摘要：AbstractThis paper proposes a dynamic trajectory planning method based on nonlinear model predictive control to avoid dynamic obstacle in actual scene. The trajectory is predetermined on the basis of sigmoid function in accordance with the road traffic condition. The obstacle may suddenly appear on the given trajectory in the form of motion, such as moving pedestrians and vehicles. In this case, to avoid obstacles more accurately, local dynamic adjustment should be made to the predetermined trajectory. Therefore, a motion trend function is built to describe the position variance of dynamic obstacle in prediction horizon. Combining the predicted obstacle positions and vehicle dynamics characteristics in prediction horizon, the predetermined trajectory is adjusted dynamically on the basis of point-mass model. Experimental results reveal that the optimized trajectory can avoid dynamic obstacle effectively and meet various requirements of collision avoidance.
关键词：KeywordsDynamic trajectory planningModel predictive controlDynamic obstaclePoint-mass model
