Energy Usage Identification Method of Integrated Energy System Based on Edge-Cloud Coordination

Integrated energy system (IES) is an import approach to improve energy utilization efficiency and promote clean energy transformation. In the practice of IES, energy usage identification after integration is an enabling technology that is able to further promote large-scale usage of IES and its refined operation. Existing methods are unable to fully utilize high-frequency sampled data from electrical systems and extract characteristics, which prevent the wide application of IES. In this work, a cloud-edge coordinated and V-I characteristic curve based energy usage identification for key component of IES is proposed. This approach follows the framework of cloud-edge coordination to achieve non-intrusive load monitoring. On the cloud side, V-I curves are processed and stored so as to generate a characteristic library. On the edge side, Adaboost algorithm from deep learning is employed to perform matching using local high-frequency sampled data stream, in order to achieve accurate identification of energy usage in IES. The effectiveness of this work is validated using a series of experiment using residential load energy usage identification.