ET-HF: A novel information sharing model to improve multi-agent cooperation

Article Computer Science, Artificial Intelligence

Bicriteria Sparse Nonnegative Matrix Factorization via Two-Timescale Duplex Neurodynamic Optimization

Hangjun Che et al.

Summary: In this article, the sparse nonnegative matrix factorization problem is formulated as a mixed-integer bicriteria optimization problem. A two-timescale duplex neurodynamic approach is used to solve the problem, achieving low error, high sparsity, and high score.

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2023)

Add to Collection

Article Computer Science, Artificial Intelligence

Multi-objective particle swarm optimization with multi-mode collaboration based on reinforcement learning for path planning of unmanned air vehicles

Xiangyin Zhang et al.

Summary: This paper proposes a multi-objective particle swarm optimization algorithm based on reinforcement learning (MCMOPSO-RL) to solve the collaborative path planning problem for multiple unmanned aerial vehicles (UAVs) in complex environments. The experimental results show that this algorithm can solve the path planning problem for multiple UAVs more efficiently and robustly.

KNOWLEDGE-BASED SYSTEMS (2022)

Add to Collection

Article Automation & Control Systems

Intelligent Knowledge Distribution: Constrained-Action POMDPs for Resource-Aware Multiagent Communication

Michael C. Fowler et al.

Summary: This article addresses the question of multiagent knowledge distribution and introduces the concepts of action-based constraints and soft probabilistic constraint satisfaction in the POMDPs framework. The paper proposes an intelligent knowledge distribution (IKD) model and validates its performance in an asset tracking problem.

IEEE TRANSACTIONS ON CYBERNETICS (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Multi-Agent Path Finding with heterogeneous edges and roundtrips

Bing Ai et al.

Summary: This paper introduces a new version of Multi-Agent Path Finding (MAPF) where agents need to complete tasks and return to their starting locations on a graph with non-uniform edge costs. Through the study of a novel algorithm, it is shown that the algorithm outperforms the state-of-the-art algorithms for dealing with MAPF-HR.

KNOWLEDGE-BASED SYSTEMS (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

Gated multi-attention representation in reinforcement learning

Dayang Liang et al.

Summary: Deep reinforcement learning has shown great success by combining deep learning with reinforcement learning techniques. Using Convolutional Neural Networks as feature extraction and incorporating attention mechanism can greatly improve the performance of DRL algorithms.

KNOWLEDGE-BASED SYSTEMS (2021)

Add to Collection

Article Computer Science, Artificial Intelligence

A semi-decentralized feudal multi-agent learned-goal algorithm for multi-intersection traffic signal control

Shantian Yang et al.

Summary: The importance of traffic signal control and the use of hierarchical deep reinforcement learning to improve its performance are introduced. The LSAC algorithm and SFM framework are proposed to address the issue of non-optimal latent goals and the fast-growing state space in multi-agent systems, respectively. The combination of these two methods as the SFM-LSAC algorithm shows superior performance in multi-intersection traffic signal control compared to existing algorithms.

KNOWLEDGE-BASED SYSTEMS (2021)

Add to Collection

Article Automation & Control Systems

Adaptive Fault-Tolerant Tracking Control for Discrete-Time Multiagent Systems via Reinforcement Learning Algorithm

Hongyi Li et al.

Summary: This article investigates the adaptive fault-tolerant tracking control problem for a class of discrete-time multiagent systems via a reinforcement learning algorithm. The direct adaptive optimal controllers are designed by combining the backstepping technique with the reinforcement learning algorithm to reduce computational burden, and adaptive auxiliary signals are established to compensate for the influence of dead zones and actuator faults.

IEEE TRANSACTIONS ON CYBERNETICS (2021)

Add to Collection

Article Automation & Control Systems

Emotion Correlation Mining Through Deep Learning Models on Natural Language Text

Xinzhi Wang et al.

Summary: Previous research in artificial intelligence has primarily focused on recognizing emotions rather than exploring reasons for incorrect emotional recognition. Through the analysis of natural language text from web news, the study reveals correlations among emotions and potential cognitive biases. Findings suggest that in subjective comments, emotions are easily mistaken for anger.

IEEE TRANSACTIONS ON CYBERNETICS (2021)

Add to Collection

Article Automation & Control Systems