论文部分内容阅读
为了进一步提高自主水下机器人(AUV)纯方位目标跟踪能力,从AUV轨迹优化方面进行了研究.采用基于距离的分段轨迹优化方法:在跟踪目标的初始阶段以定位的位置误差GDOP(geometrical dilution of precision)作为优化对象,以期在定位跟踪的各个时刻能得到最优的定位精度;针对目标运动要素(位置、速度、航向等)估计趋于收敛的情况,提出了一种基于短期预测的轨迹优化方法,AUV根据物理条件限制预测双方短期状态,计算能够反映跟踪态势特征的收益函数,根据收益函数对自身某状态进行评估,估算出自身各个预测状态的综合收益后,选出综合收益最大的那个状态作为短期目标,执行能到达该状态的行为.目标运动要素估计中使用扩展卡尔曼滤波(EKF).最后,将该轨迹优化方法与基于GDOP的轨迹优化进行仿真对比,结果表明该方法能够实现AUV与目标较快汇合.
In order to further improve the target-tracking ability of autonomous underwater vehicles (AUVs), the AUV trajectory optimization is studied.Using distance-based segmentation trajectory optimization method, the position error GDOP of precision as the object of optimization so as to obtain the best positioning accuracy at each moment of positioning and tracking. In view of the fact that the convergence of the target motion elements (position, velocity, heading and so on) tends to be convergent, a trajectory based on short-term prediction Optimization method, AUV limits the short-term state of both sides according to the physical conditions, calculates the revenue function that can reflect the characteristics of the tracking situation, evaluates a certain state according to the revenue function, estimates the comprehensive income of each forecast state, and selects the largest comprehensive income The state is used as a short-term goal to implement the state that can reach this state.EKF is used in the estimation of the target motion components.Finally, the trajectory optimization method is compared with the trajectory optimization based on GDOP, the simulation results show that this method can Achieve AUV and the goal of faster convergence.