| In an automated logistics storage system,AGV is the key to the successful completion of a series of complex tasks,such as automatic inventory and stacking of goods.The stable and high precision navigation and positioning of AGV is the key to the successful completion of the task.Because of the narrow traffic space and similar environment,the existing low cost positioning technology has a poor precision in this scene,which affects the AGV real-time navigation and normal operation,and the high precision positioning technology is complicated and cost high.Therefore,on the basis of considering the environment characteristics,it is important to study the simple and low cost AGV positioning method suitable for the warehouse cha nnel environment,and to improve the positioning accuracy of AGV,which has important engineering significance and theoretical value.In view of warehouse access environment,there is a lack of simple and accurate AGV positioning method,this paper studies the motion perception method based on the kinematic model and the motion perception method based on the matching of the laser radar scanning,and then gives the relative positioning method for the fusion of different sensor data,and then the design is designed according to the characteristics of the actual warehouse channel environment.The AGV global positioning method based on the channel right angle curve eliminates the cumulative error in the relative positioning process.Finally,the design and implementation of the AGV positioning system in the warehouse channel environment are completed.The main research contents include:(1)Dead reckoning and laser radar scanning,motion sensing based on AGV.The kinematic model based on the centroid of a car is established by the speed representation method to realize the motion estimation perception of the position speculating method.Based on the matching data of the laser radar,the motion increment perception between the adjacent time of AGV is carried out.By analyzing the characteristics of the actual channel environment,a matching method based on the corner feature is proposed.The experiment shows that the improved algorithm improves the feature extraction of the corner points.The accuracy and accuracy of ICP matching are improved.(2)Relative location based on data fusion.By analyzing the motion state of AGV obtained by different sensors,a relative location method of AGV based on EKF data fusion model is proposed.The position and pose estimation results are used as the predictive value of the system,and the position and posture information of the AGV position and pose of the previous moment and the optimal position estimation results of the previous moment are used as the observation values of the system,and the optimal estimation of the AGV position and posture at present time is obtained.The experiment shows that it has good real-time positioning accuracy.(3)AGV global positioning and pose updating based on right angled bending feature.Based on the analysis of typical environmental characteristics of warehouses,a global positioning method based on right angled bending characteristics of AGV is proposed.First of all,the model of the typical right angle bend of the warehouse is set up,then the method of distinguishing the right angle curve based on laser radar is designed.Finally,according to the key line and the intersection point feature of the straight line,the AGV position and attitude is calculated and the relative positioning results are updated.Experiments show that this method can effectively eliminate location accumulation error and improve the overall accuracy of AGV positioning.Finally,based on the above research results,we design and implement the AGV location system in warehouse channel environment.The application results show that the positioning method designed in this paper is simple and easy to improve the overall positioning accuracy of AGV with low technical cost,and has a certain generality in the warehouse environment. |
PDF Full Text Request