针对在实时动态条件下多UAV任务规划问题，提出了基于动态数据驱动的多UAV实时任务规划仿真平台，主要包括基于MultiUAV2的真实UAV群仿真平台和基于Multi-Agent的预测仿真平台两个部分。采用了A*算法对真实系统工作流进行探索，在此基础之上构建了多UAV合成工作流模型，然后针对动态数据注入运行仿真的问题，研究了传感器任务重置及传感器的预处理方法。最后，通过一个仿真实例验证了提出方法的可行性和有效性。%In order to achieve mission planning of Multi-UAV in the real-time and dynamic conditions,a Multi-UAV mission planning simulation system based on dynamic data driven application system(DDDAS)is proposed,mainly include real UAV swarm simulation platform based on MultiUAV2 and predictive simulation platform based on Multi-Agent. As a breakthrough point of the key technologies that addressed in the simulation system,the workflows within real-life system is discovered by using A* algorithm. Based on these analyses,a Multi-UAV workflows model is constructed,then the method of sensor re-tasking and sensor-based pre-processing are studied for injecting streaming sensor data into a running simulation. Finally,through a simulation example,the proposed method is feasibility and validity.
展开▼
