Research On Fault Diagnosis Of Large-scale Crushing Equipment Based On Genetic-BP Neural Network

With the improvement of production technology of open-pit mine enterprises,large-scale mining equipment is playing an increasingly important role in mine production.The management of mining equipment has become an important part of the production management of mining enterprises.Among them,crushing is an important part of mining production in open-pit mining enterprises.The gyratory crusher is an important production equipment for the crushing production of metal open-pit mines.Therefore,the advance analysis and diagnosis and early warning of the gyratory crusher is to improve the reliability of crushing production management.The important guarantee for enterprises to reduce costs and increase efficiency.At present,the failure diagnosis of the gyratory crusher is mainly based on the personnel experience,but the problem of the gyratory crusher is complicated.It is difficult to achieve effective fault diagnosis and early warning only by subjective experience.Therefore,this paper introduces the wireless sensor technology,BP neural network diagnosis method and genetic algorithm to establish a gyratory crusher fault diagnosis and early warning system,realizing the real-time fault diagnosis and early warning of the gyratory crusher.Completed the following major research content:(1)Raw data collection and analysis of a gyratory crusher.Analyze the daily maintenance records of the gyratory crusher and the domestic and foreign literatures,and obtain the failure type of the gyratory crusher and the fault characteristic parameters that need to be monitored.Use the wireless sensor technology to obtain the original data of the required gyratory crusher and according to the BP nerves.The network characteristics of the network and the operation status of the gyratory crusher,the principle of data selection are formulated to realize data selection,and the fault characteristic parameter matrix is established,which provides a reasonable data foundation for the establishment of the gyratory crusher fault diagnosis model.(2)The basic model of fault diagnosis for gyratory crusher based on BP neural network.The basic principle of BP neural network fault diagnosis is researched,and the specific application of BP neural network in the fault diagnosis of the gyratory crusher is proposed.That is,fault characteristic parameters are taken as the network input,fault types are taken as expected network output,and corresponding training parameters are set up.Fault diagnosis network,and compare the training results achieved by different network training methods and the number of neurons in the hidden layer,determine the optimal network training method and network structure,and build a basic model of the failure diagnosis of a gyratory crusher based on BP neural network.(3)Fault diagnosis optimization model of gyratory crusher based on genetic-BP neural network.Study the basic principles of BP neural network and genetic algorithm,and determine the specific method of genetic algorithm optimization BP neural network,that is,use genetic algorithm to evolve the weight threshold of BP neural network to achieve the optimization purpose,first according to the operation flow of genetic algorithm right The value threshold is evolved to obtain the optimal solution of the weight threshold,and then the optimal solution is used as the initial weight threshold of the BP neural network.The genetic-BP neural network cycle is completed according to the above-mentioned BP neural network cycle crusher fault diagnosis basic model.Construction of fault diagnosis and optimization model for crusher.(4)Application verification and testing of the fault diagnosis model of the gyratory crusher.Taking a Luoyang mine gyratory crusher as an example,according to the BP neural network fault diagnosis basic model establishment process and genetic-BP neural network fault diagnosis and optimization model establishment process,the training results and test results of the two are obtained and compared.The results show that the result of fault diagnosis optimization model of the gyratory crusher based on genetic-BP neural network is obviously better,so the genetic-BP neural network is finally determined as the fault diagnosis method of the gyratory crusher and the cycle of the genetic-BP neural network is used.The fault diagnosis model of the crusher is used as the final fault diagnosis model for the gyratory crusher.The fault diagnosis accuracy of the built model can reach 99.74%.(5)Real-time fault diagnosis and early warning system based on the gyratory crusher fault diagnosis model.Based on the accurate identification of the failure mode of the gyratory crusher fault diagnosis model,the pre-warning of the gyratory crusher was further studied,and it was concluded that the gyratory crusher failure warning could be completed by the failure type diagnosis decision and component diagnosis and decision.On the basis of this,the fault diagnosis model of the optimized gyratory crusher is used as the diagnostic method,and the original data of the gyratory crusher collected by the sensor is used as the diagnostic basis.The fault diagnosis and early warning system of the gyratory crusher is established and finally implemented.The real-time fault early warning of the gyratory crusher.This paper combines the BP neural network,genetic algorithm and the problem of fault diagnosis in the real-time gyratory crusher,scientifically and reasonably proposes the fault diagnosis model and the fault diagnosis and early warning system of the gyratory crusher,and solves the gyratory crusher in the mine production management.The problem that faults are difficult to be diagnosed and diagnosed in real time provides an effective method for fault diagnosis and early warning of the gyratory crusher,which ensures the stability of mine production,effectively reduces the cost of equipment management for open-pit mine enterprises,and improves the open-pit mine enterprise.The level of production organization and management.