To track multiple maneuvering targets hidden in the Doppler blind zone (DBZ), we have proposed an MM-GLMB-DBZ tracker based on the latest multiple model generalized labeled multi-Bernoulli (MM-GLMB) filter. To further enhance the tracking performance, this paper combines the sensor control technique to the MM-GLMB-DBZ tracker. Macroscopically, the proposed algorithm consists of the MM-GLMB-DBZ tracker and a controller. Unlike conventional control approaches where separate prediction and update implementations are usually adopted, the proposed control algorithm constructs a systematic process flow for the joint prediction and update implementation of GLMB-like filters. Moreover, inside the core controller module, we apply the previously designed safety indicator and reward function for avoiding the DBZ, and derive the Cauchy-Schwarz divergence (CSD) compatible with the tracker. Hence, this control algorithm considers such factors as the safety of the sensor itself, the DBZ avoidance, and the improvement of the tracking accuracy. Numerical examples verify the effectiveness of the proposed control scheme, showing that it is significantly better than the random control strategy, the original MM-GLMB-DBZ tracker without the control technology, and the state-of-the-art control approach.

中文翻译：

---

基于 MM GLMB 滤波器的传感器控制，用于跟踪隐藏在多普勒盲区中的多个机动目标

为了跟踪隐藏在多普勒盲区 (DBZ) 中的多个机动目标，我们提出了一种基于最新的多模型广义标记多伯努利 (MM-GLMB) 滤波器的 MM-GLMB-DBZ 跟踪器。为了进一步提高跟踪性能，本文将传感器控制技术与 MM-GLMB-DBZ 跟踪器相结合。宏观上，所提出的算法由 MM-GLMB-DBZ 跟踪器和控制器组成。与通常采用单独预测和更新实现的传统控制方法不同，所提出的控制算法构建了一个系统的过程流程，用于类 GLMB 滤波器的联合预测和更新实现。此外，在核心控制器模块内部，我们应用了先前设计的安全指标和奖励函数来避免 DBZ，并导出与跟踪器兼容​​的柯西-施瓦茨散度 (CSD)。因此，该控制算法考虑了传感器本身的安全性、DBZ 避免和跟踪精度的提高等因素。数值例子验证了所提出的控制方案的有效性，表明它明显优于随机控制策略、没有控制技术的原始 MM-GLMB-DBZ 跟踪器和最先进的控制方法。