Author name cluster

Xiaonan Wang

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

5 papers

2 author rows

AAAI Conference 2026 Conference Paper

CycleChemist: A Dual-Pronged Machine Learning Framework for Organic Photovoltaic Discovery

Hou Hei Lam
Jiangjie Qiu
Xiuyuan Hu
Wentao Li
Fankun Zeng
Siwei Fu
Hao Zhang
Xiaonan Wang

Organic photovoltaic (OPV) materials offer a promising pathway for sustainable energy generation. However, their development is hindered by the challenge of identifying high-performance donor-acceptor pairs with optimal power conversion efficiencies (PCEs). Most existing design strategies focus exclusively on either the donor or the acceptor, rather than employing a unified model capable of designing both components. In this work, we introduce a dual-pronged machine learning framework for OPV discovery, integrating predictive modeling and generative molecular design. In this study, we propose the newly curated Organic Photovoltaic Donor-Acceptor Dataset (OPV²D), the largest of its kind, comprising 2,000 experimentally characterized donor-acceptor pairs. This dataset serves as a comprehensive foundation for model training and evaluation. To enable accurate property prediction in organic photovoltaic (OPV) materials, we first introduce the Organic Photovoltaic Classifier (OPVC) to predict the likelihood that a given material exhibits OPV behavior. Complementing this, we develop a hierarchical graph neural network framework that integrates multi-task learning and cross-modal donor–acceptor interaction modeling. This framework includes the Molecular Orbital Energy Estimator (MOE²) for predicting the highest occupied molecular orbital–lowest unoccupied molecular orbital (HOMO–LUMO) energy levels, and the Photovoltaic Performance Predictor (P³) for estimating power conversion efficiency (PCE). In addition, we introduce the Material Generative Pretrained Transformer (MatGPT) to generate synthetically accessible organic semiconductors. Building on this, we propose a reinforcement learning strategy with three-objective policy optimization that guides molecular generation while preserving chemical validity. By bridging molecular representation learning with device performance prediction, our framework advances computational OPV material discovery.

PDF Details DOI

EAAI Journal 2026 Journal Article

Event-driven emergency data dissemination in vehicular named data networking based on learning automata

Xiaonan Wang
Ranran Zhang

Emergency road-safety data, such as accident notifications, can improve driving safety and efficiency, making rapid access to this data crucial. This paper proposes an event-driven emergency data dissemination method based on vehicular named data networking. The method has three key ideas: (1) When an event occurs, dissemination is triggered, leveraging a local cloud for inter-vehicle collaboration to construct a comprehensive dataset; (2) Learning automata are employed to stabilize dissemination paths; and (3) A subscription mechanism is introduced to enable path reuse for continuous, unicast-based dissemination to interested vehicles. The experimental results show the method reduces dissemination delays and overhead by nearly 40. 5 % and 68. 1 %, respectively, while improving success rates by nearly 11. 3 %.

Details DOI

EAAI Journal 2025 Journal Article

Early warning for fault-slip risk based on genetic algorithm-back propagation neural network: An intelligent mathematical evaluation model study

Hongxu Shi
Jianpo Liu
Shuxiang Wei
Song Cui
Xiaonan Wang

Excavation-induced stress disrupts fault equilibrium, significantly increasing the fault-slip risk. Currently, the limited fault-slip cases result in inadequate warning methodologies. Therefore, this study demonstrates the application of genetic algorithm-back propagation neural network (GA-BPNN) for fault-slip risk prediction by integrating monitoring data with numerical simulation. First, a numerical model is established in a deep-buried tunnel fault-slip case. This model analyzes evolution characteristics throughout the fault-slip process, revealing coordinated precursory signatures, including shear stress drop, total cracks burst, and shear crack mutation during catastrophic evolution. Then, a fault-slip case database is developed using a full-variable approach that incorporates dip angle, stress level, and friction coefficient. The model is trained on these precursors as inputs, with fault-slip risk level as the output. The model architecture is further optimized using genetic algorithm to enhance performance. The proposed model achieves a prediction accuracy of over 90 %, demonstrating strong generalization capability for fault-slip risk early warning. This research provides both theoretical foundations for stability monitoring and slip hazard prevention in deep rock mass engineering involving fault structures.

Details DOI

IROS Conference 2024 Conference Paper

Accurate and Efficient Loop Closure Detection With Deep Binary Image Descriptor and Augmented Point Cloud Registration

Jialiang Wang
Zhi Gao 0005
Zhipeng Lin
Zhiyu Zhou
Xiaonan Wang
Jianhua Cheng
Hao Zhang
Xinyi Liu 0002

Loop Closure Detection (LCD) is an essential component of Simultaneous Localization and Mapping (SLAM), helping to correct drift errors, facilitate map merging, or both by identifying previously observed scenes. Despite its importance, traditional LCD algorithms based on single sensor such as camera or LiDAR exhibit degraded performance in challenging scenarios due to their inherent limitations. To address this issue, we propose a novel LCD method based on camera-LiDAR fusion, exploiting the rich textural information from cameras and the accurate geometric data from LiDAR to ensure robustness and speed in challenging environments. Specifically, we first employ deep hashing learning to encode deep image features into binary image descriptors for extremely fast loop candidate (LC) retrieval. Then, LiDAR points are augmented with image color for accurate geometric verification. Finally, we incorporate a spatial-temporal consistency check that mandates an LC to have consistently matched neighbors to be accepted as true. Our method is extensively verified and compared with the state-of-the-art methods on various datasets encompassing both indoor and outdoor environments. Experimental results demonstrate that our method obtains the best performance, increasing the maximum recall rate at 100% precision by a significant margin of 20% while operating in real-time at an average speed of 30 fps.

Details

EAAI Journal 2024 Journal Article

Learning automata based routing and content delivery for vehicular named data networking

Xiaonan Wang
Gaoyang Wu

In Vehicular Named Data Networking (VNDN), vehicle mobility leads to frequent reverse-path interruptions and stale forwarding information, which further results in content delivery failures and disenables aggregation. Taking into account these limitations, we propose a learning automata based routing and content delivery solution for VNDN. The novelties of the solution are threefold: (1) Learning automata is leveraged to make forwarding decisions and deliver vehicular contents; (2) Various metrics including connection durations, unsatisfied requests and data response delays are utilized to compute the selection probability of each vehicle in order to select the best vehicle for content delivery; and (3) Aggregation and in-network caching are achieved through stable reverse paths. The typical application of the proposal is that vehicles rapidly access nearby road conditions to maneuver safe driving. The proposal is evaluated quantitatively. Compared with the existing solutions, the proposal reduces the content delivery latency and overheads by nearly 32. 54% and 38. 3%, respectively.

Details DOI