【摘要】 机动目标跟踪一直是智能汽车环境感知领域的研究重点和难点,为了实现对目标车辆的准确跟踪,研究了将交互多模型(interactingmultiplemodel,IMM)算法应用到智能汽车对目标的跟踪中。首先,根据目标车辆的真实运动情况提出一种基于"当前"统计模型和Singer模型的模型集,然后,在PreScan中搭建模拟场景,将此模型集应用到IMM目标跟踪算法,并将本算法与基于"当前"统计模型和基于Singer模型的两种单模型目标跟踪算法进行仿真比较。结果表明此IMM算法在智能车辆目标跟踪方面具有较高的应用价值。更多还原

【Abstract】 The maneuvering targets track is always an important and hot research field in environmental perception of Intelligent Automobile, to track a moving vehicle target, the interacting multiple model(IMM) algorithm will be used. Firstly, according to the actual movement of vehicle target, the paper establishes a new model set include a current statistical model and a Singer model. Then, the new model set is applied to IMM target trackalgorithm in Pre Scan simulation scene.Finally, the IMM algorithm is compared with the single model tracking algorithm based on current statistical modeland Singer model. The simulated results show that the IMM algorithm could get more accurate and reliable performance for vehicle positioning.更多还原