Methods Of Navigation Control And Path Recognition In The Integrated Guidance System Of Agricultural Vehicle And Implement Based On Machine Vision

Automatic guidance of agricultural machinery is one of the hotspot in precision agriculture field. The research of agricultural machinery navigation based on machine vision has important theoretical significance and practical value. In this paper, the integrated navigation system of agricultural vehicle and implement based on machine vision was researched. The research project was supported by 863 high-tech special projects and the National Key Technology Support Program of China. The main research work is as follows:(1) Design and develop the integrated guidance system of agricultural vehicle and implement.The guidance system of agricultural vehicle based on machine vision was developed, including the steering controller based on PLC and visual navigation software. Moreover, the guidance system of agricultural implement used for maize weeding was designed. In addition, the lateral controller and software of the guidance system were developed. For the purpose of improving the navigation accuracy of agricultural machiney, the integrated guidance system of agricultural vehicle and implement was designed. The experimental results indicated that the integrated guidance systems were stable and reliable.(2) The research of image preprocessing.In order to solve the problems of serious illumination interference for image segmentaion, Cg component of YCrCg color model was selected for subsequent image processing. The fuzzy C-means clustering method based on two-dimensional histogram was used for Cg component segmentation. In order to improve the accuracy of guidance line detection, a filter method based on dynamic image "AND" was proposed to eliminate noise of weeds.(3) Design of crop rows detection methods based on machine vision.Hough transform, Hough transform based on a known point and least squares method are the primary algorithms used to detect crop rows and guidance lines. However, the speed and accuracy of these algorithms need to be improved further. In this paper, three new algorithms used to detect crop rows were designed. These algorithms were crop lines detection method based on linear scanning, crop lines detection method based on improved genetic algorithm and crop rows detection method based on liner correlation coefficient constrained. The method based on liner correlation coefficient constrained was an improved algorithm of least square method. It could effectively improve the noise resistance ability of least square method. Howerver, the time-consuming of this algorithm was more than least square method. In linear scanning algorithm, pixels at the bottom and top edges of the image were selected as two endpoints of the line. Candidate lines were created by moving the position of these endpoints by pixel step. Linear scanning algorithm detected crop rows on binary image and was not sensitive to weed noise. It had high detection accuracy and real-time. In the method based on improved genetic algorithm, two points from image bottom and top side were randomly selected to code as chromosome. By multiple genetic evolutions, the highest fitness individual was chosen as the crop row line. This method had faster speed than the above two methods.The experiment results showed that the real time of the three proposed algorithms were faster than Hough transform, but more accurate than Hough transform base on a known point and least squares method.(4)The research of decision control methods for agricultural vehicle and implementAgricultural vehicle guidance system was a complex system with highly nonlinear, time-varying and large delay, therefore, an adaptive fuzzy controller was designed for path tracking control. By establishing two degrees of freedom steering model and visual preview model, lateral control equations of vehicle were described. Correction factors were introduced into the fuzzy controller and particle swarm algorithm was used to optimize the correction factors. Taking the integral time absolute error (ITAE) sum of lateral offset and heading offset as the objective function, optimal correction factors were calculated by particle swarm algorithm. Simulation and experimental results showed that the designed control algorithm could eliminate the lateral offset rapidly with less overshoot and rapid response. It not only retained the advantages of fuzzy control method but also improved the control quality of guidance system.Fuzzy control method was adopted to control the lateral motion of the agricultural implement. The implement speed and lateral error were used as the input variables of the fuzzy controller. The output variable of fuzzy controller was current signal, which was transmitted to the solenoid valve of the hydraulic system.