Load Identification Method And Analysis Platform Of Mine Hoist System Based On Operation Data

The mine hoist is one of the large-scale mine equipment,and plays an important role in improving the transportation of materials and workers in the mine production operation.Once the hoist has an accident during operation,it will not only bring huge economic and property losses to the enterprise,but also seriously affect the life safety of the coal mine workers.To date,a large part of the safety incidents that occurred during the lifting operation were caused by excessive load on the hoist.However,due to the complex production environment of the mine,the load of the mine hoist system is often difficult to measure by direct method.At the same time,with the improvement of the monitoring and management system of modern complex mechanical equipment,A large amount of data is saved to the storage side every day,and the massive historical operation data causes the pressure of the storage system.Therefore,it is very valuable to understand the data of the hoist operating system,analyze and mine the data based on some data of the hoist operating system,find the parameters that can reflect the load characteristics and then identify the hoist load.A new load identification method and prediction system are proposed based on the field data of a mine JKM2.8×6(I)A multi-rope friction mine hoist.Firstly,correlation analysis and PCA(principal component analysis)are used to feature fusion of hoist operating system data to find data reflecting the load characteristics of the hoist;secondly,a variety of neural networks are used to identify the load;finally,the hoist load analysis platform is built based onoperating conditions.The specific research process is as follows:Firstly,the research background of the subject is expounded,and the research status of the load identification method at home and abroad,especially the research status and deficiencies of the hoist load identification method are summarized.The necessity and significance of the research on the load identification method based on the hoist operation data are discussed.The kinematics of the hoist system and the relationship between the various motion parameters of the hoist system are also studied.After establishing and analyzing the dynamic model of the electromechanical transmission system of the multi-rope friction hoist,it is found that the direction of the damping torque changes when the weight is raised and lowered,so it is necessary to separate the lifting and lowering processes of the hoist.Secondly,the field load statistics are carried out,the wire rope tension difference is calculated,and the data is preprocessed by correlation analysis and PCA(principal component analysis)to find the data information that can reflect the load characteristics of the hoist.Three types of neural networks(BP,RBF,ELMAN)were used to build the hoist lifting and lowering load model.In the design of three neural networks,the network function selection and the number of hidden layer nodes are deeply studied and optimized.The identification results show that the error of load recognition by using BP neural network is the smallest for both working conditions.Finally,the BP neural network is used to build a load recognition model based on lifting/lowering conditions.Finally,the B/S-based hoist load identification platform is developed and tested by using ASP.NET technology and database technology.The platform can be visualized and the human-machine interface is friendly.The online query for the multi-rope friction mine hoist operating data and load data is realized,and the load analysis can be realized by calling MATLAB software.