分布式异步测量传感器网络机器人控制

摘要

Generally, in the distributed WSN target tracking system, we get the distance between the target and the nodes simultaneously, and then calculate the position of the target by multi-point location. According to the position, we can figute out the control input of the mobile robot. But in actual WSN, the measurement of each node was not always synchronous. Strictly speaking, several distance used in multi-point location algorithm was not the one between the target and the node at the same time. In this paper presents a measurement method of asynchronous measurement, which properly estimates several distance to different node with noise and time delay, then apply this method of Kalman filter in smoothing and estimating the actual route of the mobile robot. Finally, This method was proved to be effective in the six gird WSN.%现有的分布式无线传感器网络目标定位跟踪系统,通常根据分布在不同位置的节点同时测量目标与本节点之间距离,通过多点定位计算目标位置,据此对移动机器人进行导航控制.然而在实际无线传感器网络中,各节点测量往往不是同步进行的,多点定位算法中使用到的测距值,严格意义上并不是同一时刻目标与各节点的距离.文中提出一种基于异步测量的卡尔曼滤波方法,能有效融合包含测量噪声、时延噪声的测距数据.将处理后的数据运用到移动机器人导航控制上,能有效地对移动机器人运动轨迹进行平滑、预测,并能准确算出其位于网络死区时的坐标及航向角.此测量方式、控制算法在6网格无线传感器网络中,获得了很好的实际控制效果.