arXiv subjects

Jiaming Wu

Publications and source records attributed to Jiaming Wu.

16 recordsLinked to original sources

arXiv2026

Joint Planning and Scheduling of Modular Vehicles for Passenger-Freight Integration

This paper proposes a modular vehicle system for passenger-freight integration along a bidirectional transit corridor. The system uses homogeneous units that can be coupled into vehicles and assigned to either passenger or freight service. Freight is carried by dedicated units, with loading and unloading coordinated with docking and undocking and separated from passenger boarding and alighting. To better respond to uncertain passenger demand and integrate freight transport, vehicles can be reconfigured at intermediate stations, where they can also depart and terminate. We jointly optimize departure-specific service routes, timetables, vehicle compositions, unit schedules, and passenger-freight demand assignments, with unit reuse constrained by explicit docking and undocking times. These decisions are modeled on a space-time-state network and formulated as a stochastic mixed-integer program that minimizes unit deployment costs, passenger waiting costs, and penalties for unmet freight demand. Passenger demand uncertainty is addressed using linearized chance constraints. To solve the problem, we develop an exact Benders decomposition algorithm with valid inequalities and a warm-start strategy, together with a tailored decomposition-based heuristic for larger instances. Computational experiments on instances generated from representative transit corridors in Gothenburg demonstrate the effectiveness of the Benders algorithm for small- and medium-sized instances and the scalability of the heuristic for larger problems. Sensitivity analyses highlight the value of accounting for passenger demand uncertainty and the effects of temporal overlap between passenger and freight demand. Comparisons with benchmark transit systems further demonstrate the operational advantages of the proposed modular integrated system.

math.OC↗

arXiv2026

Towards Feedback-to-Plan Decisions for Self-Evolving LLM Agents in CUDA Kernel Generation

Large language models (LLMs) have shown strong empirical gains as self-evolving agents for CUDA kernel generation, driven by feedback-conditioned planning across generations. However, how planning decisions attribute and combine heterogeneous feedback signals remains opaque. Standard end-to-end ablations fail to resolve this question, as iterative planning amplifies early perturbations and conflates feedback effects with trajectory-dependent drift. We introduce \texttt{CUDAnalyst}, a unified analysis layer for controlled, generation-level attribution of planning decisions to feedback components via trajectory freezing and selective feedback injection. \texttt{CUDAnalyst} enables stable generation-level evaluation and principled coalitional-style attribution of feedback effects and interactions. Our results show that explicit planning is beneficial only when feedback is aligned, that effective planning emerges from structured multi-feedback interactions, and that high-level plans from stronger reasoning models can partially transfer to weaker ones. These trends hold across reference backbones, representative workloads, and reference induction regimes, indicating that the identified feedback-to-plan structure is robust within the controlled axes studied.

cs.AI↗

arXiv2026

A user-driven pricing and scheduling framework for public electric vehicle charging

Public electric vehicle (EV) charging infrastructure has expanded rapidly, yet utilization across charging stations remains uneven and often inefficient. Existing operator-determined pricing schemes offer limited flexibility to coordinate heterogeneous user demand within constrained capacity. This study proposes a user-driven pricing and scheduling framework for public EV charging. Users submit advance bids specifying acceptable time slots, bid prices, and quantity bounds. Based on these bids, the charging operator determines prices and charging slot assignments. After observing the outcomes, users decide whether to accept the resulting prices and allocations. The operator's decision problem incorporates profit objectives, user participation requirements, and capacity constraints across charging levels and time slots. The framework captures a three-stage interaction involving user bidding, operator decisions, and user acceptance. Numerical case studies reveal trade-offs among user acceptance, operator revenue, and charging capacity utilization under different charging levels. The findings provide practical guidance and insights into designing flexible pricing schemes that better accommodate heterogeneous user preferences while improving system efficiency.

math.OC↗

arXiv2026

HybridFlow: A Two-Step Generative Policy for Robotic Manipulation

Limited by inference latency, existing robot manipulation policies lack sufficient real-time interaction capability with the environment. Although faster generation methods such as flow matching are gradually replacing diffusion methods, researchers are pursuing even faster generation suitable for interactive robot control. MeanFlow, as a one-step variant of flow matching, has shown strong potential in image generation, but its precision in action generation does not meet the stringent requirements of robotic manipulation. We therefore propose \textbf{HybridFlow}, a \textbf{3-stage method} with \textbf{2-NFE}: Global Jump in MeanFlow mode, ReNoise for distribution alignment, and Local Refine in ReFlow mode. This method balances inference speed and generation quality by leveraging the rapid advantage of MeanFlow one-step generation while ensuring action precision with minimal generation steps. Through real-world experiments, HybridFlow outperforms the 16-step Diffusion Policy by \textbf{15--25\%} in success rate while reducing inference time from 152ms to 19ms (\textbf{8$\times$ speedup}, \textbf{$\sim$52Hz}); it also achieves 70.0\% success on unseen-color OOD grasping and 66.3\% on deformable object folding. We envision HybridFlow as a practical low-latency method to enhance real-world interaction capabilities of robotic manipulation policies.

cs.RO↗

arXiv2025

Learning for routing: A guided review of recent developments and future directions

This paper reviews the current progress in applying machine learning (ML) tools to solve NP-hard combinatorial optimization problems, with a focus on routing problems such as the traveling salesman problem (TSP) and the vehicle routing problem (VRP). Due to the inherent complexity of these problems, exact algorithms often require excessive computational time to find optimal solutions, while heuristics can only provide approximate solutions without guaranteeing optimality. With the recent success of machine learning models, there is a growing trend in proposing and implementing diverse ML techniques to enhance the resolution of these challenging routing problems. We propose a taxonomy categorizing ML-based routing methods into construction-based and improvement-based approaches, highlighting their applicability to various problem characteristics. This review aims to integrate traditional OR methods with state-of-the-art ML techniques, providing a structured framework to guide future research and address emerging VRP variants.

cs.AI↗

arXiv2025

Beyond Simple Graphs: Neural Multi-Objective Routing on Multigraphs

Learning-based methods for routing have gained significant attention in recent years, both in single-objective and multi-objective contexts. Yet, existing methods are unsuitable for routing on multigraphs, which feature multiple edges with distinct attributes between node pairs, despite their strong relevance in real-world scenarios. In this paper, we propose two graph neural network-based methods to address multi-objective routing on multigraphs. Our first approach operates directly on the multigraph by autoregressively selecting edges until a tour is completed. The second model, which is more scalable, first simplifies the multigraph via a learned pruning strategy and then performs autoregressive routing on the resulting simple graph. We evaluate both models empirically, across a wide range of problems and graph distributions, and demonstrate their competitive performance compared to strong heuristics and neural baselines.

cs.LG↗

arXiv2025

Collaborative Charging Scheduling via Balanced Bounding Box Methods

Electric mobility faces several challenges, most notably the high cost of infrastructure development and the underutilization of charging stations. The concept of shared charging offers a promising solution. The paper explores sustainable urban logistics through horizontal collaboration between two fleet operators and addresses a scheduling problem for the shared use of charging stations. To tackle this, the study formulates a collaborative scheduling problem as a bi-objective nonlinear integer programming model, in which each company aims to minimize its own costs, creating inherent conflicts that require trade-offs. The Balanced Bounding Box Methods (B3Ms) are introduced in order to efficiently derive the efficient frontier, identifying a reduced set of representative solutions. These methods enhance computational efficiency by selectively disregarding closely positioned and competing solutions, preserving the diversity and representativeness of the solutions over the efficient frontier. To determine the final solution and ensure balanced collaboration, cooperative bargaining methods are applied. Numerical case studies demonstrate the viability and scalability of the developed methods, showing that the B3Ms can significantly reduce computational time while maintaining the integrity of the frontier. These methods, along with cooperative bargaining, provide an effective framework for solving various bi-objective optimization problems, extending beyond the collaborative scheduling problem presented here.

math.OC↗

arXiv2025

A dynamical memory with only one spiking neuron

Common wisdom indicates that to implement a Dynamical Memory with spiking neurons two ingredients are necessary: recurrence and a neuron population. Here we shall show that the second requirement is not needed. We shall demonstrate that under very general assumptions a single recursive spiking neuron can realize a robust model of a dynamical memory. We demonstrate the implementation of a dynamical memory in both, software and hardware. In the former case, we introduce trivial extensions of the popular aQIF and AdEx models. In the latter, we show traces obtained in a circuit model with a recently proposed memristive spiking neuron. We show that the bistability of the theoretical models can be understood in terms of a self-consistent problem that can be represented geometrically. Our minimal dynamical memory model provides a simplest implementation of an important neuro-computational primitive, which can be useful in navigation system models based on purely spiking dynamics. A one neuron dynamical memory may also provides a natural explanation to the surprising recent observation that the excitation bump in Drosophila's ellipsoidal body is made by just a handful of neurons.

q-bio.NC↗

arXiv2025

The large-scale charging scheduling problem for fleet batteries: Lagrangian decomposition with time-block reformulations

There is a rise in the need for efficient battery charging methods due to the high penetration of electromobility solutions. Battery swapping, a technique in which fully or partially depleted batteries are exchanged and then transported to a central facility for charging, introduces a unique scheduling problem. For scenarios involving a large number of batteries, commercial solvers and existing methods do not yield optimal or near-optimal solutions in a reasonable time due to high computational complexity. Our study presents a novel approach that combines variable layering with Lagrangian decomposition. We develop a new, tighter time-block reformulation for one of the Lagrangian sub-problems, enhancing convergence rates when used with our partial-variable fixing Lagrangian heuristic. We also propose an ergodic-iterate-based local search method to further improve the solution quality. Lower bounds are improved by learning the relation between Lagrangian multipliers and electricity cost. Our extensive benchmarks show superior computational performance against commercial solvers. We achieved, on average, a 43% lower objective value compared to state-of-the-art methods. In 71% of the instances, we obtained near-optimal solutions (optimality gap less than 6%), and 93% of the instances were below 10%. We obtained feasible solutions for all instances, compared to only 65% feasibility using incumbent methods. The developed exact method aims to support future research on charging scheduling, especially important for micromobility industry, vehicle-to-grid (V2G) applications, and second-life utilization of batteries. Furthermore, the developed polyhedral insights can be useful in other scheduling problems with a common underlying mathematical structure.

math.OC↗

arXiv2025

YOLO-LLTS: Real-Time Low-Light Traffic Sign Detection via Prior-Guided Enhancement and Multibranch Feature Interaction

Traffic sign detection is essential for autonomous driving and Advanced Driver Assistance Systems (ADAS). However, existing methods struggle to address the challenges of poor image quality and insufficient information under low-light conditions, leading to a decline in detection accuracy and affecting driving safety. To address this issue, we propose YOLO-LLTS, an end-to-end real-time traffic sign detection algorithm specifically designed for low-light environments. YOLO-LLTS introduces three main contributions: the HRFM-SOD module retains more information about distant or tiny traffic signs compared to traditional methods; the MFIA module interacts features with different receptive fields to improve information utilization; the PGFE module enhances detection accuracy by improving brightness, edges, contrast, and supplementing detail information. Additionally, we construct a new dataset, the Chinese Nighttime Traffic Sign Sample Set (CNTSSS), covering diverse nighttime scenarios. Experiments show that YOLO-LLTS achieves state-of-the-art performance, outperforming previous best methods by 2.7% mAP50 and 1.6% mAP50:95 on TT100K-night, 1.3% mAP50 and 1.9% mAP50:95 on CNTSSS, 7.5% mAP50 and 9.8% mAP50:95 on GTSDB-night, and superior results on CCTSDB2021. Deployment on edge devices confirms its real-time applicability and effectiveness. The code and the dataset are available at https://github.com/linzy88/YOLO-LLTS.

cs.CV↗

arXiv2024

A GREAT Architecture for Edge-Based Graph Problems Like TSP

In the last years, an increasing number of learning-based approaches have been proposed to tackle combinatorial optimization problems such as routing problems. Many of these approaches are based on graph neural networks (GNNs) or related transformers, operating on the Euclidean coordinates representing the routing problems. However, such models are ill-suited for a wide range of real-world problems that feature non-Euclidean and asymmetric edge costs. To overcome this limitation, we propose a novel GNN-based and edge-focused neural model called Graph Edge Attention Network (GREAT). Using GREAT as an encoder to capture the properties of a routing problem instance, we build a reinforcement learning framework which we apply to both Euclidean and non-Euclidean variants of vehicle routing problems such as Traveling Salesman Problem, Capacitated Vehicle Routing Problem and Orienteering Problem. Our framework is among the first to tackle non-Euclidean variants of these problems and achieves competitive results among learning-based benchmarks.

cs.LG↗

arXiv2024

Less Is More -- On the Importance of Sparsification for Transformers and Graph Neural Networks for TSP

Most of the recent studies tackling routing problems like the Traveling Salesman Problem (TSP) with machine learning use a transformer or Graph Neural Network (GNN) based encoder architecture. However, many of them apply these encoders naively by allowing them to aggregate information over the whole TSP instances. We, on the other hand, propose a data preprocessing method that allows the encoders to focus on the most relevant parts of the TSP instances only. In particular, we propose graph sparsification for TSP graph representations passed to GNNs and attention masking for TSP instances passed to transformers where the masks correspond to the adjacency matrices of the sparse TSP graph representations. Furthermore, we propose ensembles of different sparsification levels allowing models to focus on the most promising parts while also allowing information flow between all nodes of a TSP instance. In the experimental studies, we show that for GNNs appropriate sparsification and ensembles of different sparsification levels lead to substantial performance increases of the overall architecture. We also design a new, state-of-the-art transformer encoder with ensembles of attention masking. These transformers increase model performance from a gap of $0.16\%$ to $0.10\%$ for TSP instances of size 100 and from $0.02\%$ to $0.00\%$ for TSP instances of size 50.

cs.LG↗

arXiv2023

Quantifying white matter hyperintensity and brain volumes in heterogeneous clinical and low-field portable MRI

Brain atrophy and white matter hyperintensity (WMH) are critical neuroimaging features for ascertaining brain injury in cerebrovascular disease and multiple sclerosis. Automated segmentation and quantification is desirable but existing methods require high-resolution MRI with good signal-to-noise ratio (SNR). This precludes application to clinical and low-field portable MRI (pMRI) scans, thus hampering large-scale tracking of atrophy and WMH progression, especially in underserved areas where pMRI has huge potential. Here we present a method that segments white matter hyperintensity and 36 brain regions from scans of any resolution and contrast (including pMRI) without retraining. We show results on eight public datasets and on a private dataset with paired high- and low-field scans (3T and 64mT), where we attain strong correlation between the WMH ($\rho$=.85) and hippocampal volumes (r=.89) estimated at both fields. Our method is publicly available as part of FreeSurfer, at: http://surfer.nmr.mgh.harvard.edu/fswiki/WMH-SynthSeg.

eess.IV↗

arXiv2023

Collaborative electric vehicle routing with meet points

In this paper, we develop a profit-sharing-based optimal routing mechanism to incentivize horizontal collaboration among urban goods distributors. This paper investigates a collaborative routing problem for urban logistics, in which the exchange of goods at meet points is optimally planned en route. We show that collaboration does not only reduce the total cost but also increases the profit of each company by sharing some customers and the related profit. Hence, we focus on solving a collaborative electric vehicle routing problem under constraints such as customer-specific time windows, opportunity charging, vehicle capacity, and meet-point synchronization. The proposed Collaborative Electric Vehicle Routing Problem with Meet Point (CoEVRPMP) is modeled as a mixed-integer nonlinear programming problem. We first present an exact method for optimal benchmarks via decomposition. To handle real-world problems, we suggest using a metaheuristic method: adaptive large neighborhood search with linear programming. The viability and scalability of the collaborative method are demonstrated via numerical case studies: (i) a real-world case of two grocery stores in the city of Gothenburg, Sweden, and (ii) a large-scale experiment with 500 customers. The results underline the importance of horizontal collaboration among delivery companies. Collaboration helps to reduce the environmental footprint (total energy consumed) and to increase the individual company's profit at the same time.

math.OC↗

arXiv2021

Deep Image-based Illumination Harmonization

Integrating a foreground object into a background scene with illumination harmonization is an important but challenging task in computer vision and augmented reality community. Existing methods mainly focus on foreground and background appearance consistency or the foreground object shadow generation, which rarely consider global appearance and illumination harmonization. In this paper, we formulate seamless illumination harmonization as an illumination exchange and aggregation problem. Specifically, we firstly apply a physically-based rendering method to construct a large-scale, high-quality dataset (named IH) for our task, which contains various types of foreground objects and background scenes with different lighting conditions. Then, we propose a deep image-based illumination harmonization GAN framework named DIH-GAN, which makes full use of a multi-scale attention mechanism and illumination exchange strategy to directly infer mapping relationship between the inserted foreground object and the corresponding background scene. Meanwhile, we also use adversarial learning strategy to further refine the illumination harmonization result. Our method can not only achieve harmonious appearance and illumination for the foreground object but also can generate compelling shadow cast by the foreground object. Comprehensive experiments on both our IH dataset and real-world images show that our proposed DIH-GAN provides a practical and effective solution for image-based object illumination harmonization editing, and validate the superiority of our method against state-of-the-art methods. Our IH dataset is available at https://github.com/zhongyunbao/Dataset.

cs.CV↗

arXiv2020

The Cooperative Sorting Strategy for Connected and Automated Vehicle Platoons

This paper presents a "cooperative vehicle sorting" strategy that seeks to optimally sort connected and automated vehicles (CAVs) in a multi-lane platoon to reach an ideally organized platoon. In the proposed method, a CAV platoon is firstly discretized into a grid system, where a CAV moves from one cell to another in the discrete time-space domain. Then, the cooperative sorting problem is modeled as a path-finding problem in the graphic domain. The problem is solved by the deterministic Astar algorithm with a stepwise strategy, where only one vehicle can move within a movement step. The resultant shortest path is further optimized with an integer linear programming algorithm to minimize the sorting time by allowing multiple movements within a step. To improve the algorithm running time and address multiple shortest paths, a distributed stochastic Astar algorithm (DSA) is developed by introducing random disturbances to the edge costs to break uniform paths (with equal path cost). Numerical experiments are conducted to demonstrate the effectiveness of the proposed DSA method. The results report shorter sorting time and significantly improved algorithm running time due to the use of DSA. In addition, we find that the optimization performance can be further improved by increasing the number of processes in the distributed computing system.

eess.SY↗