Haipeng Ren

Haipeng Ren


Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 June 2025 in Electronics
Verified
Journal Article
Open access
PDF file
BYBo Yang
Bo Yang
LGLinghang Gao
Linghang Gao
FZFangzheng Zhou
Fangzheng Zhou
HYHongge Yao
Hongge Yao
YFYan-fang Fu
Yan-fang Fu
HRHaipeng Ren
Haipeng Ren
Cooperative multi-agent reinforcement learning (MARL) has emerged as a powerful paradigm for addressing complex real-world challenges, including autonomous robot control, strategic decision-making, and decentralized coordination in unmanned swarm systems. However, it still faces challenges in learning proper coordination among multiple agents. The lack of effective knowledge sharing and experience interaction mechanisms among agents has led to substantial performance decline, especially in terms of low sampling efficiency and slow convergence rates, ultimately constraining the practical applicability of MARL. To address these challenges, this paper proposes a novel framework termed Reward redistribution and Experience reutilization based Coordination Optimization (RECO). This innovative approach employs a hierarchical experience pool mechanism that enhances exploration through strategic reward redistribution and experience reutilization. The RECO framework incorporates a sophisticated evaluation mechanism that assesses the quality of historical sampling data from individual agents and optimizes reward distribution by maximizing mutual information across hierarchical experience trajectories. Extensive comparative analyses of computational efficiency and performance metrics across diverse environments reveal that the proposed method not only enhances training efficiency in multi-agent gaming scenarios but also significantly strengthens algorithmic robustness and stability in dynamic environments.

2 May 2025 in Research Square
Verified
Preprint
Open access
PDF file
JWJin Wang
Jin Wang
XSXunhang Shang
Xunhang Shang
BYBo Yang
Bo Yang
HRHaipeng Ren
Haipeng Ren
HJHong Jiang
Hong Jiang
WLWei Li
Wei Li
As a typical meta-learning method, model-agnostic meta-learning has attracted widespread attention in the remote sensing field due to its flexibility and applicability to various problems. However, the diversity and uniqueness of meta-tasks are constrained by the influence of remote sensing datasets from different imaging devices, resolutions, and diverse application requirements, which affects the algorithm’s ability to handle different tasks. To address this issue, this paper introduces a new concept called ”central points” starting from the distribution of meta-parameters, within a variational meta-learning framework [7](VML). Based on the concept of central points, we propose a variational meta-learning framework (VML) to obtain ”task central points” through variational inference and generate model parameters from these task central points, enabling the model to adapt to different tasks. To evaluate the utility of VML, experiments were conducted on three public datasets: UC Merced, WHU-RS19, and NWPU-RESISC45. The experimental results demonstrate that compared to some common meta-learning methods, VML effectively addresses the issues of 1 task diversity and heterogeneity in remote sensing datasets, and improves the classification accuracy of remote sensing datasets.