Reliability Analysis and Evaluation Method of Harvesting Robots Based on Multi-Agent System

• Published in: 2025 2nd International Conference on Electrical Technology and Automation Engineering (ETAE) pp. 351 - 355
• Main Authors: Wan, Wenfei, Huang, Qiang, Li, Xiaobing, Xu, Pengfei
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23.05.2025
• Subjects: Collaboration; coupling condition probability; Couplings; harvesting robot; Measurement; Multi-Agent System (MAS); Multi-agent systems; multi-dimensional metrics; Reliability; reliability assessment; Reliability engineering; Robots; Stability analysis; Testing; Thermal stability
• DOI: 10.1109/ETAE65337.2025.11089763

To address the critical challenges of low reliability, poor environmental adaptability, and insufficient efficiency in harvesting robots, this paper proposes an innovative Multi-Agent System (MAS)-based reliability evaluation method. The method introduces the MAS modeling to quantify multi-dimensional reliability metrics. By analyzing the collaborative mechanisms among perception, control, and execution subsystems during complex harvesting tasks, a mission reliability block diagram (RBD) is established to integrate hardware-software fault coupling analysis. By quantifying interdependencies among agents and their collective impact on system stability, the method comprehensively evaluates the harvesting task reliability of harvesting robots in complex agricultural environments. The experimental results demonstrate that the proposed approach can be used to evaluate the reliability of harvesting robots, which provides actionable insights for fault-tolerant design and adaptive control optimization, advancing the practical deployment of intelligent agricultural robotics.