论文部分内容阅读
多传感器多目标跟踪系统进行目标状态估计的数据关联问题可以阐述为广义S维分配问题。本文提出了一种基于模糊自适应GA的广义S维分配算法,该算法利用六个模糊控制器对符号编码遗传算法的遗传操作进行自适应控制,并将S维分配问题中的目标代价函数极小化问题作为组合优化问题进行求解,同时结合极大似然方法进行目标识别和目标状态估计。在考虑虚警和漏检前提下,对算法进行了稀疏目标和密集目标两种仿真环境下的Monte Carlo试验,对试验结果进行了对比分析。
The data association problem of multi-sensor multi-target tracking system for target state estimation can be described as a generalized S-dimensional distribution problem. In this paper, a generalized S-dimensional distribution algorithm based on fuzzy adaptive GA is proposed. The algorithm uses six fuzzy controllers to adaptively control the genetic operation of the symbol-coded genetic algorithm. The objective cost function pole The minimization problem is solved as a combinatorial optimization problem, and the target recognition and the target state estimation are combined with the maximum likelihood method. Under the premise of considering false alarm and missing detection, the Monte Carlo test under sparse and dense targets was carried out, and the test results were compared and analyzed.