基于多智能体的太阳能无人机能源控制

为满足太阳能无人机对长航时飞行和高载重能力的需求,研究能源系统的储能均衡控制问题.通过将太阳能无人机中每个由光伏-储能-输出单元组成的发电节点作为一个智能体,设计基于多智能体的分布式控制器并给出满足系统约束的控制算法,实现储能单元荷电状态的一致性.分别针对连续系统模型和离散系统模型给出分布式控制协议,并通过理论分析说明连续和离散的控制协议均可实现控制目标.通过搭建半实物平台进行实测验证,采用光伏模拟器和电子负载模拟能源系统运行的外部环境,以18650锂离子电池作为储能单元,实验结果表明,分布式协同控制协议能够有效地解决光伏功率不均及电池参数差异导致的不均衡问题,使系统的充放电深度得以有效提升.

Multi-agent based energy control for solar unmanned aerial vehicles

Considering the energy supply-demand of solar-powered unmanned aerial vehicles for long-distance flight, this paper investigates the storage balance control problem of the energy system. By taking each power generation in the vehicles as an agent, which is composed of photovoltaic cells, a storage system, and a power output unit, distributed control protocols based on the multi-agent theory are designed to achieve the state-of-charge balance of the energy storage unit, and a corresponded algorithm is proposed to satisfy the constraints of the system. Distributed control protocols are designed for continuous models and discrete models, respectively, and theoretical analysis shows that both the control protocols can achieve our objective. The effectiveness of the proposed control protocols is verified in the semi-physical platform, where the photovoltaic simulator and electronic load are used to establish the operational environment of the energy system, and the 18650 lithium-ion batteries are chosen as the storage unit. The results show that the proposed protocols can effectively solve the unbalance problem caused by the difference of photovoltaic power and battery parameters, and consequently improve the depth of charging/discharging.