Design of Human-Machine Collaborative Decision Support System in Emergency Logistics Scenarios

Abstract

This paper focuses on the design of a human-machine collaborative decision support system (HMCDSS) for emergency logistics scenarios. Through systematic retrieval of literature from core databases such as CNKI and investigation of more than 10 typical emergency logistics system cases at home and abroad, it deeply analyzes the existing problems in current human-machine collaborative decision-making, including fuzzy function allocation, inefficient interaction interfaces, and fragmented decision processes. Based on ergonomics theory, an innovative design framework of "dynamic task allocation-multidimensional interaction optimization-closed-loop process iteration" is constructed: in function allocation, a Bayesian network-based dynamic human-machine task allocation model is established; the interaction interface uses eye-tracking technology to optimize information visualization layout; and the decision process integrates reinforcement learning algorithms to achieve intelligent iteration. The study expects to shorten the emergency logistics decision response time by more than 30% and improve the plan accuracy by 25%, providing theoretical support and practical paths for realizing the deep integration of human-machine advantages and constructing an efficient emergency logistics decision-making system.