Research On AGV Path Planning And Location Method Based On Machine Learning

Automatic guided vehicle(AGV)is a new type of autonomous robot equipment,more and more researchers pay attention to it.Because of its high degree of automation and stable performance,it is used in modern logistics storage environment.In order to ensure that AGV can run safely and stably in the storage environment,the improvement of AGV application technology becomes particularly critical.Therefore,this paper mainly solves three key problems of AGV: environment modeling,path planning and accurate positioning.Firstly,aiming at the problem of AGV environment modeling,this paper proposes the grid method based on feature extraction.Firstly,the obstacle grid is classified by Monte Carlo search,and then the feature grid is extracted according to the feature extraction rules.Finally,the visual graph method is used to narrow the search range of machine learning algorithm.In order to verify the applicability of the improved grid method,the improved grid method is applied to artificial potential field method,a * algorithm and ant colony algorithm.The experimental results show that the improved grid method can effectively reduce the search range of machine learning algorithm and improve the operation efficiency of the algorithm.Secondly,according to the working environment of AGV,the path planning of AGV is divided into simple environment and complex environment.In this paper,a star ant colony algorithm is applied to solve the AGV path planning task in simple environment.Firstly,the new grid method proposed in the previous chapter is used to model,then the a * algorithm is used to estimate the cost of the environment,and the pheromone concentration with guiding effect is generated.Finally,the ant colony algorithm is applied to the new grid environment with initial pheromone concentration to complete the path planning.Aiming at the AGV path planning problem in complex environment,this paper chooses the depth deterministic strategy gradient algorithm to solve the path planning problem in complex environment.The depth deterministic strategy gradient algorithm has some problems,such as slow convergence speed,long training time and local optimization.Aiming at these problems,this paper applies a new search strategy,small batch first sampling and memory playback matrix to solve them.Comparative experiments show that the improved depth deterministic strategy gradient algorithm can effectively solve the AGV path planning problem in complex environment.Finally,for the AGV positioning problem,in order to improve the AGV positioning accuracy,this paper uses Chan AGA algorithm to solve the problem.On the basis of the traditional location method,firstly,Chan algorithm is used to narrow the location range for the adaptive genetic algorithm,and then the adaptive genetic algorithm is used to complete the accurate location.Through the experimental simulation,it can be seen that this method can effectively improve the positioning accuracy of AGV,and ensure the AGV operation more accurate and stable.