Research On Grain Flow Monitoring Device Of Combine Harvester Based On Photoelectric Array

The grain flow monitoring device of combine harvester can measure the grain yield of the unit plot,which is the main link of precision agriculture.Owing to the high price of foreign combine harvesters,it is not conducive to the widespread introduction of China,and there is no domestic grain flow sensor in China.The domestic impulse grain flow monitoring scheme has the disadvantages that the sensor is not hard to be damaged and cannot fully monitor the grain flow.In the past,photoelectric grain flow monitoring scheme mainly designed the photoelectric sensor with a single optical path,which cannot fully characterize the grain accumulation on the scraper.In view of the above problems,this thesis designs a grain flow monitoring device of combine harvester based on photoelectric array,which is mainly exerted to the combine harvester of scraper grain conveying device.Using BP neural network technology and embedded technology,a grain flow monitoring system is developed,which effectively reduces the measurement error caused by the uneven distribution of grain thickness,and realizes real-time and stable online monitoring of the grain flow of the scraper grain conveying device at a lower cost.The main work of this thesis contains the following aspects:(1)The overall function designed and hardware selection of grain flow monitoring device based on photoelectric array.Development status of existing grain flow monitoring technology was analyzed,and the research route of grain flow monitoring device of combine harvester based on photoelectric array was defined.The working principle of measuring grain volume by photoelectric radiation grain flow monitoring device was analyzed,and the arrangement mode of photoelectric radiation sensor array and the grain flow monitoring scheme with STM32 single chip microcomputer as the core controller were determined.Depending on the design requirements of grain flow monitoring device,the hardware designed and selection was completed.(2)Establishment of grain accumulation model in scraper grain conveying device and research on grain flow calculation model based on BP neural network.Using the simulation software to simulate the accumulation form of grain on the scraper of elevator,the grain flow calculation method of uniform volume model was analyzed,and the linear calculation model was used to test the rice flow on the test bench.The relative error between the theoretical calculation and practice was ≤ 8.30%.In order to decrease the measurement error caused by uneven grain distribution on the scraper,a grain flow calculation model based on BP neural network was created by using the nonlinear fitting and generalization ability of BP neural network.(3)Software designed for grain flow monitoring device.The program design requirements for grain flow monitoring device were analyzed.The software was designed based on the modular idea,including the design of main program,the program design of hall velocity measurement module,the program design of grain thickness acquisition,the program designed of grain flow calculation model transplantation based on BP neural network,the program designed of grain volume and grain flow calculation,the program designed of CAN communication program and human-computer interaction module.(4)The experimental verification of grain flow monitoring device mainly includes indoor test and field harvest test.Through the speed test platform,the accuracy of the speed sensor and the speed measurement program was tested,and the relative error of speed measurement was ≤ 0.8%;In the thickness measurement tested at no-load scraper,the measured relative error of scraper thickness was ≤ 2.11%.Finally,grain flow monitoring test was performed on the grain flow test-bed.The results show that the relative error of grain flow monitoring of rice grain was less than or equal to 5.88%.Monitoring accuracy of the grain flow calculation model established by BP neural network was 2.42% higher than that of the linear calculation model of grain flow.Based on the above research,a grain flow sampling box was designed and its measurement law was calibrated on the bench.Using the sampling box for field experiment,the monitoring effect was in line with the experimental expectation.It indirectly tests the effectiveness of the grain flow monitoring device designed in this thesis and the grain flow monitoring method optimized based on BP neural network.The research shows that the grain flow monitoring device of combine harvester based on photoelectric array studied in this thesis has a stable operation,convenient installation,economy and reliability,and provides technical support for the popularization of photoelectric grain flow monitoring technology in China.