Demand response-oriented virtual power plant evaluation based on AdaBoost and BP neural network
Dawei Hu, Hengyu Liu, Yidong Zhu, Jiazheng Sun, Zhe Zhang, Luyu Yang, Qihuitianbo Liu, Bo Yang
Abstract
As an advanced technology that efficiently aggregates and optimizes renewable energy, controllable loads, and energy storage systems, virtual power plants (VPPs) can effectively promote the green and low-carbon transformation of power systems. A comprehensive assessment of VPPs is important for VPP investment and operation. However, most of the existing evaluation methods focus on the reliability, economy, and mobilizability of VPPs. In this paper, in order to better address the characteristics of demand response-oriented VPPs, three aspects of VPP operation indices, new energy indices, and demand response indices are analyzed. In this manner, it is possible to meet the principles of the construction of VPP evaluation system and also to measure the effect of demand response of the VPP. Then, on the basis of AdaBoost algorithm, combined with back propagation (BP) neural network for the evaluation and classification of demand response-oriented VPPs. The entropy value method and gray correlation are also compared to validate the superiority of the proposed method.