Peng Yi

Dixiao Wei, Peng Yi, Jinlong Lei et al.

Reinforcement Learning (RL) has demonstrated excellent decision-making potential in platoon coordination problems. However, due to the variability of coordination goals, the complexity of the decision problem, and the time-consumption of trial-and-error in manual design, finding a well performance reward function to guide RL training to solve complex platoon coordination problems remains challenging. In this paper, we formally define the Platoon Coordination Reward Design Problem (PCRDP), extending the RL-based cooperative platoon coordination problem to incorporate automated reward function generation. To address PCRDP, we propose a Large Language Model (LLM)-based Platoon coordination Reward Design (PCRD) framework, which systematically automates reward function discovery through LLM-driven initialization and iterative optimization. In this method, LLM first initializes reward functions based on environment code and task requirements with an Analysis and Initial Reward (AIR) module, and then iteratively optimizes them based on training feedback with an evolutionary module. The AIR module guides LLM to deepen their understanding of code and tasks through a chain of thought, effectively mitigating hallucination risks in code generation. The evolutionary module fine-tunes and reconstructs the reward function, achieving a balance between exploration diversity and convergence stability for training. To validate our approach, we establish six challenging coordination scenarios with varying complexity levels within the Yangtze River Delta transportation network simulation. Comparative experimental results demonstrate that RL agents utilizing PCRD-generated reward functions consistently outperform human-engineered reward functions, achieving an average of 10\% higher performance metrics in all scenarios.

Weijie Zhou, Xuantang Xiong, Yi Peng et al.

Visual reasoning in multimodal large language models (MLLMs) has primarily been studied in static, fully observable settings, limiting their effectiveness in real-world environments where information is often incomplete due to occlusion or limited field of view. Humans, in contrast, actively explore and interact with their environment-moving, examining, and manipulating objects-to gather information through a closed-loop process integrating perception, reasoning, and action. Inspired by this human capability, we introduce the Active Visual Reasoning (AVR) task, extending visual reasoning to partially observable, interactive environments. AVR necessitates agents to: (1) actively acquire information via sequential physical actions, (2) integrate observations across multiple steps for coherent reasoning, and (3) dynamically adjust decisions based on evolving visual feedback. To rigorously evaluate AVR, we introduce CLEVR-AVR, a simulation benchmark featuring multi-round interactive environments designed to assess both reasoning correctness and information-gathering efficiency. We present AVR-152k, a large-scale dataset that offers rich Chain-of-Thought (CoT) annotations detailing iterative reasoning for uncertainty identification, action-conditioned information gain prediction, and information-maximizing action selection, crucial for training agents in a higher-order Markov Decision Process. Building on this, we develop PhysVLM-AVR, an MLLM achieving state-of-the-art performance on CLEVR-AVR, embodied reasoning (OpenEQA, RoboVQA), and passive visual reasoning (GeoMath, Geometry30K). Our analysis also reveals that current embodied MLLMs, despite detecting information incompleteness, struggle to actively acquire and integrate new information through interaction, highlighting a fundamental gap in active reasoning capabilities.

Yang Lv, Jinlong Lei, Peng Yi

Two-stage Colonel Blotto game represents a typical adversarial resource allocation problem, in which two opposing agents sequentially allocate resources in a network topology across two phases: an initial resource deployment followed by multiple rounds of dynamic reallocation adjustments. The sequential dependency between game stages and the complex constraints imposed by the graph topology make it difficult for traditional approaches to attain a globally optimal strategy. To address these challenges, we propose a hierarchical graph Transformer framework called HGformer. By incorporating an enhanced graph Transformer encoder with structural biases and a two-agent hierarchical decision model, our approach enables efficient policy generation in large-scale adversarial environments. Moreover, we design a layer-by-layer feedback reinforcement learning algorithm that feeds the long-term returns from lower-level decisions back into the optimization of the higher-level strategy, thus bridging the coordination gap between the two decision-making stages. Experimental results demonstrate that, compared to existing hierarchical decision-making or graph neural network methods, HGformer significantly improves resource allocation efficiency and adversarial payoff, achieving superior overall performance in complex dynamic game scenarios.

Dixiao Wei, Peng Yi, Jinlong Lei et al.

Truck platooning technology enables a group of trucks to travel closely together, with which the platoon can save fuel, improve traffic flow efficiency, and improve safety. In this paper, we consider the platoon coordination problem in a large-scale transportation network, to promote cooperation among trucks and optimize the overall efficiency. Involving the regulation of both speed and departure times at hubs, we formulate the coordination problem as a complicated dynamic stochastic integer programming under network and information constraints. To get an autonomous, distributed, and robust platoon coordination policy, we formulate the problem into a model of the Decentralized-Partial Observable Markov Decision Process. Then, we propose a Multi-Agent Deep Reinforcement Learning framework named Trcuk Attention-QMIX (TA-QMIX) to train an efficient online decision policy. TA-QMIX utilizes the attention mechanism to enhance the representation of truck fuel gains and delay times, and provides explicit truck cooperation information during the training process, promoting trucks' willingness to cooperate. The training framework adopts centralized training and distributed execution, thus training a policy for trucks to make decisions online using only nearby information. Hence, the policy can be autonomously executed on a large-scale network. Finally, we perform comparison experiments and ablation experiments in the transportation network of the Yangtze River Delta region in China to verify the effectiveness of the proposed framework. In a repeated comparative experiment with 5,000 trucks, our method average saves 19.17\% of fuel with an average delay of only 9.57 minutes per truck and a decision time of 0.001 seconds.

Zhaojian Wang, Shengwei Mei, Feng Liu et al.

Forming (hybrid) AC/DC microgrids (MGs) has become a promising manner for the interconnection of various kinds of distributed generators that are inherently AC or DC electric sources. This paper addresses the distributed asynchronous power control problem of hybrid microgrids, considering imperfect communication due to non-identical sampling rates and communication delays. To this end, we first formulate the optimal power control problem of MGs and devise a synchronous algorithm. Then, we analyze the impact of asynchrony on optimal power control and propose an asynchronous iteration algorithm based on the synchronous version. By introducing a random clock at each iteration, different types of asynchrony are fitted into a unified framework, where the asynchronous algorithm is converted into a fixed-point problem based on the operator splitting method, leading to a convergence proof. We further provide an upper bound estimation of the time delay in the communication. Moreover, the real-time implementation of the proposed algorithm in both AC and DC MGs is introduced. By taking the power system as a solver, the controller is simplified by reducing one order and the power loss can be considered. Finally, a benchmark MG is utilized to verify the effectiveness and advantages of the proposed algorithm.

Yutao Tang, Peng Yi

In this paper, a multi-agent coordination problem with steady-state regulation constraints is investigated for a class of nonlinear systems. Unlike existing leader-following coordination formulations, the reference signal is not given by a dynamic autonomous leader but determined as the optimal solution of a distributed optimization problem. Furthermore, we consider a global constraint having noisy data observations for the optimization problem, which implies that reference signal is not trivially available with existing optimization algorithms. To handle those challenges, we present a passivity-based analysis and design approach by using only local objective function, local data observation and exchanged information from their neighbors. The proposed distributed algorithms are shown to achieve the optimal steady-state regulation by rejecting the unknown observation disturbances for passive nonlinear agents, which are persuasive in various practical problems. Applications and simulation examples are then given to verify the effectiveness of our design.