Design Of Navigation And Obstacle Avoidance System For Unmanned Spray Vehicle Based On ROS

Nowadays,with the rapid development of intelligent equipment,the application technology of unmanned vehicles has become more and more mature.A variety of unmanned vehicle equipments have been developed to meet the application needs of various scenarios.Therefore,in order to further improve the efficiency and convenience of crop management,reduce manual effort and the health effects of manual spraying,this thesis has studied and designed a set of unmanned vehicle navigation and obstacle avoidance system based on the existing unmanned vehicle technology,which is suitable for farmland crop spraying.The main research contents of this thesis are described as follows:(1)A system of unmanned vehicle positioning,navigation and obstacle avoidance is designed.The system is mainly composed of four parts: environment perception,self-positioning,path planning and motion control.The system function realization mainly depends on the hardware of 32 microcontroller,industrial control machine,sensor.The main implementation principle is to install the ros system by the industrial control machine as the upper machine.By integrating various functional packages in the ros system in the upper machine,the unique node communication mode is used to carry out data interaction of each functional part.32 microcontroller of the lower machine by the control of the displacement and direction and through serial port sent to the upper computer,the upper machine received the data,call the stored good map and combined with the current radar scanning data,through the positioning algorithm to determine the current unmanned map location,according to the global planning path and obstacles call local path planning algorithm to determine the next moment of driving direction and speed and sent to the lower machine,lower machine by pid algorithm speed approximation,cycle,until reach the end.After the software installation and debugging,the system can realize the unmanned driving function.(2)The software layer structure and technical implementation of the navigation and obstacle avoidance system are designed and explained.Firstly,according to the radar scanning of static environment map and starting point location information,call the global path planner new A * algorithm according to the operation path requirements to avoid the initial obstacle and reach the end of the global path.Secondly,when the unmanned spray car according to the global path running process,real-time call the local path planner DWA algorithm,the global path trajectory point as the real-time starting point,and according to the algorithm rules and the location of the obstacles between the local obstacle avoidance route in real time,at the same time timely back to the global path point after avoiding new obstacles.In the whole operation process of the unmanned spray vehicle,it is also necessary to estimate the real-time position of the unmanned spray vehicle according to the feedback of the encoder and combined with the positioning algorithm AMCL,so that the unmanned spray vehicle can move normally forward along the global route or local route.Finally,the Gazebo simulation environment verifies that the navigation obstacle avoidance system can operate normally.(3)This thesis presents a path planning algorithm improvement based on the traditional A star algorithm to meet the specific scene requirements of farmland crop spraying.First of all,the calculation term of the heuristic function is added to the traditional A star algorithm.The new heuristic function will be enabled when there are obstacles around the path point,and the new movement cost will be calculated according to the cost function.This heuristic function term will guide the algorithm to select the path point with obstacles around when making path selection.Secondly,because the initial improved A star algorithm will select the circuitous path when the distance between the obstacles on both sides is far away,and there will be many unnecessary path points.Therefore,the improved A star algorithm is reoptimized,so that the improved A star algorithm can automatically select the path point in the middle when the obstacles are far away.The ultimate purpose of the improved A-star algorithm is to enable the unmanned spray vehicle to advance along the crops that need to be sprayed during the application of this algorithm,and at the same time,select the best spraying route according to the distance size of the obstacles on both sides.Finally,according to the results of the simulation experiment,the improved A star algorithm completes the above functional requirements well.