Power Data Credible Decision-Making Mechanism Based on Federated Learning and Blockchain

【Author】 Li, Xin; Shang, Fangjian; Yao, Yanli; Zheng, Tianren

【Source】GAME THEORY FOR NETWORKS, GAMENETS 2022

【影响因子】

【Abstract】In modern power systems, it is an important issue to process and analyze power big data and perform reliable decision-making analysis. In response to this problem, this paper proposes a distributed computing architecture for power data based on a consortium chain, which realizes distributed and trusted shared training computing for power data while taking into account the privacy protection of the original data. To solve the problem of sample imbalance, this paper proposes a data balancing method combining SMOTE algorithm and the k-means algorithm. This paper also proposes an LSTM neural network load forecasting method based on federated learning and proves that it has higher accuracy and applicability than traditional methods through examples.

【Keywords】Power data; Federated learning; Blockchain; LSTM

【发表时间】2022

【收录时间】2023-05-31

【文献类型】实验仿真

【主题类别】

区块链应用-实体经济-电力领域