Research On AGV Positioning And Navigation Technology Based On Magnetic Nail Signal

In order to relieve the pressure of parcel sorting,the warehouse workshop is gradually replacing manual delivery with AGV to improve the sorting efficiency and provide customers with more convenient and efficient service.In AGV navigation is the key technology,in view of the AGV navigation is also a lot of work,at home and abroad mainly include: inertial navigation,optical navigation,magnetic stripe navigation,laser radar,navigation,visual navigation of code,but these navigation methods have limitations in application,such as:expensive,ground the destructive,vulnerable to interference,complex maintenance and so on.In order to explore a more stable and reliable AGV navigation technology,this topic studies the AGV positioning and navigation technology based on magnetic nail signals.The main contents of this study are as follows:1)Magnetic sensor array design and error calibration.Firstly,the matrix magnetic sensor array is designed according to the characteristics of different magnetic sensors.Secondly,due to the existence of errors,the ellipsoid fitting method is used to calibrate the magnetic sensors to obtain hard magnetic and soft magnetic errors.According to the experimental analysis of positioning accuracy before and after calibration,the pre-calibration error is 2.74 mm,and the post-calibration error is 0.68 mm.The results show that the precision of magnetic sensor is improved obviously by the calibration method.2)Research on localization algorithm of magnetic nail signal.A hybrid method of Levenberg-Marquardt(LM)and particle swarm optimization(PSO)is proposed to locate the magnetic nail.The initial value is provided by PSO algorithm,and the LM algorithm iterates the position of the magnetic nail according to the initial value.The positioning accuracy of this method can reach 0.97±0.55 mm.Secondly,according to the characteristics of the magnetic nail’s transverse position,the method of solving the unvariate quartic equation of Tian Heng formula is proposed,and the magnetic nail’s positioning accuracy is 0.72±0.54 mm by mathematical analysis.At the same time,the influence of the positioning time and magnetic nail installation error on the results is analyzed experimentally.3)Research and experimental verification of magnetic nail navigation algorithm.According to the characteristics of magnetic nail navigation scheme,an adaptive fusion algorithm was proposed to fuse the information of odometer and magnetic sensor array on the moving vehicle,determine the weight according to the reliability of magnetic field signal detected on the moving,and fuse the data,the navigation accuracy is 110±30mm.Secondly,an extended Kalman fusion algorithm is proposed for data fusion.By using odometer information as state equation and magnetic sensor information as observation equation,the navigation accuracy of the scheme is 83±15mm.To sum up,the positioning and navigation of AGV based on magnetic nail signals and odometer information can be realized under the fusion of the above fusion algorithms.Meanwhile,due to the advantages of low cost,convenience,flexibility and reliable positioning,it will be widely applied in the industrial field in the future.