基于事件触发的直流微电网无差拍预测控制

针对光伏(Photovoltaic, PV)−电池−超级电容直流微电网系统中光伏发电间歇性造成的功率失配问题, 提出一种基于事件触发的无差拍预测控制(Event-triggered deadbeat predictive control, ETDPC)方法, 以实现有效的能量管理. ETDPC方法结合事件触发控制策略和无差拍预测控制策略(Deadbeat predictive control, DPC)的优点, 根据微电网的拓扑结构构建状态空间模型, 用于设计适用于微电网能量管理的触发条件: 当ETDPC的触发条件满足时, ETDPC中无差拍预测控制模块被激活, 可以在一个控制周期内产生最优控制信号, 实现对于扰动的快速响应, 减小母线电压纹波; 当系统状态不满足ETDPC中的触发条件时, 无差拍预测控制模块被挂起, 从而消除非必要运算, 以减轻实现能量管理的运算负担. 因此, 对于电池−超级电容器混合储能系统(Hybrid energy storage system, HESS), ETDPC能够缓解间歇性光伏发电与负荷需求之间的功率失衡, 以稳定母线电压. 最后, 数字仿真和硬件在环(Hardware-in-loop, HIL)实验结果表明, 相较于传统无差拍控制方法, 运算负担减小了50.63%, 母线电压纹波小于0.73%, 验证了ETDPC方法的有效性与性能优势, 为直流微电网的能量管理提供了一种参考.

Event-triggered Deadbeat Predictive Control for DC Microgrid

This paper presents an event-triggered deadbeat predictive control (ETDPC) method for the mitigation of power mismatch in a photovoltaic (PV)-battery-supercapacitor microgrid. The proposed ETDPC method combines the event-triggered control strategy and the deadbeat predictive control (DPC) strategy and inherits their advantages accordingly. Based on the topology of the DC microgrid, the state-space model can be built for the design of the triggering condition for the energy management: When the triggering condition of ETDPC is activated, the deadbeat control block of ETDPC will be conducted and the optimal control signal can be generated within one control cycle, so that the DC bus voltage ripple can be reduced based on the fast response to the disturbance; When the state of the DC microgrid cannot satisfy the triggering condition, the deadbeat control block of ETDPC will be suspended to eliminate the redundant computations, so that the computational burden of the DC microgrid energy management can be reduced. Therefore, ETDPC can be fully utilized for battery-supercapacitor hybrid energy storage system (HESS) to mitigate the power unbalance between the load demand and the intermittent photovoltaic power generation and stabilize the bus voltage. To validate the effectiveness of the method, various simulations and hardware-in-loop (HIL) experiments are conducted based on a digital simulation system and the HIL platform, which show that the computational burden is reduced by 50.63% compared to the conventional deadbeat predictive control and the voltage ripple is regulated less than 0.73% of the reference. This work provides a reference of the control strategy for microgrid energy management.