Renewable energy sources (RES) are increasingly being developed and used to address the energy crisis and protect the environment. However, the large-scale integration of wind and solar energy into the power grid is still challenging and limits the adoption of these new energy sources. Microgrids (MGs) are small-scale power generation and distribution systems that can effectively integrate renewable energy, electric loads, and energy storage systems (ESS). By using MGs, it is possible to consume renewable energy locally and reduce energy losses from long-distance transmission. This paper proposes a deep reinforcement learning (DRL)-based energy management system (EMS) called DRL-MG to process and schedule energy purchase requests from customers in real-time. Specifically, the aim of this paper is to enhance the quality of service (QoS) for customers and reduce their electricity costs by proposing an approach that utilizes a Deep Q-learning Network (DQN) model. The experimental results indicate that the proposed method outperforms commonly used real-time scheduling methods significantly.
Background::
Data sharing can improve the utilization rate of distributed energy storage and solve the problem of data silos, but there are privacy and data security issues in distributed energy storage data sharing.
Objective::
To address the privacy and security issues in distributed energy storage data sharing.
Method::
In this paper, a distributed privacy-preserving data sharing scheme based on secure multi-party computing is proposed. Firstly, the requirements for distributed energy storage data sharing are analyzed, and the concept of distributed energy storage data sharing is defined. Secondly, the additive secret sharing technology is adopted to realize the safe transmission and storage of energy storage data between energy storage nodes. Finally, combined with the relevant forgetting transmission extension protocol, privacy-preserving data computing is realized, so as to solve the problem of secure sharing of distributed energy storage data.
Results::
The simulation results show that the privacy protection method does not affect the convergence round of specific communication. However, in order to achieve the effect of privacy protection, the number of communication times has become twice that of conventional communication, and the privacy protection communication time accounts for about 0.8 of the total communication time.
Conclusion::
Compared with the traditional data-sharing method, the proposed method realizes privacy-preserving data sharing without affecting algorithm convergence. For distributed scenarios, a flexible P2P sharing architecture is proposed, which has strong scalability and regret, and has good application prospects.
conclusion:
Experimental results show that compared with traditional data sharing methods, the proposed method has lower computational cost, achieves privacy-protecting data sharing, has strong scalability and deplorability, and good application prospects.
other:
No attention
The distributed and privacy-preserving characteristics of fine-grained
smart grid data hinder data sharing, making federated learning an
attractive approach for collaborative training among data owners with
similar load patterns. However, malicious models can interfere with
training in the federated learning aggregation process, making it
difficult to ensure the accuracy and safety of the central model in load
forecasting. Therefore, we propose a secure aggregation federated
learning method for distributed load forecasting based on similarity and
distance (Fed-SAD), which effectively eliminates the interference of
malicious models by securely aggregating models, thereby ensuring
accurate and safe distributed scenario prediction. Experimental results
demonstrate that Fed-SAD maintains high accuracy and robustness in both
the presence and absence of malicious models, while maintaining data and
model security.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.