Shibei Xue

Passive linear quantum systems are widely encountered in quantum optics due to their inherent stability and energy‐dissipation characteristics. With the increasing deployment of such systems, the dimensions of passive linear quantum networks can grow rapidly, thereby increasing the computational burden of feedback control and estimation. Hence, model reduction methods should be developed to construct lower‐order models while preserving physical realizability, a constraint absent from classical model reduction. This article proposes an ‐norm‐based model reduction approach for completely passive linear quantum systems, which yields a physically realizable reduced‐order model that closely approximates the original system dynamics. This is achieved by explicitly incorporating the physical realizability conditions that preserve the quantum nature of the reduced system into the model reduction problem. Unlike classical model reduction settings, these physical realizability conditions introduce nonlinear constraints into the optimization problem. To handle these constraints, an observability‐Gramian‐based optimization formulation is developed. By converting the nonlinear equality constraints into inequality constraints, we enable an alternating direction method of multiplier (ADMM)‐iterative LMI (ILMI) algorithm framework to be constructed for solving the model reduction optimization problem. In contrast to existing methods that primarily provide a posteriori criteria for evaluating approximation performance, the proposed approach directly incorporates the error into the construction of the reduced‐order model. Two numerical examples, involving a three‐oscillator series‐connected system and a large‐scale chain‐connected quantum system, consistently demonstrate that the proposed method preserves the dominant input–output characteristics of the original systems.

Onboard estimation of lithium-ion battery state-of-health (SOH) is essential for safe electric-vehicle operation. Nonlinear capacity fade, scarce training data and pronounced local regeneration phenomena, however, complicate reliable SOH forecasting. This paper presents LMM-PatchTSSA, a lightweight framework that couples a Local Mixing Module (LMM) with a Patch Time Series Transformer (PatchTST) backbone. The LMM extracts short-term, degradation-sensitive signatures from charge–discharge curves, exposing early-stage ageing indicators that standard transformers overlook. Within PatchTST, a Token-Statistics Self-Attention (TSSA) mechanism replaces conventional quadratic self-attention; by formulating attention through second-order moment statistics, TSSA reduces computational complexity from [[EQUATION]]O(n2) to [[EQUATION]]O(n), while the patch-based structure captures long-range dependencies at linear cost. A frozen-backbone transfer strategy further mitigates chemistry-induced domain shifts. Trained and evaluated on the public CALCE dataset, the model achieves an average RMSE of 0.0118. Under zero-shot cross-chemistry transfer from CALCE (NMC) to NASA (LCO) cells without retraining, the framework records [[EQUATION]]R2>0.95; after fine-tuning on a minimal subset of target data, [[EQUATION]]R2 reaches 0.9913 on NASA B0007. These results demonstrate accurate SOH prediction, robust cross-chemistry generalisation and low inference overhead, making the framework suitable for real-time battery management systems.

In this paper, an $\mathscr {H}_{2}$ model reduction method for linear quantum systems is presented, which can obtain a physically realizable model with a reduced order for closely approximating the original system. The model reduction problem is formulated as an optimization problem, which minimizes the $\mathscr {H}_{2}$ norm of the difference between the transfer functions of the original and reduced models. Different from classical model reduction, physical realizability conditions for guaranteeing that the reduced-order system is also a quantum system should be taken as non-convex constraints in the optimization. Based on the special properties of quantum systems, we derive necessary conditions for $\mathscr {H}_{2}$ optimality. Moreover, we employ a matrix inequality approach to transform nonlinear inequality constraints into readily solvable linear matrix inequalities (LMIs) and nonlinear equality constraints, so that it can be solved by a lifting variables approach. We emphasize that different from existing works, which only introduces a criterion to evaluate the performance after model reduction, we guide our method to obtain an optimal reduced model with respect to the $\mathscr {H}_{2}$ norm. Finally, a comparison with the results in the existing works validates the efficacy of the proposed method.

Federated learning is vulnerable to poisoning attacks by malicious adversaries. Existing methods often involve high costs to achieve effective attacks. To address this challenge, we propose a sybil-based virtual data poisoning attack, where a malicious client generates sybil nodes to amplify the poisoning model's impact. To reduce neural network computational complexity, we develop a virtual data generation method based on gradient matching. We also design three schemes for target model acquisition, applicable to online local, online global, and offline scenarios. In simulation, our method outperforms other attack algorithms since our method can obtain a global target model under non-independent uniformly distributed data.