At present,there is no unified equipment management method for the equipment status information in many coal enterprises,and the equipment status needs the subjective judgment of people which results in the high cost of coal enterprises and the equipment status can not be accurately judged.In recent years,deep learning has been widely used in big data processing fields such as image rec-ognition,text classification,and data analysis for its high recognition accuracy and fast processing of massive data.Based on this,in re-sponse to the problem of high cost and inability to make accurate judgments in predicting equipment status in coal enterprises,this paper adopted a fully connected neural network method to classify and identify equipment status of coal enterprises,and visualized the accura-cy and loss value of equipment prediction.By comparing the performance with classical algorithms SVM and decision tree models,the fully connected neural network achieved a prediction accuracy of 96.74％for the equipment status of coal enterprises,which was supe-rior to the other two machine learning algorithms,and the training convergence speed of the network model was fast.The application of fully connected neural networks in device state prediction can significantly reduce labor costs in coal enterprises,and has good develop-ment prospects and research value.
deep learningmachine learningfully connected neural networkequipment status prediction
维普
万方数据
