Research On Navigation And Control Technology Of Grain Leveling Robot Based On Vision

Grain storage is related to the national economy and people's livelihood and national economic security.As an important war preparation material in our country,the efficient and scientific storage of grain reserves is particularly important.Grain surface leveling is an important part of grain storage.A flat grain surface is conducive to the ventilation and heat dissipation of the grain,avoiding the residence of rats and finches,etc.,so as to prevent the grain from mildew during the storage period.At present,grain leveling mainly relies on labor,which has low efficiency and unsatisfactory grain surface flatness.In addition,the working environment of grain leveling is dusty and the work intensity of grain leveling is high,which has adverse effects on the health of workers who have been working on grain leveling for a long time.Based on this,thesis designs a vision-based autonomous grain leveling robot,and studies the two core issues of the visual navigation algorithm of the grain leveling robot and the control strategy of the two-machine coordinated motion,solved the problem of its straight swing and dual-machine out-of-sync problem.The main tasks are as follows:1.Design the visual navigation system of the grain leveling robot,and design and optimize the foreground straight line detection process and method for the dusty and local strong light irradiated granary environment,use the EXG method to obtain the grayscale image,and use the OTSU algorithm to do global threshold segmentation for the foreground image,and design a straight line detection algorithm for the edge of the granary aisle board under strong light conditions,that is,the slope inheritance line detection method,to solve the problem of local distortion of line detection under strong light;2.Establish a single-machine kinematic model of the grain leveling robot,establish a dual-machine cooperative kinematic model based on the leader-follower cooperative strategy,and obtain the differential expression of the dual-machine cooperative error;3.Research the dual-machine cooperative control scheme of the grain leveling robot,design the follower robot's speed and angular velocity tracking controller based on the feedback linearization theory,and analyze the stability of the system according to the Lyapunov second method.Simulink is used to simulate the dual-machine cooperative controller to verify the validity of the control model.Correct the cumulative error and systematic error of the cooperative motion process based on the visual method;4.Build the experimental platform and write the visual navigation code to verify the applicability of the slope inheritance line detection algorithm to the edge line detection of the walkway board under the condition of local strong light illumination in the foreground.And compared with the traditional Hough line detection method to verify the superiority of the former in this grain leveling robot project.