水下多源多目标不确定信息的随机有限集建模与融合跟踪研究

多目标融合跟踪是水下协同探测亟待解决的关键问题之一，也是信息融合的重要研究方向。多目标跟踪有着自身的复杂性，随机有限集（RFS）以较强的多维不确定信息表示和处理能力为其研究提供了新思路。然而RFS多目标跟踪理论还处在发展阶段，RFS框架下多源多目标融合跟踪还是一个新问题，水下异类传感器信息的差异性及水声信道延迟的特殊性增加了问题的难度，使得基于RFS的多传感器多目标融合跟踪面临挑战。.本项目针对水下多平台协同探测中多目标融合跟踪的难题，研究多源多目标不确定信息RFS建模与融合跟踪的科学问题，探索多目标融合跟踪的新理论、新方法，提升水下多传感器多目标不确定信息统一表征与融合跟踪能力。主要研究内容有：①多源多目标不确定信息的RFS建模与滤波；②RFS框架下异类传感器信息融合与多目标跟踪；③RFS框架下跨平台传感器非同步信息融合与多目标跟踪；④理论、机理和方法的消声水池模拟实验验证。

Multi-target fusion tracking is one of the key problems to be solved in underwater cooperative detection system composed of unmanned underwater vehicle (UUV), submarine and other underwater platforms, and it is also a research hotpot and an important development direction in the field of information fusion. Multi-target tracking is originally a complex problem, which is due to the uncertainty of maneuvering target state and sensor measurements in the multi-target scene. Fortunately, random finite set (RFS) based on finite set statistics (FISST), with its strong multi-dimensional uncertain information representation and processing ability, provides a new way of thinking for multi-target tracking research, which has received a great deal of attention in the field of multi-target tracking. However, the theory of random finite set multi-target tracking and its applications are still in the development stage, and many practical problems need to be solved. Multisource-multitarget fusion tracking based on RFS is still a very new topic, and there is a lack of mature schemes and effective fusion algorithms. Moreover, the difference of underwater heterogeneous sensor information and the particularity of underwater acoustic signal propagation delay increase the complexity and difficulty of the problem. These problems make multisource-multitarget fusion tracking based on RFS serious challenge..Aiming at the difficulty of multi-target fusion tracking in underwater multi-platform cooperative detection, this project intends to research the scientific problem about random finite set model of underwater multisource-multitarget uncertain information and fusion tracking. Its aim is to explore the new theory and methods of multisource-multitarget fusion tracking, so as to improve the unified representation ability of underwater multi-sensor and multi-target uncertain information and fusion tracking ability. The major research contents are as follows: ①RFS modeling of multisource-multitarget uncertain information and Bayesian filtering algorithm based on RFS. ②Heterogeneous sensor information fusion and multisource-multitarget tracking under RFS framework. ③Cross-platform sensor asynchronous information fusion and multisource-multitarget tracking under RFS framework. ④Design of anechoic water tank experiment system and experimental verification of theory, mechanism and methods.