| Intelligent transportation is an important part of the new generation of artificial intelligence development planning,and AGV,as a type of intelligent transportation,also plays a key role.In order to reduce labor costs,unmanned storage and intelligent AGV have gradually become hotspots of research.With the development of artificial intelligence technology and AGV technology,AGV is applied to more and more fields,including factories,logistics,supermarkets,hospitals.AGV based on differential drive is a widely used material distribution vehicle.Because of the relatively speed required,it is a reasonable low-cost method to get reference-trajectory based on visual sensors.At the same time,differential drive system provides the flexible and changeable motion modes.The AGV is assumed to run in a factory environment as the background in this paper.In the factory,there are advantages such as stable light sources,clear guidance lines,and low AGV speed.Therefore,vision sensors and image processing are used to obtain the guided path and centerline.Then,the classic PID control and the funzzy PID are proposed for AGV trajectory tracking.Finally experiments are verified in Matlab / Simulink and PreScan.The research contents and results are as follows:(1)With the physical structure of four-wheel AGV driven by the rear wheel differential,the kinematic model and the control system is established based on the characteristics of the differential drive.The voltage difference signal of the two driving wheels is used as input,the speed difference,the angular deviation and distance deviation of the AGV are used as state variables to obtain the state equation of the entire control system.(2)The AGV is guided by the visual sensors in this paper,following the path model of line,circular and non-circular path.During processing,the images collected by the camera are pre-processed,including image enhancement and binarized grayscale image,to obtain the path boundary of the trajectory.Then road centerline of line and circular paths is obtained by least squares fitting method.The centerline of the non-arc path is fitted by the inscribed arc correction method,and then the position and angle deviations of the three path models during the AGV operation are obtained.(3)A fuzzy PID controller is obtained by combining fuzzy control and PID control.In order to verify the effectiveness and stability of the fuzzy PID controller algorithm designed in this paper,a comparative experiments will be performed with the traditional PID controller in Simulink.The simulation results shown that the settling time of fuzzy PID is less and smaller overshoot by intelligently adjust parameters in fuzzy PID,with the same large error initial state.Therefore,the fuzzy PID controller designed in this paper has stability and effectiveness.(4)The performance of the designed fuzzy PID controller was verified in the 3D simulation software PreScan.There are four sets of experiments: the straight line path with two different initial states,line-arc mixed path,and non-circular path.The experimental results shown that the fuzzy PID controller designed in this paper can follow the reference path even with the large initial state deviation,which indicates that the vision-based factory material distribution AGV designed in this paper can perform stably and effectively on the trajectory tracking. |
PDF Full Text Request