Research On Indoor Path Planning Of Wheeled Educational Robot Based On RealSense

Robotics is a type of science related to the design,manufacture and application of robots.With the progress of society,the rapid development of the fourth industrial revolution and the rise of artificial intelligence technology,robots play an increasingly important role.The research on educational robots has become a hot issue in the scientific research field.Educational robots can be divided into wheeled robots,caterpillar robots and foot robots.Because the motion model of wheeled robots is relatively simple and the degree of freedom is relatively low,this paper chooses wheeled educational robots as the research object,puts forward the software and hardware framework of wheeled educational robots based on ROS,and studies the path planning of robots in the indoor environment under deep cameras.Educational robots in indoor environment can store real-time map information through SLAM(Synchronized Location and Map Construction)technology.It is called global path planning that chooses the corresponding algorithm from the current location to the designated point through the existing global map.However,in uncertain environment,real-time obstacles will be encountered.It is not feasible to rely solely on global path planning for navigation.At this time,visual sensors or lidar sensors are needed to detect obstacles and change the planning path locally.For the issue of path planning,single-threaded lidar or depth camera are usually leveraged as sensors to collect indoor environmental information.Depth camera is used to do the following research on indoor path planning in the paper.(1)In the process of testing wheeled educational robot,there will be problems of robot not walking straight line and speed control error.This paper proposes an improved scheme of PID control algorithm and debugs it.The PID control algorithm mainly reduces the speed control error of the robot by integrating from proportion.(2)Using the point cloud data collected by Intel Real Sense R200 depth camera,this paper studies the path planning in the local scope of the robot.If obstacles are detected during the course of the robot traveling along the established global navigation path,the depth camera is activated to grab the surrounding environment space of the obstacle,and the wider side is selected as the circumference space.The roundabout point and roundabout path are formed.At the same time,the end of the roundabout path after mapping is connected with the next node of the original global path to complete real-time obstacle avoidance.The main innovations of this paper are as follows:the integration of software and hardware framework of wheeled educational robot based on vision sensor is proposed,which includes scalable multi-sensor upper computer module and lower computer module to control motion;RealSense R200 is used as the vision sensor of wheeled robot to study the global path planning and local real-time path planning of the robot.Research and experimental results show that the wheeled education robot based on RealSense can accurately complete local real-time obstacle avoidance and return to the original global planning route.