Research On Remote Fault Diagnosis System Of Combine Harvester Based On Fuzzy Neural Network

Combine harvesters,the main machinery of grain crops,which have a high rate get malfunction after working for a certain time.However,in the domestic small and medium-sized mainstream combine harvesters,the remote fault diagnosis system is far from entering the practical stage.On this basis,the research and application of the remote fault diagnosis system of combine harvesters become more important to guarantee the harvest of crops timely and maintain the working machinery precisely.For the common faults that occur when combine harvesters are working,monitoring real-time status data of key work components,combine harvesters remote fault diagnosis system processes the monitoring data remotely to get potential fault information or real-time fault information,thus the system realizes warning or alarm function for faults remotely.The practicability and the accuracy of the diagnosis result also should be considered for the remote fault diagnosis system.It reflects practicability that the system diagnoses in real time without affecting the harvesting,and accuracy is reflected in the high accuracy of the system diagnostic results.For this need,a remote fault diagnosis system was established for the combine harvester with the fuzzy neural network algorithm.According to current status of domestic and foreign combine harvester diagnostic systems,the problems of our country in this field were analyzed.In this case,the research content and method of this topic were proposed firstly.Combine harvester onboard data monitoring module was built based on LabVIEW,then real-time working data of key components of the combine harvester were collected through the sensor,and the data were saved on local and on the server.The monitoring module alarmed through thresholding in location.Remote data was used to further data mining.Fuzzy neural network algorithm was implemented on the server.Fuzzy neural network algorithm was trained with training set and verified with testing set in turn.The results showed that the diagnostic algorithm can achieve an accuracy of more than 80%.Secondly,the system combined the monitoring module with the fault diagnosis module,by the GPRS remote communication module.Finally,a field experiment was conducted,and experimental results was analyzed.The results showed that the system could diagnose the blockage fault of the combine harvester remotely and accurately.