In view of the threshold sensitive problem of VFH series algorithms，a new adaptive threshold improvement strategy is proposed. First the initial threshold and the threshold range is determined by the hardware，motion characteristics of the robot and the target environment. Then, a threshold evaluation function is used to comprehensively evaluate each threshold and accessible direction in the optional range，so that the robot can get the threshold suitable for the current situation in real time. Finally，the algorithm is realized in ROS，and several comparative experiments are conducted using the EAI mobile robot platform. The experimental results show that when the improved policy with adaptive threshold is applied，the robot can avoid local dead zone and successfully reach the target position in complex environment, especially when the target is surrounded with obstacles.
