一种改进的机动目标拦截算法

An Improved Algorithm for Maneuverable Target Interception

机动目标的拦截问题可以用"追逃游戏"模型描述,考虑追击者的惯性、控制量有限等动力学特征时,拦截位置和控制量的确定是非常困难的.针对平面机动目标的拦截问题,以机器人拦截足球为模型,提出了一种改进的拦截算法——改进的"预测-规划-执行"法,通过预测机动目标的运动轨迹,提出了一种快速简便地确定粗略拦截位置的试探方法,在此基础上,再确定机器人的控制速度和精确拦截位置,控制机器人朝拦截位置运动.最后,给出了仿真计算结果,并应用于机器人足球比赛,仿真计算表明,算法可以准确有效地拦截运动目标.

Interception of a maneuverable target can be described as a ＂pursuit-evasion game＂ model. It is difficult to determine the position of the engagement in respect to the dynamic characters of the pursuer, such as inertia and limited control. Based on the model of a robot intercepting a soccer ball in a robot soccer competition, an improved ＂prediction, planning and execution＂ algorithm is proposed for planar maneuverable target interception. Through predicting a moving target＇s trajectory and probing a possible engagement position, a quick and convenient approach to calculating the interception location is proposed, and then a controlling strategy for the robot and an accurate engagement position are determined to drive the robot to the engaged location in given time. Finally, simulation results are obtained from the application of the algorithm to robot soccer games. They demonstrate that the algorithm can accurately and effectively intercept the maneuverable target.