Research And Implementation Of Indoor Service Robot Navigation System Based On Lidar

The indoor service robot industry is developing rapidly,and the demand for service robots is increasing.Indoor service robots have also received widespread attention as a branch of mobile robots.Indoor service robots need not only the ability to recognize environmental information,but also the ability to navigate and avoid obstacles in the environment,so that they can be widely used in different scenarios.At present,scientific research institutions and production companies have conducted in-depth research on key technologies in robotic navigation.Robot navigation technology mainly includes the construction of environmental maps,robot positioning and path planning in the environment.Through these techniques,the robot builds a map in an unknown environment,performs path planning on the constructed map,avoids obstacles,and finally reaches the target point.Due to the different application scenarios of indoor service robots,the navigation requirements for robots are also different.The application scenario of the service robot in this paper is the indoor office environment,the relatively cumbersome work of transporting documents between different offices,and avoiding pedestrians and obstacles during the transportation process.Aiming at these application scenarios,the existing navigation technology foundation is analyzed,and the current navigation method is improved.The laser radar-based indoor service robot navigation system is designed.The main contents of this article are as follows:1.In the related theory of robot navigation,the accuracy of sensor data is crucial for the robot's autonomous navigation.However,during the movement of the robot,the laser data will produce motion distortion,which will affect the robot's autonomous navigation effect.In this paper,the use of mileage data to correct laser data,remove motion distortion,improve the accuracy of the sensor data expression environment,is the basis for subsequent robot navigation.2.Aiming at the problems of particle depletion and poor estimation accuracy for the simultaneous localization and mapping(SLAM)scheme based on particle filter algorithm,a particle filter algorithm based on improved resampling method is proposed to improve the accuracy of pose estimation while improving particle diversity.The effectiveness of the algorithm is verified by the simulation results of two models,one-dimensional non-linear tracking(UNG)and two-dimensional pure angle tracking(BOT).The environment map is effectively constructed based on the algorithm in navigation system.3.Since the existing obstacle avoidance algorithm can't avoid the obstacles that are close to the robot,an active obstacle avoidance algorithm based on improved dynamic window method(DWA)is proposed.Select the appropriate path based on the score of the speed corresponding to the simulated path.When the distance from the obstacle is too small,the back command is executed first,and then the path to the target point is re-planned,and finally,when the robot is too close to the obstacle,the obstacle is effectively avoided and the planned path is reproduced.And verify the effectiveness of the algorithm in the actual scenario.4.According to the functional requirements of the robot navigation system,combined with the laser SLAM module,positioning module and navigation module,the laser-based indoor service robot navigation system is designed and implemented,and the effectiveness of the improved algorithm is tested and verified on the system.