Haigang Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Haigang Liu has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Chu Sun
School of Electrical Engineering and Computer Science, KTH Royal Institute of Technology, Stockholm, Sweden
2 shared publications - Na Wang
School of Automation Science and Electrical Engineering, Beihang University, Beijing, China
2 shared publications - David B. Hitchcock
Department of Statistics, University of South Carolina, Columbia, USA
1 shared publication - Diyin Tang
School of Automation Science and Electrical Engineering, Beihang University, Beijing, China
1 shared publication - Jian Zhang
School of Management Science and Engineering, Beijing Information Science and Technology University, Beijing 102206, China
1 shared publication - Jie Yang
The Hangzhou International Innovation Institute, Beihang University, Hangzhou 311115, China
1 shared publication
Latest Publications
- 3 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 73 (1) , 107-119
Explaining Anomalous Events in Flight Data of UAV With Deep Attention-Based Multi-Instance Learning
Ensuring the safety and reliability of unmanned aerial vehicles (UAVs) has become a critical issue as they continue to advance. Analyzing anomalous events using event-based explanations is an effective approach to identifying key anomalous behaviors and mitigating potential risks. However, this task is challenging because anomalous events in flight data lack time-step labels in real-world UAV flight scenarios. To address this challenge, we propose a dual attention-based multi-instance learning (DA-DI-MIL) for pinpointing anomaly instances and automatically explaining anomalous events. Our MIL-based framework, from a weakly supervised learning perspective, treats a segment of flight data as a ‘bag’ with an available label, and its time steps as ‘instances’ without labels. The dual attention mechanism in DA-DI-MIL combines a temporal pseudo-label, predicted by temporal attention, and sensor variable importance obtained from delta feature attention to better pinpoint anomalies from instance-level labels. We use these labels in our proposed MIL-based framework to establish the relationships between anomalous events and anomalous behaviors. We conducted extensive experiments on real UAV flight data with engine failures to demonstrate the effectiveness and robustness of our proposed method compared to existing state-of-the-art methods. DA-DI-MIL achieves near 90% in evaluation metrics with less than 5 time steps' delay of anomalous event detection and around 85% in temporal anomaly detection. Additionally, we present an illustration of both global and local interpretations in the time domain and feature space, providing comprehensive insights into anomalous events. Our proposed method has efficiently and automatically explained anomalous events in UAV flight data, contributing to high-level aviation safety and reliability.
- 3 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 73 (1) , 107-119
Explaining Anomalous Events in Flight Data of UAV With Deep Attention-Based Multi-Instance Learning
Ensuring the safety and reliability of unmanned aerial vehicles (UAVs) has become a critical issue as they continue to advance. Analyzing anomalous events using event-based explanations is an effective approach to identifying key anomalous behaviors and mitigating potential risks. However, this task is challenging because anomalous events in flight data lack time-step labels in real-world UAV flight scenarios. To address this challenge, we propose a dual attention-based multi-instance learning (DA-DI-MIL) for pinpointing anomaly instances and automatically explaining anomalous events. Our MIL-based framework, from a weakly supervised learning perspective, treats a segment of flight data as a ‘bag’ with an available label, and its time steps as ‘instances’ without labels. The dual attention mechanism in DA-DI-MIL combines a temporal pseudo-label, predicted by temporal attention, and sensor variable importance obtained from delta feature attention to better pinpoint anomalies from instance-level labels. We use these labels in our proposed MIL-based framework to establish the relationships between anomalous events and anomalous behaviors. We conducted extensive experiments on real UAV flight data with engine failures to demonstrate the effectiveness and robustness of our proposed method compared to existing state-of-the-art methods. DA-DI-MIL achieves near 90% in evaluation metrics with less than 5 time steps' delay of anomalous event detection and around 85% in temporal anomaly detection. Additionally, we present an illustration of both global and local interpretations in the time domain and feature space, providing comprehensive insights into anomalous events. Our proposed method has efficiently and automatically explained anomalous events in UAV flight data, contributing to high-level aviation safety and reliability.
- Open Access
- 10 July 2023
- Published by MDPI AG in Applied Sciences
- Vol. 13 (14) , 8059
Design of Energy Recovery Control for General Virtual Synchronous Machines Based on Various Forms of Energy Storage
The reduced inertia in the power system due to renewable energy integration introduces operation challenges in frequency stability and control. The current options for virtual inertia and frequency support are limited by the energy resources and the power electronic interface. Considering the demand on response speed and energy capacity, a general virtual synchronous machine (VSM) control based on various forms of energy storage systems (ESS) is proposed. The steady-state energy variation of energy storage is found to be proportional to the virtual damping or governor gain, while inversely proportional to the integral gain of system frequency control. It is found that the size of energy storage can be at the second time scale (for example, 6.8 p.u.·s) for VSM implementation, which is significantly smaller than the conventional hour-scale energy storage in the power system. Based on energy dynamic analysis, stability requirement, and bandwidth separation rules, an energy recovery control is designed to maintain constant state of charge (for example, 50%) while avoiding conflicts with frequency regulation. The time scale of the designed energy recovery control loop (for example, hundreds of seconds) is longer than the secondary frequency control. The effectiveness of the proposed control is verified through comprehensive case studies.
- 10 July 2023
- Published by MDPI AG in Applied Sciences
- Vol. 13 (14) , 8059
Open AccessDesign of Energy Recovery Control for General Virtual Synchronous Machines Based on Various Forms of Energy Storage
The reduced inertia in the power system due to renewable energy integration introduces operation challenges in frequency stability and control. The current options for virtual inertia and frequency support are limited by the energy resources and the power electronic interface. Considering the demand on response speed and energy capacity, a general virtual synchronous machine (VSM) control based on various forms of energy storage systems (ESS) is proposed. The steady-state energy variation of energy storage is found to be proportional to the virtual damping or governor gain, while inversely proportional to the integral gain of system frequency control. It is found that the size of energy storage can be at the second time scale (for example, 6.8 p.u.·s) for VSM implementation, which is significantly smaller than the conventional hour-scale energy storage in the power system. Based on energy dynamic analysis, stability requirement, and bandwidth separation rules, an energy recovery control is designed to maintain constant state of charge (for example, 50%) while avoiding conflicts with frequency regulation. The time scale of the designed energy recovery control loop (for example, hundreds of seconds) is longer than the secondary frequency control. The effectiveness of the proposed control is verified through comprehensive case studies.
- Open Access
Analysis of Synchronous Generator Self-Excitation under Capacitive Load Condition in Variable-Frequency Aviation Power System
As power electronic converters become more widely used in aviation power systems, the associated capacitive loads in the harmonic filter circuits increase accordingly. The risk of self-excitation of aeronautical synchronous generators due to capacitive loads is thus increased. Compared with the self-excitation of a generator in a conventional fixed-frequency power system, this process is more complicated in a variable-frequency aviation power supply (360–800 Hz), as both the varied frequency and the loading conditions contribute to the self-excitation. To quantify this effect, in our study, a series-parallel model of simplified RLC loads under a variable-frequency power supply was built. The criterion of generator self-excitation, given in terms of the generator impedance and the load impedance, was then derived. To facilitate the load configuration design in the case of an aviation power system, a comprehensive analysis of the influences of the varied load power and system frequency on the load impedance was conducted. A graphical approach was proposed to determine self-excitation by comparing the series load reactance and resistor with three critical impedances corresponding to three self-excitation criteria, which is more intuitive and enables one to visualize the tendency of self-excitation with varied frequencies and loading conditions more effectively. Finally, the influence of variable frequency on the self-excitation of the aeronautical synchronous generator was verified by the simulation results.
Open AccessAnalysis of Synchronous Generator Self-Excitation under Capacitive Load Condition in Variable-Frequency Aviation Power System
As power electronic converters become more widely used in aviation power systems, the associated capacitive loads in the harmonic filter circuits increase accordingly. The risk of self-excitation of aeronautical synchronous generators due to capacitive loads is thus increased. Compared with the self-excitation of a generator in a conventional fixed-frequency power system, this process is more complicated in a variable-frequency aviation power supply (360–800 Hz), as both the varied frequency and the loading conditions contribute to the self-excitation. To quantify this effect, in our study, a series-parallel model of simplified RLC loads under a variable-frequency power supply was built. The criterion of generator self-excitation, given in terms of the generator impedance and the load impedance, was then derived. To facilitate the load configuration design in the case of an aviation power system, a comprehensive analysis of the influences of the varied load power and system frequency on the load impedance was conducted. A graphical approach was proposed to determine self-excitation by comparing the series load reactance and resistor with three critical impedances corresponding to three self-excitation criteria, which is more intuitive and enables one to visualize the tendency of self-excitation with varied frequencies and loading conditions more effectively. Finally, the influence of variable frequency on the self-excitation of the aeronautical synchronous generator was verified by the simulation results.
- 29 June 2019
- Published by Springer Nature
- p. 31-50
Spatial and Spatio-Temporal Analysis of Precipitation Data from South Carolina
- 29 June 2019
- Published by Springer Nature
- p. 31-50
Spatial and Spatio-Temporal Analysis of Precipitation Data from South Carolina