Zhenwei Yu

Mr. Zhenwei Yu

orcid

Research Focus & Metrics

Academic Career Length

16Years (2011-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 November 2024 in IEEE Transactions on Smart Grid
Verified
Journal Article
Closed access
QLQiao Lin
Qiao Lin
Li DingLi Ding
Li Ding
Zhengmin KongZhengmin Kong
Zhengmin Kong
ZYZhenwei Yu
Zhenwei Yu
XLXin Li
Xin Li
HWHaijin Wang
Haijin Wang
The microgrid (MG) integrating clean renewable energy has increasingly emerged as a critical solution for addressing energy challenges and promoting sustainable energy development. However, the inherent uncertainties in renewable energy output and load consumption present significant challenges to the economic and stable operation of the MG. This paper investigates a multi-time model predictive control (MSMPC) strategy for the optimal scheduling of grid-connected MG. The proposed method can dynamically update the optimal scheduling of the MG on two-time scales based on real-time measurement data. The dispatchable thermostatically controlled loads (TCLs) are incorporated into the demand side management (DSM) system to enhance flexibility and satisfy the future trend of a larger proportion of controllable TCLs. Furthermore, the TCL model considers the aging problem associated with excessive compressor cycling and the satisfaction of end users. Simulation results demonstrate that the proposed method significantly improves the economic performance and robustness of the MG system. Moreover, a real-time experiment conducted using RT-LAB further verifies the feasibility of the proposed approach.

17 March 2025 in Energies
Verified
Journal Article
Open access
PDF file
YXYong Xiao
Yong Xiao
XJXin Jin
Xin Jin
TPTingzhe Pan
Tingzhe Pan
ZYZhenwei Yu
Zhenwei Yu
Li DingLi Ding
Li Ding
Accurate residential load forecasting plays a crucial role in optimizing demand-side resource integration and fulfilling the needs of demand-side response initiatives. To tackle challenges, such as data heterogeneity, constrained communication resources, and data security in smart grid load prediction, this study introduces a novel differential privacy federated learning algorithm. Leveraging the federated learning framework, the approach incorporates weather and temporal factors as key variables influencing load patterns, thereby creating a privacy-preserving load forecasting solution. The model is built upon the Long Short-Term Memory (LSTM) network architecture. Experimental results demonstrate that the proposed algorithm enables federated training without the need for sharing raw load data, facilitating load scheduling and energy management operations in smart grids while safeguarding user privacy. Furthermore, it exhibits superior prediction accuracy and communication efficiency compared to existing federated learning methods.

18 March 2024 in IEEE Systems Journal
Verified
Journal Article
Closed access
ZMZilong Mi
Zilong Mi
Zhengmin KongZhengmin Kong
Zhengmin Kong
Tao HuangTao Huang
Tao Huang
PSPeng Shi
Peng Shi
ZYZhenwei Yu
Zhenwei Yu
Li DingLi Ding
Li Ding
With the growing integration of unpredictable renewable energy sources into the grid, achieving power balance has become an increasingly crucial challenge. To address this challenge, demand response has emerged as a promising solution. This article proposes a new demand-side flexible thermostatically controlled loads response strategy framework. Our method employs a hierarchical control framework that covers three layers of control, which consist of the optimization layer, coordination layer, and local control layer. The optimization layer employs a dynamic average consensus algorithm for economic optimization scheduling to maximize the sum of the aggregators' welfare functions. In the coordination layer, power is distributed fairly based on the comfort state, generating reference signals for the local control layer. The local control layer tracks these reference signals and employs integral sliding mode control to suppress the influence of unknown disturbances. The control objectives of the entire framework can be achieved in a fixed time, and the parameters in the framework are heterogeneous. Furthermore, the relationships between controller parameters and tracking performance are derived, and the upper bounds of settling time are estimated. Finally, we demonstrate the validity of our theoretical results through numerical simulations.

14 August 2024 in IEEE Transactions on Smart Grid
Verified
Journal Article
Closed access
ZYZhenwei Yu
Zhenwei Yu
Li DingLi Ding
Li Ding
QLQiao Lin
Qiao Lin
SZShi-Rong Zhang
Shi-Rong Zhang
ZLZhiwei Liu
Zhiwei Liu
Thermostatically controlled loads (TCLs) obtain significant potential for providing auxiliary services to the power grid. Aggregators typically utilize centralized strategies when implementing direct load control (DLC) to manage TCLs, but these strategies consume substantial communication resources when dealing with large-scale TCLs. This paper proposes a distributed control strategy for ON/OFF state TCLs to alleviate the aggregator’s communication burden. The concept of temperature-based priority is introduced to ensure fair scheduling among TCLs. TCLs can determine their priority order in the cluster by communicating with neighboring TCLs and locally decide their operation states to achieve reference power tracking. A sub-priorities splitting method is introduced to further enhance power tracking accuracy. Furthermore, the proposed strategy utilizes an asynchronous communication mechanism in the coordination process, making it highly suitable for practical implementation. The study conducts numerical simulations and semi-physical experiments to validate the practical feasibility of the proposed strategy. The results conclusively demonstrate that the proposed strategy can effectively achieve accurate reference power tracking by the TCLs cluster.

22 January 2025 in IEEE Transactions on Systems, Man, and Cybernetics: Systems
Verified
Journal Article
Closed access
XLXin Li
Xin Li
Li DingLi Ding
Li Ding
YCYi-Ru Chen
Yi-Ru Chen
ZYZhenwei Yu
Zhenwei Yu
QLQiao Lin
Qiao Lin
Demand response (DR) programs are an effective means of mitigating power shortages. This article proposes a novel incentive-based DR program for peak shaving situations. The system operator (SO) determines the response power amount and provides tiered subsidy policies. The load aggregators (LAs) are rational decision-makers and formulate multiobjective optimization problems (MOPs) to make compromises for incentive income, user comfort, and environmental contribution. The innovation lies in the proposed distributed event-triggered solution methodology, including a distance-minimization algorithm to find the decision closest to the ideal point from the Pareto front, and a weighting coefficients optimization algorithm to allocate the importance of each objective at a predefined time. The distributed solution framework maintains the autonomy of each LA, and the event-triggered communication mechanism saves communication resources. Numerical simulations validate the effectiveness of the proposed solution methodology, including the realization of the response power target given by the SO, the optimal compromise of the MOP, and the saving of communication resources in the solution process.

11 October 2025 in ISA Transactions
Verified
Journal Article
Closed access
XLXin Li
Xin Li
Li DingLi Ding
Li Ding
ZYZhenwei Yu
Zhenwei Yu
Zhengmin KongZhengmin Kong
Zhengmin Kong
ZLZhiwei Liu
Zhiwei Liu