Zhongmei Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Zhongmei Li has a total of 120 co-authors. Here are the top co-authors by number of shared publications:
- Wenli Du
East China University of Science and Technology
16 shared publications - Rong Nie
Institute of Artificial Intelligence, School of Future Technology, Shanghai University, Shanghai, China
6 shared publications - Xin Peng
East China University of Science and Technology
6 shared publications - Shuping He
Anhui University
5 shared publications - Weihua Gui
Central South University
5 shared publications - Weimin Zhong
East China University of Science and Technology
5 shared publications
Latest Publications
- 13 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 23 (15455955) , 7880-7892
A Novel Likelihood Gradient-Based Incipient Fault Detection Approach for Avionics Systems
This paper presents a gradient-based fault detection method for pitch control systems in avionics. On the basis of the dynamic model of the airplane, the proposed method detects both operator and sensor faults by monitoring the online data. By integrating fault-related behaviors over an extended time window, the method effectively amplifies small changes caused by incipient faults, improving detectability. Theoretical analysis reveals that under normal flight conditions, the gradient has a zero expected value and a finite, analytically tractable variance. These characteristics make the method compatible with traditional fault detection approaches. Sufficient tests on real flight data verify its ability to detect hard-to-identify faults.
- 13 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 23 (15455955) , 7880-7892
A Novel Likelihood Gradient-Based Incipient Fault Detection Approach for Avionics Systems
This paper presents a gradient-based fault detection method for pitch control systems in avionics. On the basis of the dynamic model of the airplane, the proposed method detects both operator and sensor faults by monitoring the online data. By integrating fault-related behaviors over an extended time window, the method effectively amplifies small changes caused by incipient faults, improving detectability. Theoretical analysis reveals that under normal flight conditions, the gradient has a zero expected value and a finite, analytically tractable variance. These characteristics make the method compatible with traditional fault detection approaches. Sufficient tests on real flight data verify its ability to detect hard-to-identify faults.
- 10 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man, and Cybernetics: Systems
- Vol. 56 (4) , 2253-2261
Distributed Aggregative Optimization of MASs Subject to Coupled Inequality Constraints
This article investigates the distributed aggregation optimization problem in multiagent systems (MASs), with a particular focus on addressing the aggregation effect commonly encountered in modern engineering and technological applications. In such scenarios, the local objective function of an agent depends not only on its own decision variables but also interacts with the decision variables of other agents, resulting in complex coupling relationships. To solve these challenges while ensuring that the optimization variables satisfy the coupled inequality constraints, this article introduces a novel framework called distributed aggregative parameter projection (DAPP). Specifically, the proposed distributed protocol is based on an improved parameter projection, including two direction updates, which minimizes the cost function and keeps the search direction obeying the inequality at each iteration. In addition, the linear convergence performance of the proposed scheme over the undirected and connected graph is ensured by rigorous theoretical proof with mild assumptions. Finally, simulation results demonstrate the superior performance of DAPP with smaller global function and faster convergence speed in comparison to the existing method.
- 10 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man, and Cybernetics: Systems
- Vol. 56 (4) , 2253-2261
Distributed Aggregative Optimization of MASs Subject to Coupled Inequality Constraints
This article investigates the distributed aggregation optimization problem in multiagent systems (MASs), with a particular focus on addressing the aggregation effect commonly encountered in modern engineering and technological applications. In such scenarios, the local objective function of an agent depends not only on its own decision variables but also interacts with the decision variables of other agents, resulting in complex coupling relationships. To solve these challenges while ensuring that the optimization variables satisfy the coupled inequality constraints, this article introduces a novel framework called distributed aggregative parameter projection (DAPP). Specifically, the proposed distributed protocol is based on an improved parameter projection, including two direction updates, which minimizes the cost function and keeps the search direction obeying the inequality at each iteration. In addition, the linear convergence performance of the proposed scheme over the undirected and connected graph is ensured by rigorous theoretical proof with mild assumptions. Finally, simulation results demonstrate the superior performance of DAPP with smaller global function and faster convergence speed in comparison to the existing method.
- 9 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 1-16
Lightweight and Interpretable Channel Selection for EEG Measurement Systems in Epileptic Seizure Prediction
Utilizing multichannel electroencephalogram (EEG) signals to predict epileptic seizures is crucial for reducing patient distress. However, devices equipped with numerous electrodes are often cumbersome, expensive, and impractical for long-term or home-based monitoring. To address this challenge, we propose a novel EEG channel selection framework that combines LayerCAM with reinforcement learning (RL), aiming to enhance the accuracy and interpretability of epilepsy prediction models. In our framework, LayerCAM is integrated into a specially designed convolutional neural network (CNN) to accurately align class activation mapping with physical electrodes, thereby ensuring precise localization. This localization serves as the foundation for effective channel selection. Subsequently, we employ a transformer architecture termed decision transformer (DT) to treat channel selection as a sequence prediction task, and class activation mapping is utilized as a state within a Markov decision process (MDP). This strategy captures the dependencies between different channels and the decisions to activate or deactivate them, facilitating the learning of temporal and causal relationships. To validate our approach, we conducted extensive experiments on two datasets. The results demonstrated significant improvements in prediction accuracy across three distinct neural network architectures compared to utilizing all available EEG channels. In particular, our approach achieved an accuracy of 93.69% with eight channels on the CHB-MIT dataset and 95.41% with only five channels on the Renji dataset. These results demonstrate that reliable seizure prediction is achievable with simplified sensing configurations, thus enabling the design of portable and cost-effective EEG monitoring systems for practical use.
- 9 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 75 (00189456) , 1-16
Lightweight and Interpretable Channel Selection for EEG Measurement Systems in Epileptic Seizure Prediction
Utilizing multichannel electroencephalogram (EEG) signals to predict epileptic seizures is crucial for reducing patient distress. However, devices equipped with numerous electrodes are often cumbersome, expensive, and impractical for long-term or home-based monitoring. To address this challenge, we propose a novel EEG channel selection framework that combines LayerCAM with reinforcement learning (RL), aiming to enhance the accuracy and interpretability of epilepsy prediction models. In our framework, LayerCAM is integrated into a specially designed convolutional neural network (CNN) to accurately align class activation mapping with physical electrodes, thereby ensuring precise localization. This localization serves as the foundation for effective channel selection. Subsequently, we employ a transformer architecture termed decision transformer (DT) to treat channel selection as a sequence prediction task, and class activation mapping is utilized as a state within a Markov decision process (MDP). This strategy captures the dependencies between different channels and the decisions to activate or deactivate them, facilitating the learning of temporal and causal relationships. To validate our approach, we conducted extensive experiments on two datasets. The results demonstrated significant improvements in prediction accuracy across three distinct neural network architectures compared to utilizing all available EEG channels. In particular, our approach achieved an accuracy of 93.69% with eight channels on the CHB-MIT dataset and 95.41% with only five channels on the Renji dataset. These results demonstrate that reliable seizure prediction is achievable with simplified sensing configurations, thus enabling the design of portable and cost-effective EEG monitoring systems for practical use.
- 2 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 12, 199-211
Distributed Robust Output-Constrained Control of Unknown Nonlinear Multiagent Systems Under Switching Topologies
The leader-following problem for the heterogeneous nonlinear multiagent systems under output constraints is investigated in this paper. It is focused on the cases where the underlying communication graph is directed and switching, and each agent is a multi-input multi-output block-triangular system with totally unknown and inherently nonlinear dynamics. They render the existing distributed control solutions infeasible. To surmount this challenge, a novel distributed mixed control strategy, composed of a group of robust decoupling constrained controllers and a set of PI-type gain adaption-based controllers, is developed in this paper. It ensures that the outputs of each follower evolve inside the prescribed constraint bands and track the leader outputs with the tunable settling time and accuracy, despite the topology switching. Moreover, it exhibits a prominent simplicity, in spite of the severe model uncertainties, without parameter identification, function approximation, disturbance estimation, or derivative calculation or estimation. The theoretical findings are illustrated by a comparative simulation on a team of planar mobile robots.
- 2 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 12, 199-211
Distributed Robust Output-Constrained Control of Unknown Nonlinear Multiagent Systems Under Switching Topologies
The leader-following problem for the heterogeneous nonlinear multiagent systems under output constraints is investigated in this paper. It is focused on the cases where the underlying communication graph is directed and switching, and each agent is a multi-input multi-output block-triangular system with totally unknown and inherently nonlinear dynamics. They render the existing distributed control solutions infeasible. To surmount this challenge, a novel distributed mixed control strategy, composed of a group of robust decoupling constrained controllers and a set of PI-type gain adaption-based controllers, is developed in this paper. It ensures that the outputs of each follower evolve inside the prescribed constraint bands and track the leader outputs with the tunable settling time and accuracy, despite the topology switching. Moreover, it exhibits a prominent simplicity, in spite of the severe model uncertainties, without parameter identification, function approximation, disturbance estimation, or derivative calculation or estimation. The theoretical findings are illustrated by a comparative simulation on a team of planar mobile robots.
- 22 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (3) , 3000-3020
Dynamic Trust Empowerment Mechanism for Enhanced Security in Intelligent Connected Vehicle Networks Under Zero-Trust Framework
Traditional security mechanisms often fail to accurately assess Intelligent Connected Vehicle (ICV) nodes in highly dynamic environments. To address this, this study proposes a Dynamic Trust Empowerment (DTE) mechanism integrated within a Zero-Trust Framework. First, we construct four core evaluation models to quantify connectivity, importance, trustworthiness, and reliability using a modified graph-theoretic approach. Subsequently, an integrated aggregation algorithm is designed to compute the global weighted trust level. A key novelty is the introduction of a time-decaying dynamic empowerment function, which strictly enforces the “Never Trust, Always Verify” principle by naturally degrading trust values over time in the absence of continuous verification. Experimental results demonstrate that the DTE mechanism significantly outperforms static models, effectively identifying potential threats and ensuring timely security responses in high-mobility vehicular networks.
- 22 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Intelligent Transportation Systems
- Vol. 27 (3) , 3000-3020
Dynamic Trust Empowerment Mechanism for Enhanced Security in Intelligent Connected Vehicle Networks Under Zero-Trust Framework
Traditional security mechanisms often fail to accurately assess Intelligent Connected Vehicle (ICV) nodes in highly dynamic environments. To address this, this study proposes a Dynamic Trust Empowerment (DTE) mechanism integrated within a Zero-Trust Framework. First, we construct four core evaluation models to quantify connectivity, importance, trustworthiness, and reliability using a modified graph-theoretic approach. Subsequently, an integrated aggregation algorithm is designed to compute the global weighted trust level. A key novelty is the introduction of a time-decaying dynamic empowerment function, which strictly enforces the “Never Trust, Always Verify” principle by naturally degrading trust values over time in the absence of continuous verification. Experimental results demonstrate that the DTE mechanism significantly outperforms static models, effectively identifying potential threats and ensuring timely security responses in high-mobility vehicular networks.
- 13 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 23 (3) , 5355-5370
A Trust Assessment Method for Intelligent Connected Vehicles Based on Data Consistency Verification Matched With Adaptive Leader Election in a Zero-Trust Framework
The use of intelligent connected vehicles (ICVs) is an emerging concept in transportation systems, where the platoon leader processes various types of information from followers and roadside units to make accurate decisions. However, in complex network environments, the leader election process can become complicated and inefficient due to potential network attacks and abnormal node behaviours. To address this challenge, this paper presents a leader vehicle election mechanism based on dynamic trust assessment within a zero-trust framework. The proposed mechanism comprises two key components: dynamic trust assessment and adaptive leader election. The dynamic trust assessment method involves a trust evaluation mechanism composed of direct trust assessment and recommendation trust assessment based on the state information of vehicle nodes, where the dynamic trust assessment results are adjusted according to node behaviour information. The leader election mechanism involves periodic elections among the vehicle nodes on the basis of the results of the dynamic trust assessment, in which the vehicle node with the most votes is selected as the leader. Through comparative experiments against a full trust scenario, the proposed method demonstrates superior performance in resisting network attacks and identifying abnormal nodes, thereby significantly improving the safety, stability, and operational efficiency of the vehicle platoon.
- 13 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Dependable and Secure Computing
- Vol. 23 (3) , 5355-5370
A Trust Assessment Method for Intelligent Connected Vehicles Based on Data Consistency Verification Matched With Adaptive Leader Election in a Zero-Trust Framework
The use of intelligent connected vehicles (ICVs) is an emerging concept in transportation systems, where the platoon leader processes various types of information from followers and roadside units to make accurate decisions. However, in complex network environments, the leader election process can become complicated and inefficient due to potential network attacks and abnormal node behaviours. To address this challenge, this paper presents a leader vehicle election mechanism based on dynamic trust assessment within a zero-trust framework. The proposed mechanism comprises two key components: dynamic trust assessment and adaptive leader election. The dynamic trust assessment method involves a trust evaluation mechanism composed of direct trust assessment and recommendation trust assessment based on the state information of vehicle nodes, where the dynamic trust assessment results are adjusted according to node behaviour information. The leader election mechanism involves periodic elections among the vehicle nodes on the basis of the results of the dynamic trust assessment, in which the vehicle node with the most votes is selected as the leader. Through comparative experiments against a full trust scenario, the proposed method demonstrates superior performance in resisting network attacks and identifying abnormal nodes, thereby significantly improving the safety, stability, and operational efficiency of the vehicle platoon.
- 12 December 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 303
A two-stage large-scale multiobjective evolutionary algorithm based on offset direction sampling and dual-layer competition
- 12 December 2025
- Published by Elsevier in Expert Systems with Applications
- Vol. 303
A two-stage large-scale multiobjective evolutionary algorithm based on offset direction sampling and dual-layer competition