Research And Application Of Navigation Strategy For Intelligent AGV

Automatic Guided Vehicle(AGV)is also known as a handling robot.With the arrival of the e-commerce,more and more mobile robots are applied to the logistics warehouse.The demand for warehousing and logistics industry is increasing rapidly.Therefore,how to study the navigation strategy of the transport robot and realize the automation of logistics has become the current research focus.This paper starts with the actual application scenario and takes intelligent logistics warehouse as the research background.Using the method of combinating the QR codes and magnetic stripe navigation to index the landmarks.It analyzes the reason why multi-AGV is inefficient,and it has a lot of congestion and deadlocks.Design and research are based on two aspects: hybrid navigation library system modeling and multi-path multi-task scheduling control strategy.First,according to the storage characteristics of QR codes to store a warehouse shelf and path location information.Then the warehouse was modeled systematically,using the grid method and topology method to multi-objective continuous space discretization,which greatly reduces the amount of data transmission.Restrain multipath channels which had been formed as one-way paths.At the same time,using conflict resource control of the Bundle area to establish resource monitoring for the conflict area which effectively reduces the occurrence of deadlocks during multi-AGV navigation and facilitates subsequent scheduling of the control center.Then,the final solution of the navigation plan in intelligent warehouse is developed.Dijkstra algorithm and improved ant-colony algorithm are used for the original path planning,This paper focuses on the genetic manipulation of the population of single-point crossover and single-point mutation in the path planning based on improved genetic algorithm.The influence of environmental complexity on the application of this algorithm and the influence of population size,crossover probability and mutation rate on the output quality of the algorithm are both researched,the parameters are set as well.The dynamic operation efficiency of AGV is improved by applying the method of multilevel subtask which overlay the same shelf tasks in multi-task parallel scheduling strategy.Finally,the real environment of a warehouse is modeled,the AGV status attribute is configured and the dynamic task order is added through OpenTCS simulation experiment platform.The experiment results show that there is no such phenomenon as collision,deadlock and long jam in the dynamic operation of multi-AGV and multi-task order.The optimal path is monitored and all the next tasks are successfully completed,which verifies the feasibility,effectiveness and robustness of the navigation system.