高斯混合标签多伯努利滤波器的传感器控制策略

针对多目标跟踪优化中的传感器控制问题,提出一种基于高斯混合标签多伯努利滤波器的快速传感器控制策略,并将其用于基于目标威胁度评估的风险评估中.首先,给出高斯混合标签多伯努利滤波器的实现形式,并对两个高斯混合之间的柯西施瓦兹散度求解.其次,利用多伯努利部分替代传感器控制中标签多伯努利伪更新步骤,即在多伯努利更新过程中采用了...

Sensor control strategy of Gaussian mixture labeled multi-Bernoulli filter

Aiming at the sensor control problem in the optimization of multi-target tracking, a fast sensor control strategy based on Gaussian mixture labeled multi-Bernoulli filter is proposed, and is used in the risk assessment based on the target threat assessment. First, the paper presents the realization form of the Gaussian mixture labeled multi-Bernoulli filter, and studies the calculation of the Cauchy Schwarz divergence between two Gaussian mixtures. Secondly, multi-Bernoulli is used to partially replace the labeled multi-Bernoulli filter pseudo update step in sensor control, that is, removing tagging is adopted in the process of multi-Bernoulli update, which significantly reduces the complexity of sensor control algorithm. In addition, based on the general sensor control method, the ultimate control scheme is solved by maximizing the target information gain of the maximum threat degree as the criterion. Simulation results show the effectiveness and rapidity of the proposed algorithm.