Port AGV Path Planning And Navigation Algorithm Research Based On Multi-sensor Information Fusion

As the main mode of trade transportation,the volume of maritime transportation has increased with the increasing frequency of trade between countries,and the turnover of China’s import and export trade has also increased gradually.As a transit station for maritime transport,the automation of port operation is of great significance to improve the efficiency of maritime transport,and the automation of port AGV is an important part of the automation of port operation.The automated operation of port AGVs is an important part of the automated operation of ports.The precise positioning and path planning of port AGVs are inseparable from the automated operation.This paper focuses on the precise positioning and path planning of port AGVs,and the research work is as follows:(1)In view of the high construction and maintenance costs associated with the current mainstream electromagnetic navigation and positioning method of port AGVs,which is not conducive to the re-engineering of old terminals and the development needs of automation in new terminals,a more intelligent and efficient multi-sensor information fusion method is used for positioning.In this paper,INS and GPS are selected for multi-sensor information fusion.In order to synchronize the time of different output frequency sensors,the factor map algorithm is used as the multi-sensor information fusion algorithm.To address the problem that the GPS signal is susceptible to interference and the missing signal degrades the positioning accuracy of the whole system,a random forest regression-assisted factor graph combination navigation algorithm is proposed on the basis of the factor graph algorithm.When the GPS signal is normal,a random forest regression model is established and trained with the velocity and position increments solved by INS and GPS at adjacent moments.When the GPS signal is missing,the velocity and position information from INS is used to simulate the velocity and position information from GPS at the current moment to maintain the integrity of the combined navigation system,thus suppressing the error growth and ensuring the good positioning accuracy of the port AGV.The simulation and physical tests show that the improved algorithm can suppress the error growth and maintain good positioning accuracy in the absence of GPS signal.(2)To address the problem that the traditional BAS path planning algorithm has a random search direction,which may plan poor path points and affect the rapid convergence of the algorithm,the relative position direction is introduced to make the path search more targeted and reduce the impact of the random search direction on the planned path and planning time.The results show that the improved BAS path planning algorithm can plan shorter paths in less time.(3)A turnaround mechanism is proposed to address the drawback that the traditional BAS path planning algorithm is prone to fall into the depressions of complex obstacles,i.e.,into the local optimal region.When the algorithm falls into the local optimal region and cannot jump out,it achieves the purpose of jumping out of the local optimal region by reversing the orientation at this point.By designing simulation experiments,it is verified that the turnaround mechanism can substantially increase the probability of jumping out of the local optimal region.