【作者】 胡士强； 敬忠良；

【Author】 Hu Shiqiang Jing Zhongliang (Department of Automation, Shanghai Jiaotong University, Shanghai 200030, China)

【机构】 上海交通大学自动化系；

【摘要】 给出一种基于噪声方差估计的稳健融合算法。针对集中式多传感器融合结构,利用随机逼近算法分析了各传感器权的最优分配原则,并给出了一种基于方差估计的权的自适应调整算法。结合多传感器融合跟踪系统,为了提高非线性系统的跟踪性能,滤波算法采用了非线性逼近能力更好的Unscented卡尔曼滤波算法。通过计算机仿真实验,获得了良好的融合效果。与经典的加权方法相比,在相同仿真航路条件下,融合精度和跟踪精度都得到了显著提高。更多还原

【Abstract】 A robust fusion algorithm based on noise variance estimation is presented. Aiming at the centralized structure of multi-sensor fusion, the optimal weighting coefficients are obtained using the stochastic approximation theory, and an algorithm that is able to adapt itself to the changes of sensor’s measurements noise is given. For the multi-sensor tracking system, in order to improve the tracking ability of non-linear system, Unscented Kalman filter is used, which has the good character of non-linear approximation. Computer simulation has shown that under the condition of same simulation’s trajectory, this algorithm brings up higher fusion and tracking precision compared with that of traditional weight method.更多还原