecd the AGB combined prediction model method.The combined prediction model is verified by adjusting each single model weight parameter.Secondly,aiming at the problems of large amount low utilization rate and slow speed of mass data mining by single machine of data in operating state of electromechanical equipment of coal mine,a double MapReduce minling prediction framework is established by using MapReduce technology,establish a dual-MapReduce operation state data mining prediction gmodel.Among them,MapReducel is responsible for feature extraction of monitoring data.MapReduce2 is responsible for predictie analysis of feature data,and performs the Map process and Reduce implementation process.Thirdly,It is hard to quantify the health status of mechanical and electrical equipment in coal mine objectively,the fault prone parts of mechanical and electrical equipment in coal mine are deeply studied.Establish a health status evaluation system,using principal component analysis to determine each exvaluation indicatorpredict the parameters of the shearer evaluation index through the AGB combined forecasting model;introduce the concept of deterioration degree of coal mine electromechanical equipment,construct a multi-index deterioration degree evaluation matrix,and use AHP to determine the weight of components and iindicators.Establish a health status assessment model for coal mine electromechanical equipment.Finally,build a Hadoop platform to verify the efficiency of the data mining framework and study the relationship between Hadoop cluster nodes and parallel processing speed...