| Nowadays,the logistics industry mainly depends on labor for loading and unloading goods of containers or vans,it has an urgent need for smart loading and unloading mobile robots.The self-positioning and motion control are key technologies of mobile robots,and they are the premise of automatic loading and unloading inside the carriage,navigation and control during the movement process,and collaborative work.This article conducts research on the two key technologies,the main work and conclusions are as follows:Based on the working environment and system performance requirements of mobile robots for loading and unloading,a combined positioning scheme consisting of a laser ranging sensor,an ultrasonic distance sensor array,driving wheel encoders,and a gyroscope is proposed to effectively solve the problems of driving wheel slippage and positioning accuracy requirements in different regions.The odometry and ranging method positioning models are deduced,and the characteristics of multiple data sources and sensors of the same kind in each positioning sub-scheme are analyzed.Based on Kalman filter and Bayesian estimation,three kinds of multi-sensor data fusion methods are put forward,in which,the positioning data is processed in phases and implemented,the sensor data and system model are fully utilized to improve the system positioning accuracy.Based on the artificial potential field method,a selection method of Bezier curve control points is proposed.The distance between control points is calculated by the obstacle distance and influencing area,and the obstacle avoidance problem in the Bessel trajectory planning with obstacles is solved.For the trajectory data,a segmented velocity planning method based on the special inflection point of the trajectory is proposed to avoid the robot from turning too fast and reduce the impact of motion.Optimize and verify the smoothness parameters and operation cycle in the speed planning to achieve the time optimal path planning under the condition of the performance limit.Simulation and experimental have verified the dynamic follow-up positioning accuracy of the mobile robot in loading and unloading which meet the system requirements.The positioning fluctuation is 2 mm in x direction,0.1 mm in y direction,and 0.5 degree in angle direction.The motion control accuracy is about 10 mm,where the x direction is about 8 mm,and the y direction is about 3mm.The odometer and gyro combined positioning method with positioning method and data fusion method proposed in this paper effectively solve the positioning problem in the narrow and semi-enclosed space of the mobile robot,realized the robot's motion control needs are realized which is combined with path planning and navigation. |
PDF Full Text Request