基于信息决策树分支剔除的传感器资源调度

针对低轨星座对高速运动目标连续跟踪的传感器调度问题,将传感器调度的长时信息增量模型描述为信息决策树,并通过分支剔除技术搜索最优解,提出一种基于信息决策树分支剔除的传感器调度方法.仿真实验表明,长时信息增量有效克服了短时信息增量过于频繁调度的问题,且决策树分支剔除的引入大大降低了长时信息增量最优搜索的运算量.

Sensor Resource Scheduling Based on the Pruning of Information Decision Tree

To deal with the sensor scheduling in high-speed object tracking of low earth orbit constellation, the non-myopic information increment was described as an information decision tree. Then, the optimal solution was searched with pruning technology, and a method based on the pruning of information decision tree was proposed. The simulation results have indicated that the excessive scheduling is reduced effectively with non-myopic information increment. Furthermore, the introduction of pruning technology has greatly reduced computation burden.