【摘要】 提出一种面向煤矿井下线性拟合和卡尔曼滤波相结合的改进分布式目标跟踪算法。根据移动目标的当前位置建立动态簇,簇头节点集中处理簇成员节点发来的最新观测数据,结合线性拟合算法和卡尔曼滤波算法对移动目标进行预测,将线性拟合的预测值和卡尔曼滤波预测值作为真正的预测值,得到目标的状态估计,通过这样的改进可实时的修正预测值。仿真结果表明,与传统的分布式目标跟踪算法比较,改进算法集中了2种算法的优点,有很好的跟踪性能。更多还原

【Abstract】 We propose an improved distributed target tracking algorithm that combines linear fitting with Kalman filtering for coal mines. We establish dynamic cluster by the current position of moving targets, and cluster head nodes will process the latest observation data from cluster member nodes. Finally, we forecast the moving targets by combining linear fitting method and Kalman filtering algorithm, and take the predicting values obtained from the linear fitting and the predicted values from Kalman filtering as the real predictive values to obtain the state estimation of targets. Through such improvement, we can correct the predicted values timely. Simulation results show that the improved algorithm combined the advantages of two algorithms has a good tracking performance compared with the traditional distributed target tracking algorithm.更多还原