Research On AGV Autonomous Navigation And Control Strategy Based On Vision

As a kind of automatic logistics equipment,AGV(Automated guided vehicle)is increasingly being used in material handing and assembly process.Navigation technology is the core technology of AGV.Visual navigation,as one of the navigation modes of AGV,has the advantages of large amount of information,simple and flexible path setting,easy maintenance and reconstruction,and easy identification of multi-branch paths and various parking stations,etc.It has a broad prospect of development and application.In this paper,the Pioneer 3-AT mobile platform was used as the experimental platform of AGV.The circular coded target were used as the path identification identifier of visual navigation,and the fuzzy PID controller was designed to realize the navigation of AGV.At the same time,the topological method was used to establish the map and the monocular visual ranging was used to calculate the AGV position information.The main research contents of this paper are as follows:(1)This paper introduced the software and hardware structure of Pioneer 3-AT mobile platform,and carried out model simplification and kinematics analysis.The mathematical model of Pioneer 3-AT was established.The difference between the left and right speeds was used as the input variable of the open-loop control system.It laid the foundation for the subsequent navigation control.(2)This paper redesigned the circular coded target and proposed a corresponding detection algorithm for the problems of small number and accurate recognition of traditional circular coded target.This paper added three locators on the basis of traditional circular coded target to achieve precision position and increase the number of targets.Firstly,the algorithm detected the locator coordinates and their positions in the targets.Then,the perspective transformation of the coded target was used to achieve image distortion correction.Finally,used the method based on the ring scan to achieve the decoding of the circular coded target.The experimental results show that the algorithm proposed in this paper has a good detection and recognition effect on circular coded target under arbitrary rotation angles,different shooting angles and complex backgrounds.And in the detection and identification of circular coded target,the navigation information of the AGV can be calculated according to the navigation model.(3)In order to realize the navigation control of AGV,the fuzzy PID controller was designed and used as the visual navigation controller.Firstly,the structure of fuzzy PID controller and the detailed process of its design were introduced.Then,this paper explained the basic principles of particle swarm optimization and improved the algorithm.The improved algorithm set the upper and lower limits of inertial weight coefficients in PSO and decreased nonlinearly with the iteration times in the form of gamma function,at the same time,the inertia weight coefficient and learning factor of particles were dynamically adjusted according to the fitness of particles,made the particle keep the reasonable motion inertia and learning ability,improved the self-adaptive ability of the particle.And the algorithm test function was used to test.The test results show that the improved particle swarm optimization algorithm has the advantages of fast convergence speed and high convergence precision.Finally,the proposed improved particle swarm optimization algorithm was used to optimize the parameters of the PID controller and the optimized parameters were used in the fuzzy PID control system.The simulation results show that the proposed algorithm achieves better control performance and effect.(4)In order to realize the calculation of the AGV position information and the visual display of the AGV position in real time.Firstly,using the coded marker points as nodes,the topological method was used to establish the electronic map of visual navigation,and the distance from the coded marker point to the AGV was calculated by monocular vision ranging,and then the current position of the AGV was calculated by querying the electronic map.Followed by software design,software design included the software design of the upper computer and the lower computer(AGV ontology).The host computer adopts the MFC development platform in VS2013 to design it,and used GDI+ drawing method to draw navigation maps by double buffering technology.The communication mode between the upper computer and the lower computer was completed by the communication method of UDP protocol;the lower computer included image acquisition,detection and identification of coded points,navigation information calculation,data transmission and AGV control and so on.Experiments show that the AGV has the ability to track the navigation line and its location information in real time.