Research On Data-Driven Fault Diagnosis And Early Warning Technology Of Mine Hoisting Equipment

The mine hoist is the core mechanical equipment in the production of mineral resources and its main function is to transport personal and equipment.The safety of its operation is directly related to the normal operation of the mine and the safety of the life and property of the operators.Therefore,the research on the early warning method for the key systems and parts of the hoist is of great significance to the safety of coal mine production.In the field of elevator fault early warning,traditional early warning methods mainly rely on human subjectivity,both in terms of accuracy and timeliness,are limited by the experience level of maintenance personnel and some other human factors.Therefore,it is difficult to obtain stable guarantees in terms of precision and accuracy.In addition,most of the existing hoist fault warnings focus on the monitoring of hoist parameters and the extraction of signal features.After a fault occurs,time is invested to analyze the cause of the fault,which greatly reduces the working efficiency of the equipment.In order to fully fit the uniqueness of the characteristic parameters of the hoist under different operating environments and improve the accuracy of fault early warning,this paper takes the multi-rope friction mine hoist of a mine as the research object,and carries out a data-driven research on the fault early warning technology of mine hoisting equipment.The purpose of this paper is to use the monitoring data of mine hoist operating conditions to construct an effective mine hoist fault early warning model by analyzing the key operating parameters of the hoist,so as to provide timely,accurate and effective early warning for the potential hidden troubles of the hoist.This paper firstly analyzes and studies the safe operation requirements,structural composition and failure mechanism of the multi-rope friction type mine hoist,and determines the relationship between the common failure manifestations and failure types.On this basis,this paper establishes the acquisition scheme of the hoist parameters,and completes the arrangement of the sensors and the construction of the monitoring system.Finally,after analyzing and comparing the advantages of the algorithms,a fault early warning method for mine hoists based on LSTM(Long Short-Term Memory)is studied.The method based on the LSTM-based fault prediction model predicts the change trend of the hoist parameters,and uses the sliding proportional mean to analyze the residual of the prediction result.Finally,the residual analysis result is used as the early warning standard,the early warning threshold is determined,and a complete fault early warning method is established.(1)A hoist data prediction model based on deep learning is built to provide data basis for hoist faults.Through the comparison and selection of the advantages and disadvantages of various algorithms and the prediction accuracy,this paper selects the long-term and short-term memory network algorithm with the best prediction effect,and establishes a prediction model on this basis,and selects the Adam optimizer for parameter optimization.The current popular Python programming language is used to implement the hoist prediction model.It has been verified that the model can well grasp and predict the changing trend of hoist parameters.(2)A fault early warning method based on sliding proportional mean is proposed.In order to reduce the influence of uncertain factors such as the environment on the early warning effect,this paper uses the sliding window proportional mean to process the residual,and uses the processed residual as the early warning basis.On this basis,a complete early warning method process is established and tested.(3)A mine hoist fault early warning platform has been established.This paper took the Ruoyi development system as the basic framework,integrated the hoist data early warning model,and developed a mine hoist early warning platform including functions such as fault early warning,equipment monitoring,in-use equipment management and related industry standard management.After experimental verification,it is proved that the big data-driven mine hoist fault early warning method proposed in this paper is effective,which can effectively grasp the changing trend of hoist parameters and respond to faults.At the same time,the early warning platform pushes the early warning model to practicality and can be applied to actual production,thereby ensuring the safety of coal mine production.