斜抛轨迹预估算法在毽球机器人中的应用

Track Prediction of Projectile Motion in Shuttlecock Robot

介绍一种利用卡尔曼滤波对视觉传感器传入的不规则形状物体的位置信息进行优化和滤波的算法,以及该算法在毽球轨迹预估中的应用。毽球属于一种形状不规则的物体,同时由于质量分布较为不均且运动时受空气阻力影响较大,所以无论是对它的视觉识别还是对它的模型分析都会难免遇到不小的噪声干扰,采用卡尔曼滤波对测得的毽球的位置信息利用毽球运动学模型进行滤波,同时由于对算法计算时间上较高的要求,将毽球的运动进行空间分解,这样减小了状态向量的维度,大大缩短了算法的执行周期。仿真实验证明了算法的有效性;在"毽球机器人"科研平台上有效地实现了对毽球运动轨迹的快速,准确的预估。

The algorithm of Kalman filter is used for optimizing and filtering the position signal of shuttlecock obtained by the vision servo system.The position of shuttlecock and the kinematic model of shuttlecock motion are so difficult to be obtained accurately because of several kinds of noise that Kalman filter is used to filter the position signal of shuttlecock before track fitting.After the state transition equations of Kalman filter is decomposed into 'x,y and z' three axes,the total time spent on matrix operations in Kalman filter algorithm is much less,which is important to control Shuttlecock Robot fast and accurately.The simulation results show that the accuracy of track prediction is improved effectively.When applied on 'Shuttlecock Robot',track prediction is also accomplished fast and accurately.