Research On Resource Management Algorithm For Multi-UAV SAR System

Multi-base synthetic aperture radar(SAR)is one of the focuses of current SAR research.Multi-base SAR has good target identification capability due to its transceiver and transmitter splitting and diverse configurations.Multi-UAV platforms are suitable for carrying multi-base SAR systems because of their robustness and deployment flexibility.Multi-base SAR systems mounted on multi-UAV platforms can perform multiple imaging missions and acquire more target information in complex environments.In order to make the multi-UAV SAR system collect and process the required data effectively and obtain good imaging performance,its resource management is required.The resources of the multi-UAV SAR system include UAV spatial configuration,power,bandwidth,number of UAVs,etc.By reasonably allocating these resources and determining the mission execution sequence and path,the rational utilization of the resources of the multi-UAV SAR system is achieved,and the system performance is integrated and optimized.In this thesis,theoretical analysis,methodological research and simulation validation of the imaging requirements in a multi-UAV SAR system and the assignment and path planning of multiple tasks in a multitasking scenario are carried out as follows:1.To establish a model of multi-UAV SAR system,analyze the imaging characteristics of multi-UAV SAR,and lay the foundation for the subsequent research of resource management algorithm of multi-UAV SAR system.2.A resource allocation method for multi-UAV SAR systems is proposed.The re-source allocation problem in resource management is analyzed,and the problem is modeled as a multi-objective optimization problem with constraints.A multi-objective genetic algorithm is used to allocate the spatial configuration,power,and bandwidth of the UAVs in the system,so that the system imaging capability is inte-grated and optimal in three aspects: observation angle diversity,image quality,and resolution.3.A multi-UAV SAR system task allocation and path planning method is proposed.A discrete ion swarm algorithm is used to assign the number of UAVs and mission execution sequences,and a reinforcement learning-based path planning algorithm is proposed to perform path planning for the UAVs completed by the prior sequence as-signment.The method enables the system to accomplish multiple imaging missions in complex terrain with the combined optimal system fuel consumption,operation time and imaging performance.The above models and algorithms are validated by simulation,and can realize the resource management of multi-UAV SAR system and accomplish multi-UAV SAR imaging tasks.