Darong Huang

Proton exchange membrane electrolyzer (PEMELE) water electrolysis has garnered considerable interest for its high efficiency, particularly in microgrid applications. However, the diffusion processes within PEMELE introduce significant nonlinearity in the voltage-current relationship and exacerbate current ripple during operation. Excessive current ripple not only compromises hydrogen production efficiency and system performance but also accelerates device degradation. To mitigate these issues, this article introduces a fractional-order equivalent circuit model that accurately captures the dynamics of PEMELE. Based on this model, a dynamic fractional-order model predictive control (DFOMPC) strategy is developed, utilizing both real-time and historical current data to predict voltage-current behavior. By explicitly accounting for the nonlinear and memory-dependent characteristics of PEMELE, the proposed approach achieves superior control performance. Simulations and hardware experiments demonstrate that, compared to conventional MPC, the proposed method reduces current ripple by 39.7% to 42.4% and voltage ripple by 18.5%–27.6%.

Accurate state-of-health (SOH) prediction of lithium-ion batteries is critical for ensuring operational safety, reliability, and extending service lifetime in energy storage systems. Existing methods for predicting the SOH of lithiumion batteries have predominantly focused on per-cycle estimation, often overlooking two critical longer-range patterns: latent temporal dependencies across cycles within single battery and continual sharing knowledge across battery packs, both of which are critical for capturing degradation patterns. To address this, we propose a graph continual learning framework with longer-range patterns discovering. First, to explicitly characterize cross-cycle temporal dependencies, a temporal graph representation is formulated through an enhanced multi-order dynamic time warping, wherein directed edges deterministically encode inter-cycle relational patterns. Then, an edge-oriented network is designed to depict temporal representations and aggregation along directed edges, further augmented with a global cell state to guide the capture of persistent memory across batteries. After that, a replay continual learning strategy is adopted to mitigate catastrophic forgetting, combined with lightweight graph distillation to significantly reduce memory overhead while ensuring practical deployment. Finally, the proposed method achieves an average root mean square error (RMSE) of 0.693% in a real-world lithium-ion battery dataset, yielding a 23.09% average improvement over baseline methods, even with only 1.25% of original data volume, thereby demonstrating its effectiveness and strong generalization capability.

This paper deals with the distributed resource allocation problem with a nonconvex objective function over a weight-unbalanced directed communication network, where a coupled equality constraint and a private convex set are considered. To handle this issue, a distributed consensus-based estimator is introduced to track the coupled constraint within a finite time. Then, based on this estimator, a primal-dual distributed optimization approach is designed, where the local dual variable guarantees the global constraint and cooperatively achieves the optimal centralized dual variable. To ensure the stability of the primal subsystem, based on the penalty-like function method, a convex term is added to the augmented Lagrangian multiplier function such that its Hessian matrix over the primal variable is positive definite. Finally, the developed method's asymptotical convergence rate is established, and its effectiveness is evaluated via two case simulations.

This paper presents a stability analysis of virus propagation models for Connected and Autonomous Vehicle (CAV) platoons with time-varying transmission rates, particularly focusing on scenarios within a fast-switching network environment. By defining the dynamic nature of a CAV platoon virus propagation model, we first analyze the stability behavior of the virus propagation model under fixed transmission rates, serving as a baseline. Furthermore, we introduce a layered network framework that accommodates fast switching-reflecting the highly dynamic nature of real-world CAV platoon networks. Then, we introduce a new variable that captures the essential dynamics of virus propagation in CAV platoons under fast-switching conditions. By leveraging Lyapunov's method, we provide a rigorous proof of the system's stability. Specifically, we demonstrate that despite the time-varying nature of the virus transmission rates and the fast switching of the network layers, the CAV platoon virus propagation model exhibits bounded stable behavior under certain conditions. This study proposes a novel analytical framework for investigating the dynamics and stability of virus propagation in CAV platoons. Through theoretical analysis and simulations, we demonstrate the impact of time-varying transmission rates and fast-switching networks on the stability of the CAV platoon virus propagation model, thereby advancing our knowledge in CAV cybersecurity and virus mitigation strategies-especially in dynamic traffic scenarios.

The hyper-redundant robots (HRRs) exhibit diverse locomotion and manipulation capabilities based on backbone curves. However, the low computational efficiency of the existing backbone curve discretization methods limits their applicability. In this article, we proposed the space rotation minimizing frame bellows model (SRMF bellows model) and an improved discretization method based on the double reflection method. The proposed method directly maps backbone curves to HRR configurations by solving relative rotations between adjacent SRMFs, thereby eliminating the numerical integration of curvature and torsion required in conventional approaches and improving the motion planning efficiency of HRRs. To evaluate the performance of the proposed method, the computational burden of motion generation from universal time-invariant and time-varying backbone curves is analyzed and compared between the proposed and existing methods. Finally, experimental results from comparative real-world experiments involving hump rolling and bottle-carrying tasks, under both time-invariant and time-varying backbone curves, demonstrate the effectiveness and robustness of the proposed method. Compared with existing discretization approaches, the proposed method achieves a 47.2% reduction in execution time and a 94.6% improvement in rolling forward velocity.

In this paper, when actuator failure occurs, an adaptive fault-tolerant control (AFTC) strategy is proposed to solve the problem of attitude tracking control and vibration suppression in a flexible satellite system described by an ODE–PDE interconnected system. The moment of inertia matrix of the system and which actuator fails and when the failure occurs are unknown. To solve these problems, an adaptive control strategy is designed based on uncertainty parametrization and symmetric positive definite matrix–diagonal matrix–unit upper triangular matrix (SDU) factorization. The modified Lyapunov method and the assumed mode principle are used to prove that attitude tracking and vibration suppression can be realized. Finally, the simulation results show that AFTC strategy can be realized and is effective.