Research On Speed Control System Of Combine Harvester Based On Feed Quantity Forecast

With the rapid development of cloud computing,Internet of Things and other technologies,China’s agricultural production has gradually entered the stage of smart agriculture.Traditional combine harvester operation in the field completely depends on the operator’s experience.During the busy harvest season,the complex field conditions and high intensity and long time operation will make drivers prone to fatigue and improper speed control,which will lead to the abnormal feeding amount and the blockage of the drum,greatly reducing the harvest efficiency,or even missing the best harvest time.The existing operating speed control device mainly uses the roller speed and other events that have occurred to establish the control model,which has a high delay and lag problem.Therefore,in order to improve the real-time performance,reduce the failure rate and improve the harvest efficiency.This paper designs a combined harvester operation speed control system based on feeding quantity prediction.Combined with the working characteristics of combine harvester,the feeding quantity forecasting device and operation speed control system are developed by using machine vision and embedded technology.The main work is as follows:(1)The overall scheme design and hardware selection of automatic control system of operation speed of combine harvester.Speed control and feed volume forecast overall scheme design and hardware selection.Firstly,the operation process of the combine harvester is briefly analyzed and the importance of feeding quantity in the whole harvest process is expounded.It will affect the rotation speed of the key components and the performance index of the harvested crops.Then,an automatic control scheme of operation speed based on feed volume prediction based on embedded platform is determined.Hardware selection is carried out according to the overall scheme and requirements of speed control,mainly including the selection of main controller,Hall sensor,Beidou speed measuring module,human-computer interaction module and circuit design.The hardware selection of the feed volume prediction device includes image acquisition module,image processing module and communication module.Through the system,the next feeding amount CAN be predicted,and the value CAN be transmitted to the main controller through CAN module.The main controller calculates the optimal target speed according to this value,key component speed and harvest performance parameters.Finally,the output voltage value of the D/A module is used to control the solenoid valve and then control the operation speed of the combine harvester.The relevant parameters are displayed on the serial port screen.(2)Establishment of feeding quantity prediction model based on machine vision.Firstly,theoretical analysis and research were carried out on the mold of rice in the laboratory.The machine vision technology was used to extract the pixel value of the ear layer of the crops in front and the feeding density was predicted by combining with the optimized BP neural network prediction model,so as to get the feeding amount at the next moment.The process of image processing mainly includes area of interest extraction,color space conversion,filtering and threshold segmentation,etc.The average processing time of each image is about 0.8s.Then the pixel value of ear layer,stubble height,water content and grass to grain ratio were taken as the input of neural network,and the total weight of rice was taken as the output to establish an optimized BP neural network prediction model,and simulation training was carried out through MATLAB.The results show that the R of the whole data set is 0.98823,and the fitting effect is good.(3)Research on Automatic Control Strategy of Combine Harvester Operation Speed.First,the real-time operation speed of the combine harvester was analyzed by the Beidou module,and the extended Kalman algorithm was used to improve the measurement accuracy.Secondly,the optimal target velocity is calculated according to the predicted feeding amount,the rotation speed of the key components and the harvest performance parameters(including impurity,crushing and loss rate).In addition,a fuzzy PID speed controller is designed.The deviation of speed and the rate of deviation change are taken as input values,the voltage value is taken as output,and the speed obtained by the sensor is taken as feedback information.Realize the automatic control of operation speed of combine harvester.The results show that the proposed control algorithm has strong robustness and real-time performance.(4)Operation speed control and feed volume prediction system software design.Firstly,the main program is designed and the development environment and interface of the main controller are introduced.Secondly,the underlying driver of the main control unit is designed,including Beidou speed measurement,Hall speed measurement,CAN module communication,automatic speed control and humancomputer interaction software.Then configure the environment of the feed volume prediction module and the installation of related library files.Write the program of image acquisition,image processing and feed volume prediction.Through CAN communication to the master controller.Finally,the main controller realizes the functions of data acquisition and processing,speed automatic control,interface display and data storage.(5)Verification of the performance of the operating speed automatic control system based on feeding quantity prediction in laboratory and field tests.First of all,the basic function test of the whole system is carried out indoors to ensure the normal operation of each module.Secondly,the site calibration test was carried out.Understand the accuracy of each sensor.Then field trials were carried out.This includes the verification of the accuracy of the feed volume prediction module and the harvest of the combine under manual and automatic modes.Record the relevant test data and compare and analyze the two control methods.The results show that the accuracy of the device can reach 94%.Compared with manual control,the harvest efficiency of the combined harvester operating speed control system based on feeding quantity prediction is increased by about 9.9%,which basically achieves the desired effect.It provides a technical support for the intelligence of the combine harvester.