Research On Auxiliary Navigation System Of Combine Harvester Based On Beidou And Vision

Agricultural machinery navigation technology is one of the most important supporting technologies for precision agriculture at home and abroad,and it is also an effective way to realize agricultural intelligentization.At present,agricultural machinery navigation is mainly based on satellite navigation,and high-precision satellite navigation and positioning equipment is used to obtain positioning information so as to realize global path planning.The satellite navigation technology has certain deficiencies in real-time alignment of the rice and wheat harvesting boundary.In addition,the cost of purchasing base station equipment or long-term application of CORS service network is high.Therefore,it is of great practical significance to develop an auxiliary navigation system for combine harvester,which features local and global path planning,low cost,simple structure and easy installment,to improve grain harvesting efficiency,reduce production cost and popularize equipment.This thesis combines the development trend of combine harvesters and the characteristics of operation and studies the auxiliary navigation system through the combination of BeiDou Navigation Satellite System and visual method.The main research content are as follows:First,the overall scheme design and hardware selection of the auxiliary navigation system for combine harvester were carried out.The positioning information of the harvester and the images of the work boundary are acquired by the BeiDou positioning module and two vision sensors respectively,which were displayed on the human-computer interface in real time.The main controller selected the vision sensors according to the path planning information,calculated the lateral distance and the heading angle of the harvester relative to the work boundary through the algorithms.Besides,deflection angle of the harvester was calculated according to the angle information of the angle sensor of the steering wheel so as to steering mechanism to carry out auxiliary navigation.Second,the path planning and boundary prediction methods based on BeiDoupositioning information were studied.According to the information of BeiDou positioning and farmland shape,a navigation path planning model was established to realize the global path planning of navigation operation.Based on the BeiDou Positioning Information,a boundary prediction model based on distance and operation speed was established to switch automatic mode to manual mode before the harvester was reaching the field boundary.Third,this thesis proposed the work boundary recognition algorithm based on dynamic region of interest selection.By comparing the processing effects of different color space,Cb components in YCrCb color space were selected for subsequent image processing to reduce the influence of illumination.Bilateral filtering,Otsu algorithm,morphological operation,Canny edge detection,cumulative probability,Hough transform and other algorithms were used to extract the boundary line.The accuracy of operation boundary detection was improved by twice screening.In addition,the dynamic region of interest was selected twice based on Kalman filter and operation boundary line position prediction respectively,which reduced the average single image processing time from 0.176087 s to 0.064547 s,and can better meet the requirements of the auxiliary navigation system for image processing time.Fourth,A navigation control method for the auxiliary navigation of the combine harvester was designed.The perspective distortion of the image was corrected to obtain the lateral distance and the course angle of the harvester relative to the work boundary.The auxiliary navigation of the combine harvester was realized by using the designed navigation controller based on fuzzy neural network and PID steering controller.The rationality of the design of the navigation and steering controller was verified through simulation tests.Fifth,the thesis carried on the software design of the auxiliary navigation system of the combine harvester.The functions of data receiving and processing,navigation control,interface display and data storage were realized.The performance of the auxiliary navigation system of the combine harvester was verified through field tests.Field test results showed that when the combine harvester harvests rice at the speed of0.4m /s,0.8m /s and 1.2m /s respectively,the maximum lateral deviation was 6.5cm,13.7cm and 19.2cm.Based on the analysis of the development trend and operation characteristics of the combine harvester,this thesis studied the working path planning,working boundary recognition algorithm,hardware selection and software design of the auxiliary navigation system,and designed a auxiliary navigation system of combine harvester based on BeiDou and vision.Finally,the reliability and control accuracy of the auxiliary navigation system were analyzed through field experiments.The test results showed that the maximum lateral deviation is 19.2cm when harvesting at the speed of 1.2m/s,which had good auxiliary navigation performance.It provides the feasibility and technical support for the research of the simple and low-cost automatic navigation system of combine harvester.