微型电脑应用

在应用屋顶光伏时会存在内部不稳定情况，因此，研究基于元学习的源网储融合型屋顶光伏能量双向交互控制模型来解决该问题。利用多元学习方法构建负荷预测模型，利用该模型实现设备层中源、网、储等多个控制器的负荷预测，经微电网层传达负荷预测结果并以此为依据控制配电网主网和微电网，能量交换层接收该控制结果，处理后制定最优控制方案重新传输至微电网层与设备层光伏能量实现交互控制。模型中的能量交换器使用边缘计算框架采集与分析控制数据，将这些分析结果作为多工况下能量交换器控制策略的依据实现光伏能量双向交互控制。试验结果表明，该控制模型能够准确预测负荷变化，同时能够多工况下控制功率变化，具有较为良好的控制效果。

There is internal instability in the practical application of rooftop photovoltaic. Therefore, a bidirectional interactive control model for rooftop photovoltaic energy of source network storage integration is studied based on source network and meta-learning to solve this problem. The load prediction model is constructed by the multivariate learning method, and the load prediction of multiple controllers such as source, network and storage in the equipment layer is realized by the model. The load prediction results are conveyed through the microgrid layer and used as the basis to control the main network and microgrid of the distribution network. The energy exchange layer receives the control results. After processing, the optimal control scheme is formulated and then transmitted to the photovoltaic energy of the microgrid layer and the equipment layer for interactive control. The energy exchanger in the model uses the edge computing framework to collect and analyze the control data, and these analysis results are used as the basis for the control strategy of the energy exchanger under multiple working conditions to realize the bidirectional interactive control of photovoltaic energy. The test results show that the control model can accurately predict the load variation and control the power variation under multiple working conditions, and has a relatively good control effect.