Blockage Fault Diagnosis And Alarm System For Combine Harvester Based On BPNN And DS Theory

In recent years, along with our country supporting for agriculture and promoting the deep integration of digital and intelligent technology with agricultural equipment, the development of combine harvester and related equipment is ushering in a new opportunity. Currently domestic combine harvester is small and medium-size, low degree of automation, lack of intelligent monitoring equipment, and during grain harvesting operations the frequent blockage failt seriously affect work efficiency and quality. In order to reduce the failure rate of blockage, improve work efficiency and driving operating environment, a system should be designed, which has a higher level intelligence and can mornitor blockage fault.This thesis combines BP neural network(BPNN) with DS evidence theory as combine harvester blockage fault diagnosis algorithm. The working condition of combine harvester can be divided into four categories:Normal, Slight blockage, Blockage, Serious blockage. Set these four kinds of working conditions as the identification framework of DS evidence theory. Analyzing the speed information of harvester rotating components through the BP neural network and assigning a value to basic probability of each proposition within the identification framework. Finally, fusing the information at different times to get the diagnosis results by using Dempster synthesis rules.Based on the intelligent blockage fault analysis method, the system of blockage fault diagnosis and alarm for combine harvester is designed. Through analyzing the various components working step and blockage fault of combine harvester, the system selects cutting table auger, conveying channel, threshing cylinder and transport grain auger as the mornitoring objects, and the forward speed as the controlling object. Using C8051F020 as the system controller and touch screen as man-machine interactive system, the system presents the user with speed information of each rotating member, forward speed and the alarm information of combine harvester work status, and store data in the CF card for offline analysis. Design a automatic control system for forward speed, which adjust forward speed in real time according to combine harvester blockage fault diagnosis.The system achieves the purposes of optimizing working state real-timely and reducing the failure rate of blcockage for combine harvester ultimately by adjusting the forward speed to regulate the amount of feed grain.Using MATLAB to simulate the blockage fault diagnosis algorithm, the test verify the feasibility of the algorithm. Similarly, the system was debugged in laboratory and tested in field. Based on the manual control mode of forward speed, field test further validated the accuracy of blockage fault diagnosis and feasibility of the algorithm,and the system achieves “Uncertain, Slight blockage, Blockage, Serious blockage” four pre-alarm and the “Uncertain, Slight blockage” warning time is up to 2 seconds or more, which provides valuable time for the driver to take measures to exclude blockage fault. Based on the automatic control mode of forward speed, field test validated that the system is able to take actions timely for blockage fault warning and to optimize combine harvester work state continuously to maintain efficient working state.