AGV Dynamic Based On Visual Road Condition Information Research On Path Planning

With the development of "Industry 4.0" and "Made in China 2025",the traditional manufacturing industry has gradually transformed into a smart manufacturing industry.In this process,Automated Guided Vehicles(AGV)is becoming increasingly prominent role and widely used in sorting,transportation,assembly and other fields for its accuracy,high flexibility and stability.This paper studies the image processing algorithm and path planning algorithm of AGV visual navigation to realize AGV autonomous driving.This paper designs the AGV software and hardware platform according to the requirements of designing visual navigation.It designs the six-wheel chassis structure which mainly driven by the intermediate wheels and partly driven by the front and rear wheels.Through the analysis of the AGV's motion performance and load-bearing performance,the appropriate drive motor is selected to design the AGV's vision system,control system and obstacle avoidance system hardware structure.Together with the constructed AGV software system platform and software system architecture,it constructs a practical experimental AGV to provide an experimental platform for the verification of navigation algorithms.The AGV's visual inspection mainly includes the detection of the foreground image and the side view image.In this paper,the improved FT algorithm is applied to detecting the foreground image which performs visual saliency processing on the image to determine the position of the two-dimensional code,and then the two-dimensional code recognition is performed after the image is cut to determine the current position of the AGV;for the side view image,the inverse perspective transformation algorithm is used for the image correction,and then the Hough transform is used to extract the road marking information,and calculate the deviation angle and deviation distance of AGV.For the path planning of AGV,this paper analyzes the global and local path planning algorithms,and improves the Jump Point Search(JPS)and Artificial Potential Field(APF)methods.Combining the advantages of the two algorithms,JA algorithm is proposed in this paper to realize AGV path planning under complex road conditions,which realizes the need to quickly find the shortest path with the fewest inflection points in the AGV working environment,and can realize the real-time autonomous obstacle avoidance function.Finally,the visual inspection function and path planning function of the visual navigation AGV are experimentally verified on the designed AGV experiment platform.The results of visual inspection experiments show that the visual navigation AGV can detect the surrounding environment information in real time,and perform real-time positioning and posture deviation correction.The results of path planning experiments show that AGV can autonomously plan a shortest path with the fewest inflection points according to the mission goal,and also adjust the path in real time according to the road condition information to achieve intelligent obstacle avoidance.