基于空调自适应修正模型的户用微电网能量优化

空调负荷作为居民用户中的主要负荷,成为户用微电网能量管理的重要对象。以空调负荷为研究重点,结合户用微电网中其他可控负荷及储能设备,提出一种计及空调模型自适应修正的户用微电网能量管理策略。针对空调所属建筑物热力学模型中空调能效比系数、建筑物参数不易获取的问题,提出根据空调在当前场景下运行的历史数据,用遗传算法拟合得到适应当前场景的更为准确的热力学模型参数。针对室内人员活动、其他冷热型负荷工作状态改变等环境动态变化因素导致室内温度偏离优化温度的问题,用Q学习算法对空调设置温度进行在线自适应修正。最后,通过算例和实验证明了所提策略可以适应不同场景和环境的动态变化,提高了能量管理的适用性和准确性。

Energy Optimization of Household Microgrid Based on Adaptive Adjustment Model of Air Conditioning

As the major load of residential electricity system, air conditioning has been of great significance for the energy management of household microgrid because of its thermal storage ability. With air conditioning load as the focal point and an eye to other controllable loads and energy storage system, an energy management strategy considering the adaptive adjustment model of air conditioning is proposed. In view of the difficulty in obtaining coefficients of performance of air conditioning and the parameters of the building, it is proposed to use the historical data on air conditioning operation in current scenario to obtain more accurate fitting method using the genetic algorithm to get more accurate parameters. Besides, the human activity and the effect of other loads may lead to the indoor temperature deviate from the reference temperature. Thus a Q-learning algorithm is proposed to adjust the set temperature of air conditioning online to improve the accuracy of energy management. Finally, the cases in point verify the correctness and effectiveness of the proposed household energy management strategy.