Wenbo Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Wenbo Li has a total of 61 co-authors. Here are the top co-authors by number of shared publications:
- Chengrui Liu
Beijing Institute of Control Engineering, Beijing 100094, China
4 shared publications Hanyu Liang
National Key Laboratory of Science and Technology on Space Intelligent Control, Beijing Institute of Control Engineering, Beijing, China
2 shared publicationsJinxiu Zhang
2 shared publicationsPing Liu
2 shared publications- Wenchao Xue
State Key Laboratory of Mathematical Sciences, Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China
2 shared publications - Wenjing Liu
National Key Laboratory of Space Intelligent Control, Beijing Institute of Control Engineering, Beijing 100094, China
2 shared publications
Latest Publications
- Open Access
- 12 October 2026
- Published by Elsevier in SSRN Electronic Journal
Reducing the Impacts of Incomplete Data in Anomaly Detection for Time Series – A Model-Ensemble-Based Method
Anomaly detection in time series is critical in many applications, yet practical performance is limited because monitoring data are frequently incomplete: packets are dropped, buffers overflow, and sensors are switched off in certain operating modes. Missing values break the temporal and inter-metric dependencies on which data-driven detectors rely, and most existing methods impute the gaps before detection. Imputation relies on missingness assumptions that often fail in edge and industrial settings, and the substituted values can distort the relationships a detector has learned. We propose GMDE (graph-guided missing-pattern deletion and ensemble), which avoids imputation. A graph neural network trained on the complete portion of the data produces a dependency graph at every timestamp; missing metric sets whose onset changes the graph most are compressed by frequent-itemset mining into a few patterns. For each pattern we delete the affected metrics, train a sparse autoencoder on the remaining subset, and use it at detection time only when that subset is admissible. Scores are fused with weights proportional to the number of retained metrics; our analysis shows that deterministic imputation introduces a distortion that persists even for an optimal imputer, whereas targeted deletion preserves the conditional structure of the retained metrics, and that the fusion is a minimum-variance combination of the per-model scores. Experiments on five real-world datasets show that GMDE outperforms fifteen impute-then-detect combinations and a jointly optimized imputation-detection model in most configurations, and the advantage grows with the missing rate.
- 12 October 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessReducing the Impacts of Incomplete Data in Anomaly Detection for Time Series – A Model-Ensemble-Based Method
Anomaly detection in time series is critical in many applications, yet practical performance is limited because monitoring data are frequently incomplete: packets are dropped, buffers overflow, and sensors are switched off in certain operating modes. Missing values break the temporal and inter-metric dependencies on which data-driven detectors rely, and most existing methods impute the gaps before detection. Imputation relies on missingness assumptions that often fail in edge and industrial settings, and the substituted values can distort the relationships a detector has learned. We propose GMDE (graph-guided missing-pattern deletion and ensemble), which avoids imputation. A graph neural network trained on the complete portion of the data produces a dependency graph at every timestamp; missing metric sets whose onset changes the graph most are compressed by frequent-itemset mining into a few patterns. For each pattern we delete the affected metrics, train a sparse autoencoder on the remaining subset, and use it at detection time only when that subset is admissible. Scores are fused with weights proportional to the number of retained metrics; our analysis shows that deterministic imputation introduces a distortion that persists even for an optimal imputer, whereas targeted deletion preserves the conditional structure of the retained metrics, and that the fusion is a minimum-variance combination of the per-model scores. Experiments on five real-world datasets show that GMDE outperforms fifteen impute-then-detect combinations and a jointly optimized imputation-detection model in most configurations, and the advantage grows with the missing rate.
- 6 July 2026
- Published by Elsevier in Aerospace Science and Technology
- Vol. 178
A lightweight model-data hybrid driven framework for fault diagnosis in spacecraft attitude control system
- 6 July 2026
- Published by Elsevier in Aerospace Science and Technology
- Vol. 178
A lightweight model-data hybrid driven framework for fault diagnosis in spacecraft attitude control system
- 26 February 2026
- Published by Elsevier in Aerospace Science and Technology
- Vol. 177
Few-shot anomaly detection for aerospace telemetry data based on model-agnostic meta-learning
- 26 February 2026
- Published by Elsevier in Aerospace Science and Technology
- Vol. 177
Few-shot anomaly detection for aerospace telemetry data based on model-agnostic meta-learning
- Open Access
Spacecraft Health Status Monitoring Method Based on Multidimensional Data Fusion
To address the difficulty of detecting on-orbit faults of spacecraft under complex operating conditions in time, rational monitoring and assessment of spacecraft health status are essential for ensuring its safe, stable, and reliable operation. Considering the complexity, coupling, and multidimensionality of telemetry data, this paper proposes a method for monitoring the health status of spacecraft based on multidimensional data fusion for a key electromechanical component of a spacecraft control system. The method first extracts the explicit and implicit features of the multidimensional coupled telemetry parameters via physical feature formulas and a stacked autoencoder. Then, the extracted features are fused and filtered to obtain the health factor—a performance degradation trend described the evolution law of key component health status over runtime. Moreover, the different degradation stages are identified via an unsupervised clustering algorithm. Finally, a Bidirectional Long Short-Term Memory (Bi-LSTM) is used to construct a health status prediction model in stages. By taking Control Moment Gyroscopes (CMGs) as experimental verification subjects, the proposed method demonstrates significantly superior performance compared to other methods across prediction accuracy metrics including MSE, RMSE, and . This study provides robust technical support for health status monitoring of key spacecraft electromechanical components under specific fault modes.Open AccessSpacecraft Health Status Monitoring Method Based on Multidimensional Data Fusion
To address the difficulty of detecting on-orbit faults of spacecraft under complex operating conditions in time, rational monitoring and assessment of spacecraft health status are essential for ensuring its safe, stable, and reliable operation. Considering the complexity, coupling, and multidimensionality of telemetry data, this paper proposes a method for monitoring the health status of spacecraft based on multidimensional data fusion for a key electromechanical component of a spacecraft control system. The method first extracts the explicit and implicit features of the multidimensional coupled telemetry parameters via physical feature formulas and a stacked autoencoder. Then, the extracted features are fused and filtered to obtain the health factor—a performance degradation trend described the evolution law of key component health status over runtime. Moreover, the different degradation stages are identified via an unsupervised clustering algorithm. Finally, a Bidirectional Long Short-Term Memory (Bi-LSTM) is used to construct a health status prediction model in stages. By taking Control Moment Gyroscopes (CMGs) as experimental verification subjects, the proposed method demonstrates significantly superior performance compared to other methods across prediction accuracy metrics including MSE, RMSE, and . This study provides robust technical support for health status monitoring of key spacecraft electromechanical components under specific fault modes. - 11 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 22 (15455955) , 23984-23995
Encoding–Decoding-Based Fault-Tolerant Consensus Control for Multi-Agent Systems With Markovian Switching Topologies and Logarithmic Quantizers
This paper focuses on the fault-tolerant consensus problem of multi-agent systems under an encoding-decoding framework with Markovian switching topologies. Due to the influence of complex environments, the communication network changes over time, and thus, a Markov chain is introduced to describe the random topology switching. Under the stochastic characteristics and encoding-decoding framework, Lyapunov’s second method is applied for the first time instead of the original matrix analysis method, with both uniform and logarithmic quantizers considered. To obtain the unmeasurable states and the unknown sensor faults in the form of first-order differences, augmented estimation is performed, followed by encoding-transmission-reception-decoding for controller design. Under uniform quantization, the relationship between the scaling function and the parameters in the Riccati inequality is established, enabling boundedness analysis of data transmission. Under logarithmic quantization, a quantization-dependent Lyapunov function is constructed, and its rationality is rigorously proven. Finally, numerical simulations under both uniform and logarithmic quantization are provided to demonstrate the achievement of consensus and poly-quadratically stable in mean square sense.
- 11 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 22 (15455955) , 23984-23995
Encoding–Decoding-Based Fault-Tolerant Consensus Control for Multi-Agent Systems With Markovian Switching Topologies and Logarithmic Quantizers
This paper focuses on the fault-tolerant consensus problem of multi-agent systems under an encoding-decoding framework with Markovian switching topologies. Due to the influence of complex environments, the communication network changes over time, and thus, a Markov chain is introduced to describe the random topology switching. Under the stochastic characteristics and encoding-decoding framework, Lyapunov’s second method is applied for the first time instead of the original matrix analysis method, with both uniform and logarithmic quantizers considered. To obtain the unmeasurable states and the unknown sensor faults in the form of first-order differences, augmented estimation is performed, followed by encoding-transmission-reception-decoding for controller design. Under uniform quantization, the relationship between the scaling function and the parameters in the Riccati inequality is established, enabling boundedness analysis of data transmission. Under logarithmic quantization, a quantization-dependent Lyapunov function is constructed, and its rationality is rigorously proven. Finally, numerical simulations under both uniform and logarithmic quantization are provided to demonstrate the achievement of consensus and poly-quadratically stable in mean square sense.
- 4 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cybernetics
- Vol. 55 (11) , 5286-5299
Detectability Driven Recommendation of Anomaly Detection Models for Time-Series Data
Anomaly detection for time-series data has been viewed widely in many practical applications and caused lots of research interests. A popular solution based on deep learning techniques is first building an anomaly detection model by offline learning on a specific training data and subsequently utilizing the model to detect anomalies in the online setting. On the one hand, previous works have introduced a plenty of learning methods for building anomaly detection models, but it is well known that no one can perform best in all settings. Therefore, there is usually high demand for collecting multiple detection models in practical applications. On the other hand, in view of the limited computing resources in many online anomaly detection applications, it is almost impossible to run multiple detection models simultaneously due to the high time cost. Then, a natural idea is to enhance the general anomaly detection procedure with an effective mechanism for selecting proper models to avoid the high cost caused by executing too many detection models. Previous works focusing on the recommendation either are time inefficient or usually show weak performance, suffering from the missing labels and heterogeneous data characteristics met in real applications. Therefore, it is highly needed to design effective recommendation methods for automatically choosing anomaly detection models. Motivated by the technical challenges, a novel recommendation method for anomaly detection models is proposed in this article. First, a model recommendation framework based on the concept of detectability is introduced, where the detectability of an anomaly detection model is defined using a fine-grained strategy for comparing data characteristics. Then, based on efficient techniques for computing detectability, an efficient model recommendation algorithm is designed. Finally, extensive experimental results are produced on real time series and typical anomaly detection methods, and show that the proposed method is both effective and efficient.
- 4 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cybernetics
- Vol. 55 (11) , 5286-5299
Detectability Driven Recommendation of Anomaly Detection Models for Time-Series Data
Anomaly detection for time-series data has been viewed widely in many practical applications and caused lots of research interests. A popular solution based on deep learning techniques is first building an anomaly detection model by offline learning on a specific training data and subsequently utilizing the model to detect anomalies in the online setting. On the one hand, previous works have introduced a plenty of learning methods for building anomaly detection models, but it is well known that no one can perform best in all settings. Therefore, there is usually high demand for collecting multiple detection models in practical applications. On the other hand, in view of the limited computing resources in many online anomaly detection applications, it is almost impossible to run multiple detection models simultaneously due to the high time cost. Then, a natural idea is to enhance the general anomaly detection procedure with an effective mechanism for selecting proper models to avoid the high cost caused by executing too many detection models. Previous works focusing on the recommendation either are time inefficient or usually show weak performance, suffering from the missing labels and heterogeneous data characteristics met in real applications. Therefore, it is highly needed to design effective recommendation methods for automatically choosing anomaly detection models. Motivated by the technical challenges, a novel recommendation method for anomaly detection models is proposed in this article. First, a model recommendation framework based on the concept of detectability is introduced, where the detectability of an anomaly detection model is defined using a fine-grained strategy for comparing data characteristics. Then, based on efficient techniques for computing detectability, an efficient model recommendation algorithm is designed. Finally, extensive experimental results are produced on real time series and typical anomaly detection methods, and show that the proposed method is both effective and efficient.
- 7 February 2025
- Published by Springer Nature in Archive of Applied Mechanics
- Vol. 95 (2) , 1-20
Modeling method for boundary conditions of the rectangular thin plate by using the power series polynomial
- 7 February 2025
- Published by Springer Nature in Archive of Applied Mechanics
- Vol. 95 (2) , 1-20
Modeling method for boundary conditions of the rectangular thin plate by using the power series polynomial
- 8 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 26 (3) , 3402-3416
Novel Finite-Time Controller With Improved Auxiliary Adaptive Law for Hypersonic Vehicle Subject to Actuator Constraints
A novel adaptive finite-time controller (NAFTC) is proposed for flexible air-breathing hypersonic vehicle (FAHV) with actuator saturations, composing of two controllers designed for velocity and height subsystem respectively. Firstly, an adaptive dynamic inversion control is presented for velocity subsystem. The influence of actuator saturation is solved by an improved auxiliary adaptive law (IAAL). Compared with conventional adaptive law, the IAAL can achieve faster convergent speed of tracking error and weaken dramatical change for control signal effectively. Secondly, an adaptive continuous sliding mode control is designed for height subsystem, in which integral sliding surface is established based on a continuous fast higher-order sliding mode algorithm (CFSMA). Compared with conventional finite-time high-order regulator, CFSMA can drive states to converge faster and adjust respond speed of system conveniently without complicated parameters selection. Meanwhile, IAAL is combined to deal with the influence of elevator saturation. Ultimately, with the aid of NAFTC, FAHV subject to actuator constraints can also achieve faster convergent speed. In addition, NAFTC can realize higher tracking precision and faster respond speed compared with existing conventional adaptive controllers.
- 8 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 26 (3) , 3402-3416
Novel Finite-Time Controller With Improved Auxiliary Adaptive Law for Hypersonic Vehicle Subject to Actuator Constraints
A novel adaptive finite-time controller (NAFTC) is proposed for flexible air-breathing hypersonic vehicle (FAHV) with actuator saturations, composing of two controllers designed for velocity and height subsystem respectively. Firstly, an adaptive dynamic inversion control is presented for velocity subsystem. The influence of actuator saturation is solved by an improved auxiliary adaptive law (IAAL). Compared with conventional adaptive law, the IAAL can achieve faster convergent speed of tracking error and weaken dramatical change for control signal effectively. Secondly, an adaptive continuous sliding mode control is designed for height subsystem, in which integral sliding surface is established based on a continuous fast higher-order sliding mode algorithm (CFSMA). Compared with conventional finite-time high-order regulator, CFSMA can drive states to converge faster and adjust respond speed of system conveniently without complicated parameters selection. Meanwhile, IAAL is combined to deal with the influence of elevator saturation. Ultimately, with the aid of NAFTC, FAHV subject to actuator constraints can also achieve faster convergent speed. In addition, NAFTC can realize higher tracking precision and faster respond speed compared with existing conventional adaptive controllers.