Research And Design Of Positioning And Navigation System Of Orchard Robot

With the development of computer technology and sensor technology,intelligent robot has been applied to many fields,such as industry,service industry,agriculture and so on.Orchard mechanization is an important part of agricultural production mechanization.Using orchard robot instead of manual operation can effectively improve the automation degree of orchard operation and improve the efficiency of production so that the production scale of orchard can be expanded.The positioning and navigation technology of orchard robot is a foundation of realizing various functions.In this paper,according to the characteristics of orchard environment,a positioning and navigation system for orchard robot is designed,and the main work is completed as follows:(1)Through the analysis of the functional requirements of the positioning and navigation system of the orchard robot,the overall design of the software and hardware has been carried out.The hardware integration and the selection of the key hardware are completed.The software framework is designed based on the basic principles.(2)According to the characteristics of orchard environment,a method of dead reckoning assisted by inertial measurement element based on odometer is designed,which is not dependent on the external environment.At the same time,the advantages and disadvantages of dead reckoning and laser positioning are both analyzed.So that based on the improved Cubature Kalman filter algorithm,a positioning method that fuses laser positioning and dead reckoning is designed,which can effectively utilize the advantages of multi-sensor to complement each other and improve the positioning accuracy.(3)Based on the analysis of orchard environment,the orchard map model has been established by grid method.On this basis,an improved genetic ant colony optimization is designed for efficient route planning,which makes up for the blindness of the initial search of the ant colony optimization by using the genetic algorithm.The heuristic function and pheromone update strategy of the ant colony optimization are also improved,so as to speed up the convergence of the algorithm and improve the efficiency of route planning.(4)In view of the fact that the actual motor speed is limited,the Lyapunov function is used to construct the trajectory tracking control law and limit the output.The global asymptotic convergence of the trajectory tracking error is analyzed under the designed trajectory tracking control law,and the effective tracking of the desired trajectory is realized.Then,according to the actual motor control situation,the discrete points between the adjacent points are constructed based on the control cycle to form a complete discrete trajectory,and the trajectory is optimized to make the movement of the orchard robot smoother and to reduce the tracking error and the mechanical wear of the robot's wheels.(5)Based on the ROS platform,three functional modules of the positioning and navigation system that are positioning,route planning and motion control are realized and tested.The experiment shows that the positioning and navigation system designed in this paper can meet the expected functions of the subject.