| Intelligent fault diagnosis system is necessary for the automatic upgrading of the wooden door production line.It is an effective measure to improve the quality and efficiency of fault diagnosis and ensure its reliable operation.This dissertation takes wood door production line as the research object and develop a troubleshooting expert system suitable for the current stage of wood door production line by combing the Internet of Things technology.Firstly,starting from the current status of the production line and existing problems,the overall framework of the troubleshooting expert system is designed based on combing the manufacturing process and site investigation of the wooden door production line.Secondly,in view of the difficulty of collecting the wooden door production line data,the data acquisition system of the scene is developed to realize the integration of multi-source heterogeneous production data,and in order to improve the effectiveness and generalization ability of the fault diagnosis expert system,the fault diagnosis algorithm based on improving the convolutional neural network is designed according to the K-fold cross validation and the self-organization mapping algorithm.Meanwhile,the principal component analysis is integrated to reduce the input data dimension and improve the speed of model training,then,the expert system knowledge database and inference engine are designed,the quantitative analysis of fault events is completed based on the fault analysis method,the extracted minimum fault cut set and reliability are stored in the knowledge database as knowledge rules,and the inference engine is designed by fusing forward reasoning with the designed improved fault diagnosis algorithm.Finally,the expert system for fault diagnosis of wood production line is developed by the softwares such as Qt Creator and Redis,and the correctness and feasibility of the system designed in this dissertation are verified through the corresponding fault examples.The system of fault diagnosis experts of wood door production line designed in this dissertation realizes real-time production and manufacturing process status monitoring,fault diagnosis and maintenance of key equipment of production line,providing effective solutions to realize the informationization of wood door production line process,ensuring safe and reliable production in the workshop,completing the delivery task on time,and improving the production efficiency of enterprises. |