Research On Obstacle Avoidance Method Of Automated Guided Vehicle (AGV)

Automatic guided vehicle(AGV),as an automatic transporting tool,can reduce the production cost and improve the production efficiency of the enterprise,it has been widely used in the storage industry.In order to ensure that AGV can avoid obstacles on the moving path and accomplish the handling work safely,AGV needs to have accurate obstacle avoidance function.This paper aims at studying on the problem of target unreachable and local minimum existing in the traditional artificial potential field method,and the problem of the low accuracy of the obstacle avoidance and the poor adaptability of the fuzzy control algorithm.The main contents are as follows:Firstly,a method to optimize the improved artificial potential field method by using chaos optimization is proposed: on the basis of the traditional artificial potential field method,the repulsive force potential field function is modified by adding the relative distance between AGV and target,making the target point as the minimum point of the global potential field,the resultant force of the AGV at the target point is zero,and the problem of the target unreachable is solved;the chaos optimization algorithm is used to perform the global search to obtain the global optimal value and avoiding falling into the local minimum problem.The simulation results show that the method can effectively solve the problem of target unreachable and local minimum existing in the traditional artificial potential field method,and obtain a relatively smooth motion trajectory.Secondly,the fuzzy neural network algorithm based on T-S model is studied: on the basis of fuzzy control algorithm,the BP neural network algorithm with self-learning ability is integrated,the network parameter learning method is used to the off-line training of the initial membership function of the fuzzy neural network,which improves the adaptability and control ability of the AGV in a complex environment,and makes up for the defects of the low accuracy of the obstacle avoidance and the poor adaptability of the fuzzy control algorithm.The simulation results show that the fuzzy neural network algorithm does not have a path point close to the obstacle compared to the fuzzy control algorithm,improves the obstacle avoidance accuracy of fuzzy control algorithms,and can more safely meet the obstacle avoidance requirements.