arXiv subjects

Qian Long

Publications and source records attributed to Qian Long.

At least 19 recordsLinked to original sources

arXiv2026

Limitations of RMS-based Stability Assessment of Converter-based Grids: A Case Study

The increasing penetration of power electronic converters in industrial grids introduces stability challenges at frequencies well above the electromechanical range of traditional power systems. Commercial software tools typically perform eigenvalue-based stability analysis using root-mean-square (RMS) models that assume a quasi-stationary network, thereby neglecting electromagnetic dynamics of assets --- which is standard practice in power system analysis. This paper presents a case study of an industrial grid where the conventional RMS-based eigenvalue analysis predicts stable operation, while a detailed electromagnetic transient (EMT) simulation reveals growing oscillations, indicating instability. To bridge this gap, we introduce an alternative modeling approach that formulates the network dynamics in the rotating $dq$ reference frame, retaining the differential equations of electromagnetic dynamics. The resulting model enables eigenvalue analysis that correctly identifies the unstable modes, consistent with the EMT simulation results. Our findings highlight a fundamental limitation of RMS-based stability assessment for converter-dominated grids and demonstrate that dynamic impedance models in the dq frame provide a viable path toward accurate small-signal analysis encompassing high-frequency converter-network interactions.

eess.SY↗

arXiv2026

Beyond Global Replanning: Hierarchical Recovery for Cross-Device Agent Systems

Real-world computer-use tasks often span multiple applications and devices, requiring agents to coordinate heterogeneous environments under dynamic runtime failures. Existing multi-device agent systems support task decomposition and cross-device assignment, but recovery remains largely coarse-grained: when execution fails, they typically retry the same strategy, reassign the subtask, or revise the global plan, without systematically modeling the device-local strategy space. This limits their ability to distinguish failures that can be repaired within the current device from those that require cross-device replanning. We propose \textbf{H-RePlan}, a hierarchical replanning framework for multi-device agents with unified API--CLI--GUI execution. H-RePlan equips each device with interchangeable execution strategies and separates device-local strategy recovery from orchestrator-level global replanning through a compact cross-layer failure abstraction. To evaluate this capability, we introduce \textbf{HeraBench}, a fault-injected benchmark that constructs cross-device workflows over Linux and Android devices and injects strategy- and device-level failures. Experiments show that H-RePlan substantially outperforms single-strategy and coarse-grained multi-device baselines, achieving higher completion, instruction adherence, and perfect-pass rates while reducing the token cost required for reliable end-to-end success. These results demonstrate that scope-aware hierarchical recovery is essential for robust multi-device agent execution.

cs.CL↗

arXiv2026

Can Current Agents Close the Discovery-to-Application Gap? A Case Study in Minecraft

Discovering causal regularities and applying them to build functional systems--the discovery-to-application loop--is a hallmark of general intelligence, yet evaluating this capacity has been hindered by the vast complexity gap between scientific discovery and real-world engineering. We introduce SciCrafter, a Minecraft-based benchmark that operationalizes this loop through parameterized redstone circuit tasks. Agents must ignite lamps in specified patterns (e.g., simultaneously or in timed sequences); scaling target parameters substantially increases construction complexity and required knowledge, forcing genuine discovery rather than reliance on memorized solutions. Evaluating frontier models including GPT-5.2, Gemini-3-Pro, and Claude-Opus-4.5 under a general-purpose code agent scaffold, we find that all plateau at approximately 26% success rate. To diagnose these failures, we decompose the loop into four capacities--knowledge gap identification, experimental discovery, knowledge consolidation, and knowledge application--and design targeted interventions whose marginal contributions serve as proxies for corresponding gaps. Our analysis reveals that although the general knowledge application capability still remains as the biggest gap across all models, for frontier models the knowledge gap identification starts to become a major hurdle--indicating the bottleneck is shifting from solving problems right to raising the right problems for current AI. We release SciCrafter as a diagnostic probe for future research on AI systems that navigate the full discovery-to-application loop.

cs.AI↗

arXiv2026

Scaling World Model for Hierarchical Manipulation Policies

Vision-Language-Action (VLA) models are promising for generalist robot manipulation but remain brittle in out-of-distribution (OOD) settings, especially with limited real-robot data. To resolve the generalization bottleneck, we introduce a hierarchical Vision-Language-Action framework \our{} that leverages the generalization of large-scale pre-trained world model for robust and generalizable VIsual Subgoal TAsk decomposition VISTA. Our hierarchical framework \our{} consists of a world model as the high-level planner and a VLA as the low-level executor. The high-level world model first divides manipulation tasks into subtask sequences with goal images, and the low-level policy follows the textual and visual guidance to generate action sequences. Compared to raw textual goal specification, these synthesized goal images provide visually and physically grounded details for low-level policies, making it feasible to generalize across unseen objects and novel scenarios. We validate both visual goal synthesis and our hierarchical VLA policies in massive out-of-distribution scenarios, and the performance of the same-structured VLA in novel scenarios could boost from 14% to 69% with the guidance generated by the world model. Results demonstrate that our method outperforms previous baselines with a clear margin, particularly in out-of-distribution scenarios. Project page: \href{https://vista-wm.github.io/}{https://vista-wm.github.io}

cs.RO↗

arXiv2026

Electrical Design of a Clean Offshore Heat and Power (CleanOFF) Hub

This paper presents an innovative offshore solution where oil & gas platform clusters are powered by a wind farm and a hydrogen hub. The results show a feasible off-grid design as an alternative to conventional electrification solutions. To address the challenges of design and operation of such a system, a power system model of the equipment and control was developed in a power system simulator called Process Power Simulator (PPSim). Power fluctuations in the wind farm are modelled using a state-of-the-art method encompassing turbulence and wakes. Various operation scenarios were used to evaluate the system design and find the right equipment size. An expensive component to over dimension is the battery energy storage system (BESS). The BESS power rating and energy capacity were found by running a combination of scenarios with extreme and natural wind variations, and contingencies. The control strategy and ramp rates of electrolyzers have significant impact on both system performance and design. A ramp rate in the order of seconds as opposed to minutes will decrease the required BESS size by 60-70%. Choosing synchronized control of the electrolyzers can further reduce the BESS size by 15-20%. The simulations also revealed challenges to achieve self-sufficiency of hydrogen and potential design improvements are suggested.

eess.SY↗

arXiv2025

Muon-Accelerated Attention Distillation for Real-Time Edge Synthesis via Optimized Latent Diffusion

Recent advances in visual synthesis have leveraged diffusion models and attention mechanisms to achieve high-fidelity artistic style transfer and photorealistic text-to-image generation. However, real-time deployment on edge devices remains challenging due to computational and memory constraints. We propose Muon-AD, a co-designed framework that integrates the Muon optimizer with attention distillation for real-time edge synthesis. By eliminating gradient conflicts through orthogonal parameter updates and dynamic pruning, Muon-AD achieves 3.2 times faster convergence compared to Stable Diffusion-TensorRT, while maintaining synthesis quality (15% lower FID, 4% higher SSIM). Our framework reduces peak memory to 7GB on Jetson Orin and enables 24FPS real-time generation through mixed-precision quantization and curriculum learning. Extensive experiments on COCO-Stuff and ImageNet-Texture demonstrate Muon-AD's Pareto-optimal efficiency-quality trade-offs. Here, we show a 65% reduction in communication overhead during distributed training and real-time 10s/image generation on edge GPUs. These advancements pave the way for democratizing high-quality visual synthesis in resource-constrained environments.

cs.CV↗

arXiv2024

TeamCraft: A Benchmark for Multi-Modal Multi-Agent Systems in Minecraft

Collaboration is a cornerstone of society. In the real world, human teammates make use of multi-sensory data to tackle challenging tasks in ever-changing environments. It is essential for embodied agents collaborating in visually-rich environments replete with dynamic interactions to understand multi-modal observations and task specifications. To evaluate the performance of generalizable multi-modal collaborative agents, we present TeamCraft, a multi-modal multi-agent benchmark built on top of the open-world video game Minecraft. The benchmark features 55,000 task variants specified by multi-modal prompts, procedurally-generated expert demonstrations for imitation learning, and carefully designed protocols to evaluate model generalization capabilities. We also perform extensive analyses to better understand the limitations and strengths of existing approaches. Our results indicate that existing models continue to face significant challenges in generalizing to novel goals, scenes, and unseen numbers of agents. These findings underscore the need for further research in this area. The TeamCraft platform and dataset are publicly available at https://github.com/teamcraft-bench/teamcraft.

cs.AI↗

arXiv2024

Inverse Attention Agents for Multi-Agent Systems

A major challenge for Multi-Agent Systems is enabling agents to adapt dynamically to diverse environments in which opponents and teammates may continually change. Agents trained using conventional methods tend to excel only within the confines of their training cohorts; their performance drops significantly when confronting unfamiliar agents. To address this shortcoming, we introduce Inverse Attention Agents that adopt concepts from the Theory of Mind (ToM) implemented algorithmically using an attention mechanism trained in an end-to-end manner. Crucial to determining the final actions of these agents, the weights in their attention model explicitly represent attention to different goals. We furthermore propose an inverse attention network that deduces the ToM of agents based on observations and prior actions. The network infers the attentional states of other agents, thereby refining the attention weights to adjust the agent's final action. We conduct experiments in a continuous environment, tackling demanding tasks encompassing cooperation, competition, and a blend of both. They demonstrate that the inverse attention network successfully infers the attention of other agents, and that this information improves agent performance. Additional human experiments show that, compared to baseline agent models, our inverse attention agents exhibit superior cooperation with humans and better emulate human behaviors.

cs.AI↗

arXiv2024

T2V-Turbo-v2: Enhancing Video Generation Model Post-Training through Data, Reward, and Conditional Guidance Design

In this paper, we focus on enhancing a diffusion-based text-to-video (T2V) model during the post-training phase by distilling a highly capable consistency model from a pretrained T2V model. Our proposed method, T2V-Turbo-v2, introduces a significant advancement by integrating various supervision signals, including high-quality training data, reward model feedback, and conditional guidance, into the consistency distillation process. Through comprehensive ablation studies, we highlight the crucial importance of tailoring datasets to specific learning objectives and the effectiveness of learning from diverse reward models for enhancing both the visual quality and text-video alignment. Additionally, we highlight the vast design space of conditional guidance strategies, which centers on designing an effective energy function to augment the teacher ODE solver. We demonstrate the potential of this approach by extracting motion guidance from the training datasets and incorporating it into the ODE solver, showcasing its effectiveness in improving the motion quality of the generated videos with the improved motion-related metrics from VBench and T2V-CompBench. Empirically, our T2V-Turbo-v2 establishes a new state-of-the-art result on VBench, with a Total score of 85.13, surpassing proprietary systems such as Gen-3 and Kling.

cs.CV↗

arXiv2024

SocialGFs: Learning Social Gradient Fields for Multi-Agent Reinforcement Learning

Multi-agent systems (MAS) need to adaptively cope with dynamic environments, changing agent populations, and diverse tasks. However, most of the multi-agent systems cannot easily handle them, due to the complexity of the state and task space. The social impact theory regards the complex influencing factors as forces acting on an agent, emanating from the environment, other agents, and the agent's intrinsic motivation, referring to the social force. Inspired by this concept, we propose a novel gradient-based state representation for multi-agent reinforcement learning. To non-trivially model the social forces, we further introduce a data-driven method, where we employ denoising score matching to learn the social gradient fields (SocialGFs) from offline samples, e.g., the attractive or repulsive outcomes of each force. During interactions, the agents take actions based on the multi-dimensional gradients to maximize their own rewards. In practice, we integrate SocialGFs into the widely used multi-agent reinforcement learning algorithms, e.g., MAPPO. The empirical results reveal that SocialGFs offer four advantages for multi-agent systems: 1) they can be learned without requiring online interaction, 2) they demonstrate transferability across diverse tasks, 3) they facilitate credit assignment in challenging reward settings, and 4) they are scalable with the increasing number of agents.

cs.AI↗

arXiv2023

Sim2Plan: Robot Motion Planning via Message Passing between Simulation and Reality

Simulation-to-real is the task of training and developing machine learning models and deploying them in real settings with minimal additional training. This approach is becoming increasingly popular in fields such as robotics. However, there is often a gap between the simulated environment and the real world, and machine learning models trained in simulation may not perform as well in the real world. We propose a framework that utilizes a message-passing pipeline to minimize the information gap between simulation and reality. The message-passing pipeline is comprised of three modules: scene understanding, robot planning, and performance validation. First, the scene understanding module aims to match the scene layout between the real environment set-up and its digital twin. Then, the robot planning module solves a robotic task through trial and error in the simulation. Finally, the performance validation module varies the planning results by constantly checking the status difference of the robot and object status between the real set-up and the simulation. In the experiment, we perform a case study that requires a robot to make a cup of coffee. Results show that the robot is able to complete the task under our framework successfully. The robot follows the steps programmed into its system and utilizes its actuators to interact with the coffee machine and other tools required for the task. The results of this case study demonstrate the potential benefits of our method that drive robots for tasks that require precision and efficiency. Further research in this area could lead to the development of even more versatile and adaptable robots, opening up new possibilities for automation in various industries.

cs.RO↗

arXiv2022

How Close Dark Matter Halos and MOND Are to Each Other: Three-Dimensional Tests Based on Gaia DR2

Aiming at discriminating different gravitational potential models of the Milky Way, we perform tests based on the kinematic data powered by the Gaia DR2 astrometry, over a large range of $(R,z)$ locations. Invoking the complete form of Jeans equations that admit three integrals of motion, we use the independent $R$- and $z$-directional equations as two discriminators ($T_R$ and $T_z$). We apply the formula for spatial distributions of radial and vertical velocity dispersions proposed by Binney et al., and successfully extend it to azimuthal components, $σ_θ(R,z)$ and $V_θ(R,z)$; the analytic form avoids the numerical artifacts caused by numerical differentiation in Jeans-equations calculation given the limited spatial resolutions of observations, and more importantly reduces the impact of kinematic substructures in the Galactic disk. It turns out that whereas the current kinematic data are able to reject Moffat's Modified Gravity (let alone the Newtonian baryon-only model), Milgrom's MOND is still not rejected. In fact, both the carefully calibrated fiducial model invoking a spherical dark matter (DM) halo and MOND are equally consistent with the data at almost all spatial locations (except that probably both have respective problems at low-$|z|$ locations), no matter which a tracer population or which a meaningful density profile is used. Because there is no free parameter at all in the quasi-linear MOND model we use, and the baryonic parameters are actually fine-tuned in the DM context, such an effective equivalence is surprising, and might be calling forth a transcending synthesis of the two paradigms.

astro-ph.GA↗

arXiv2022

Identification of new M31 star cluster candidates from PAndAS images using convolutional neural networks

Context.Identification of new star cluster candidates in M31 is fundamental for the study of the M31 stellar cluster system. The machine-learning method convolutional neural network (CNN) is an efficient algorithm for searching for new M31 star cluster candidates from tens of millions of images from wide-field photometric surveys. Aims.We search for new M31 cluster candidates from the high-quality $g$- and $i$-band images of 21,245,632 sources obtained from the Pan-Andromeda Archaeological Survey (PAndAS) through a CNN. Methods.We collected confirmed M31 clusters and noncluster objects from the literature as our training sample. Accurate double-channel CNNs were constructed and trained using the training samples. We applied the CNN classification models to the PAndAS $g$- and $i$-band images of over 21 million sources to search new M31 cluster candidates. The CNN predictions were finally checked by five experienced human inspectors to obtain high-confidence M31 star cluster candidates. Results.After the inspection, we identified a catalogue of 117 new M31 cluster candidates. Most of the new candidates are young clusters that are located in the M31 disk. Their morphology, colours, and magnitudes are similar to those of the confirmed young disk clusters. We also identified eight globular cluster candidates that are located in the M31 halo and exhibit features similar to those of confirmed halo globular clusters. The projected distances to the M31 centre for three of them are larger than 100\,kpc.

astro-ph.GA↗

arXiv2022

Hierarchical Frequency Control of Hybrid Power Plants Using Frequency Response Observer

Frequency control (FC) enables utility-scale grid-connected hybrid power plants (HPPs) to operate in compliance with grid code requirements while to capture value streams from provision of frequency control services (FCSs). In this paper, a novel hierarchical FC approach is proposed to allow HPPs to provide three types of FCSs, namely fast frequency response (FFR), frequency containment response (FCR) and frequency restoration response (FRR). To accommodate state-of-the-art fast FC, controllers for fast FCSs, such as FFR and FCR, are implemented at asset controllers, while controllers for slow FCSs like FRR are implemented at plant controllers or the HPP controller (HPPC). Control counteraction issue, which arises across control hierarchy, is then discussed. To solve this issue, an innovative frequency response observer (FROB) is proposed. Inspired by the concept of disturbance observer (DOB), FROB at plant controllers and the HPPC accurately estimates frequency response initiated at asset controllers, and the obtained estimation is used for control compensation at plant controllers and the HPPC to avoid control counteraction. This scheme achieves robust performance even when there are system uncertainties existing in HPPs, such as parameter uncertainty, unknown control malfunction, and time-varying communication delays. The proposed approach is implemented in a power system dynamic model in MATLAB/Simulink to highlight its effectiveness and robustness.

eess.SY↗

arXiv2020

Diesel Generator Model Parameterization for Microgrid Simulation Using Hybrid Box-Constrained Levenberg-Marquardt Algorithm

Existing generator parameterization methods, typically developed for large turbine generator units, are difficult to apply to small kW-level diesel generators in microgrid applications. This paper presents a model parameterization method that estimates a complete set of kW-level diesel generator parameters simultaneously using only load-step-change tests with limited measurement points. This method provides a more cost-efficient and robust approach to achieve high-fidelity modeling of diesel generators for microgrid dynamic simulation. A two-stage hybrid box-constrained Levenberg-Marquardt (H-BCLM) algorithm is developed to search the optimal parameter set given the parameter bounds. A heuristic algorithm, namely Generalized Opposition-based Learning Genetic Algorithm (GOL-GA), is applied to identify proper initial estimates at the first stage, followed by a modified Levenberg-Marquardt algorithm designed to fine tune the solution based on the first-stage result. The proposed method is validated against dynamic simulation of a diesel generator model and field measurements from a 16kW diesel generator unit.

eess.SY↗

arXiv2020

Deep Learning for Strong Lensing Search: Tests of the Convolutional Neural Networks and New Candidates from KiDS DR3

Convolutional Neutral Networks have been successfully applied in searching for strong lensing systems, leading to discoveries of new candidates from large surveys. On the other hand, systematic investigations about their robustness are still lacking. In this paper, we first construct a neutral network, and apply it to $r$-band images of Luminous Red Galaxies (LRGs) of the Kilo Degree Survey (KiDS) Data Release 3 to search for strong lensing systems. We build two sets of training samples, one fully from simulations, and the other one using the LRG stamps from KiDS observations as the foreground lens images. With the former training sample, we find 48 high probability candidates after human-inspection, and among them, 27 are newly identified. Using the latter training set, about 67\% of the aforementioned 48 candidates are also found, and there are 11 more new strong lensing candidates identified. We then carry out tests on the robustness of the network performance with respect to the variation of PSF. With the testing samples constructed using PSF in the range of 0.4 to 2 times of the median PSF of the training sample, we find that our network performs rather stable, and the degradation is small. We also investigate how the volume of the training set can affect our network performance by varying it from 0.1 millions to 0.8 millions. The output results are rather stable showing that within the considered range, our network performance is not very sensitive to the volume size.

astro-ph.CO↗

arXiv2020

Evolutionary Population Curriculum for Scaling Multi-Agent Reinforcement Learning

In multi-agent games, the complexity of the environment can grow exponentially as the number of agents increases, so it is particularly challenging to learn good policies when the agent population is large. In this paper, we introduce Evolutionary Population Curriculum (EPC), a curriculum learning paradigm that scales up Multi-Agent Reinforcement Learning (MARL) by progressively increasing the population of training agents in a stage-wise manner. Furthermore, EPC uses an evolutionary approach to fix an objective misalignment issue throughout the curriculum: agents successfully trained in an early stage with a small population are not necessarily the best candidates for adapting to later stages with scaled populations. Concretely, EPC maintains multiple sets of agents in each stage, performs mix-and-match and fine-tuning over these sets and promotes the sets of agents with the best adaptability to the next stage. We implement EPC on a popular MARL algorithm, MADDPG, and empirically show that our approach consistently outperforms baselines by a large margin as the number of agents grows exponentially.

cs.LG↗

arXiv2020

A Robust Real-Time Computing-based Environment Sensing System for Intelligent Vehicle

For intelligent vehicles, sensing the 3D environment is the first but crucial step. In this paper, we build a real-time advanced driver assistance system based on a low-power mobile platform. The system is a real-time multi-scheme integrated innovation system, which combines stereo matching algorithm with machine learning based obstacle detection approach and takes advantage of the distributed computing technology of a mobile platform with GPU and CPUs. First of all, a multi-scale fast MPV (Multi-Path-Viterbi) stereo matching algorithm is proposed, which can generate robust and accurate disparity map. Then a machine learning, which is based on fusion technology of monocular and binocular, is applied to detect the obstacles. We also advance an automatic fast calibration mechanism based on Zhang's calibration method. Finally, the distributed computing and reasonable data flow programming are applied to ensure the operational efficiency of the system. The experimental results show that the system can achieve robust and accurate real-time environment perception for intelligent vehicles, which can be directly used in the commercial real-time intelligent driving applications.

cs.CV↗