基于深度强化学习技术的光伏–固体氧化物燃料电池混合能源系统多场景控制

Deep Reinforcement Learning-based Multiple Scenario Control Strategy for PV-SOFC Hybrid System

固体氧化物燃料电池(solidoxidefuelcell,SOFC)运行安静、清洁高效,具有广泛的应用前景。SOFC能够与其他发电装置和储能装置灵活耦合形成混合能源系统,该文所研究的光伏-燃料电池(photovoltaic-solidoxidefuelcell,PV-SOFC)混合能源系统将太阳能通过电转气技术(powerto gas,P2G)转换为氢能供SOFC使用,能够有效弥补太阳能波动性大、间歇性强的缺点。针对SOFC混合能源系统多变量、强非线性的特征,该文采用深度强化学习(deep reinforcement learning,DRL)技术研究一种智能运行控制策略。首先,搭建PV-SOFC混合能源系统的仿真模型,该模型以SOFC为核心,侧重于系统中的功率流向和氢气流向,并考虑了系统内部的运行条件约束。其次,分别提出联网运行模式与孤岛运行模式下基于深度确定性策略梯度算法(deep deterministic policy gradient,DDPG)的系统智能运行控制策略。最后,通过算例仿真验证该算法的有效性。结果表明,孤岛运行模式下DDPG算法能够在不同环境和负荷需求条件下使混合能源系统保持较低的系统失电概率和弃电概率,有效提高系统的运行可靠性和经济性。联网运行模式下DDPG算法能够利用不同时间段的电价差和剩余储氢与电网灵活交易,从而获取更高的系统长期收益。

Solid oxide fuel cell(SOFC)has a wide application prospect because of its quiet operation,clean output and high efficiency.SOFC can also be flexibly coupled with other power generation devices and energy storage devices to form a hybrid energy system.The photovoltaic fuel cell(PV-SOFC)hybrid energy system studied in this paper converts solar energy into hydrogen for SOFC through power-to-gas(P2G)technology,which could effectively make up for the disadvantages of solar energy fluctuation and intermittency.To avoid the multivariable and strong nonlinear characteristics of SOFC hybrid energy system,this paper used deep reinforcement learning(DRL)technology to study its operation and control strategy.Firstly,this paper built a simulation model of PV-SOFC hybrid energy system environment,which took SOFC as the core,focusing on the flow direction of power and hydrogen in the system,and considered the constraints of internal operation conditions.Secondly,this paper proposed the system operation control strategies based on deep deterministic policy gradient(DDPG)algorithm in the grid-connected operation mode and the island operation mode respectively.Finally,the effectiveness of the proposed algorithm was demonstrated and verified by several simulation cases.The results showed that the DDPG algorithm could keep low probability of power loss and power abandonment of the hybrid energy system under different environment and load requirements,improving the system reliability and economic efficiency.Under the grid-connected operation mode,DDPG algorithm could make use of the electricity price difference in different time periods and the flexible transaction of surplus hydrogen storage with the power grid,so as to obtain higher long-term benefits of the system.