Research On Early Warning System Of Five-axis Cnc Machine Failure Based On Fuzzy Reasoning

To implement Man-Machine interaction and equipment dialog in future industrial 4.0 requires the equipment with intelligence. Research using intelligent method for prediction of equipment state becomes active. As the foundation of the national equipment, CNC machine intellectualization research is increasingly important. In this thesis, To process knowledge acquisition for real-time data of five-axis CNC machine through data mining, To predict the probability of system failure of five-axis CNC machine by reasoning method based on fuzzy Petri net for the purpose of fault warning. Focusing on the five-axis CNC machine’s real-time data modeling and model evaluating, through data mining knowledge acquisition, and fault reasoning based on fuzzy Petri nets, etc. The contents are divided into four parts as follows:1. Data modeling research which based on five-axis CNC machine time series model real-time status. Five-axis CNC machine in real-time status data has the features of massive data, this thesis presents a multi-dimensional time-series model, on this basis,equipment state model, device status prediction model and equipment knowledge acquisition model has established, and further confirm the definition of model measurement.2. Research of five-axis CNC machine fault knowledge acquisition method based on data mining. Based on equipment knowledge acquisition model and its measurement, using similarity prediction algorithm and clustering analysis of K-Means algorithm to process real-time data mining by analyzing the data in real-time status of five-axis CNC machine, thus to acquire the knowledge of real-time machine data, and predict the status of machine. This laid a solid foundation of warning machine fault by knowledge based inference. Finally, verified the effectiveness of the model and algorithm through the analysis of the experimental data in real-time of CNC machine, and put forward the improvement methods.3. Research of five-axis CNC machine fault knowledge reasoning based on Adapt Fuzzy Petri Net(AFPN), After the prediction of the data status of equipment, presenting the experience knowledge and acquired fault prediction knowledge in forms of APPN through fuzzy production regulations, and obtaining failure prediction results by AFPN recursive formula. The last, a numerical example are used to verify the correctness of AFPN reasoning algorithm.4. Developed a five-axis CNC machine failure warning system based on Microsoft DotNet platform, achieved the core function of basic data and parameters configuration management and the function of fault reasoning by data mining.To proved the correctness of warning system by application testing for different data.