Research On The Method Of Visual Navigation And Motion Control For Express Carrying Robot

With the rapid development of e-commerce,There has a large accumulation of express goods in the logistics center.At present,the operation mode of manual handling and automatic transportation system developed by most logistics centers are far from sufficient in China,so it is urgent to research and develop an kind of express carrying robot applied in the logistics centers.Starting from the actual application scene of the express carrying robot in the logistics center,this paper studies key technologies of the robot through understanding about related technologies at home and abroad.Firstly,the robot model is introduced.It's mainly divided into two parts of lifting structure and driving structure: in the lifting structure,it's used to control the rotational tray to load and unload expresses for the positive and negative rotation of upper stepper motor;in the driving structure,it's used to adjust the velocity difference output of the left and right servomotors to drive autonomously through the deflection angle along the present track being located with an industrial camera.Secondly,on the one hand,the control method of S-type speed curve for stepper motor in lifting structure is studied,of which the sampling way of discrete frequency points is improved by multi-frequency and non-uniform method on the basis of the simplifying five S-type acceleration and deceleration curve;on the other hand,the visual positioning technology for the navigation in the driving structure is studied,of which an improved Faster R-CNN is put forward to detect four small rectangles of different colors in the designed navigation pattern.For the improved Faster R-CNN,convolution layers are added on the basis of Faster R-CNN network model,and then anchor frames are generated on multilevel feature maps.Later,these generated anchor frames are respectively convoluted through two 3 * 3 convolution kernels,classifying and regressing directly.Next,Harris corner detection algorithm is used to extract reference corners of the location area and Pn P algorithm is used to compute industrial camera pose during shooting followed by.Further the real-time deflection angle of the robot being obtained.Finally,the fuzzy control method is adopted to control the robot driving along the present track,viewing the deviation angle and its change rate as input variables and the speed difference of servo motors on the left and right sides as a output variable.Through the introduction of the mode and motion principle of the express carrying robot,firstly the rising and declining speed of the lifting structure is analyzed and improved,and then a visual navigation and positioning strategy of the drive structure is put forward and its target detection algorithm is optimized,finally a reasonable fuzzy controller is designed to control the pose of the robot when it drives along the preset track.It enables the robot to meet the requirements of stability and intelligent precise navigation.Therefore,the Robot can be used in the logistics center to improve the efficiency of express goods carrying.