Abstract: To identify the differences of the importance of human-machine interaction information in the cockpit of deep-sea manned submersible during complex mission,a complex network-based human-machine interaction information importance assessment method was proposed.Taking the hover sampling task as the research object,the decision ladder model and task network modeling techniques were applied to identify the human-machine interaction information elements and logical associations related to the complex task,and a complex network model of the interaction information was constructed.Based on the network topology feature parameters,the node global and local influence effect assessment indexes were proposed from the perspective of the logical influence relationship between nodes,and the integrated value of interaction information node importance was calculated by combining the interpretive structural model hierarchical weight indexes to derive the hover sampling task human-computer interaction information importance ranking.The ranking results were compared with the other four algorithms,and the submariners were invited to evaluate the subjective importance of the interaction information in the context of the actual mission process based on the proof that the calculation was valid.The results showed that the importance ranking results obtained from the study were strongly correlated with the subjective ranking results of the submariners,and the identification of important interaction information elements was consistent with the subjective perception of the submariners,which validated the effectiveness of the proposed method to identify the differences of the importance of human-machine interaction information for complex tasks.

Key words: manned submersible, human machine interaction, information importance, complex network, interpretive structural modeling

摘要： 为更好地识别复杂任务环境下深海载人潜水器驾驶舱人机交互信息重要度差异,提出一种基于复杂网络的人机交互信息重要度评估方法。以悬停采样任务为研究对象,运用决策阶梯模型及任务-网络建模技术识别复杂任务相关人机交互信息元及逻辑关联,构建交互信息复杂网络模型。基于网络拓扑特征参数,从节点间逻辑影响关系的角度提出节点全局及局部影响效应评估指标,结合解释结构模型层级权重指标对交互信息节点重要度综合值进行计算,得出悬停采样任务人机交互信息重要度排序。将排序结果与其他4种算法进行对比,在证明计算有效的基础上,邀请潜航员结合实际任务流程对交互信息进行主观重要度评估。结果显示,研究所得重要度排序结果与潜航员主观排序结果呈强相关,且对重要交互信息元的识别结果与潜航员主观感知相符合,验证了所提方法识别复杂任务人机交互信息重要度差异的有效性。

关键词: 载人潜水器, 人机交互, 信息重要性, 复杂网络, 解释结构模型