XYXusheng Yang

Xusheng Yang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xusheng Yang has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Edge Resilient Control for Feeder Automation System Under Hybrid Attacks: A Hybrid Physical Model-Driven and Data-Driven Approach

    by Rong Zeng, Yong Li, Sijia Hu, Xia Shao, Yong Xu, Xusheng Yang, Yijia Cao

    The extensive application of the distributed feeder automation system (DFAs) is a prominent feature of the power distribution system (PDS). To enhance the resilience of DFAs to cyberattacks and hybrid attacks (contains both physical and cyber-attacks), this paper proposes an edge resilient control approach (ERCA) for DFAs, which integrates attack detection and mitigation. Two key modules are developed: 1) Edge Cyberattack Detection and Correction Module (ECDCM), achieving quick cyberattack detection and correction based on machine learning and physical constraints; 2) Consensus-based Fault Section Localization Module (CFSLM), which incorporates causality function related to agent overcurrent states and fault location to correctly localize the fault section in hybrid attack scenarios. In addition, the effective boundaries of ERCA are theoretically deduced in this paper, which proves that ERCA can retain resilience to cyberattacks and hybrid attacks even if undetected cyberattacks exist due to the failures of machine learning-based detection submodules. Finally, the proposal’s effectiveness and performance are tested and verified using the IEEE 123-bus system and the real two-feeder test platform with DFAs. By adopting ERCA, outages and load losses caused by cyber attacks are 100% avoided, and the fault caused by physical attack is correctly isolated with more than 97.6% reliability in 100,000 times-level Monte Carlo Numerical simulations of hybrid attacks.


  • Bi-Level Resilient Control Solution for Distributed Feeder Automation System Under Hybrid Attack

    by Rong Zeng, Yong Li, Sijia Hu, Jiayan Liu, Huidi Wu, Jie Chen, Yong Xu, Xusheng Yang, Yijia Cao

    Distributed feeder automation system (DFAs), as a promising protection technology for power distribution system (PDS) with distributed generation, its vulnerability to cyberattacks and hybrid attacks (contains both physical and cyberattacks) is gradually recognized and haunts utilities, creating potential risks for its large-scale applications. This paper proposes a novel bilevel resilient control solution (BRCS) deployed to DFAs without hardware burden. Two key modules are developed: 1) Lightweight distributed cyberattack detection module (DCDM), deployed into agents of DFAs, based on unsupervised learning to realize the quick detection and reporting of cyberattacks; 2) Robust centralized fault section localization module (CFSLM), installed in DFAs’ host workstation located in control center, achieving the correctly fault section localization and the high dimensionally awareness of attack events in cyberattack and hybrid attack scenarios. By adopting BRCS, outages and load losses caused by cyberattacks can be 100% avoided, and faults caused by physical attacks can be correctly isolated at once. Finally, the effectiveness and performance of the proposal are verified and conducted by the real two-feeder test platform with DFAs. In this process, a digital high-dimensional awareness and control unit is created against cyber and hybrid attacks, contributing to the system-level application of risk management and resilient control.


  • A Load Forecasting Framework Considering Hybrid Ensemble Deep Learning With Two-Stage Load Decomposition

    by Sisi Zhou, Yong Li, Yixiu Guo, Xusheng Yang, Mohammad Shahidehpour, Wei Deng, Yujie Mei, Lei Ren, Yi Liu, Tong Kang, et al.

    Accurate load forecasting is essential for the operational management of distribution network. This paper proposes a hybrid ensemble deep learning (HEDL) load forecasting framework. Two-stage load decomposition is based on multiple seasonal-trend decomposition using loess (MSTL) and variational mode decomposition (VMD), which simplifies the observed load sequence with complex variation patterns on multi-timescales and multi-frequencies. Multidimensional feature matrices considering component variability are constructed based on the maximum information coefficient (MIC) and the sliding window. The component forecasting models based on temporal convolutional network (TCN) are built, which fully capture the long-term and short-term dependencies of loads through distinct receptive fields. A weighted ensemble method is utilized to obtain more accurate forecasting results for residual components with high uncertainties. The final load forecasting result is then ensembled by summing up the multiple component forecasting results. The results of case studies demonstrate the effectiveness of the proposed HEDL. Comparative experiments with five benchmark models and two advanced frameworks have verified the superiority of the proposed HEDL framework in load forecasting accuracy, generalizability and robustness.