基于任务的复杂人机交互系统操纵适配性度量与优化

Large scale complex industrial and weapon systems, such as high speed trains and military ships, greatly increase the complexity of man-machine interaction operation, which brings new challenges to efficiency of human work and safety of system operation. Enhancing the compatibility of man-machine interaction can effectively improve the safety and efficiency of the system. Research on measurement method of operation compatibility and optimization technology of complex man-machine interactive system is a critical work to be carried out in the field of ergonomics. A novel measurement and optimization method of complex man-machine interactive system operation compatibility based on task is presented. This method builds a task-device-operation hierarchical index model through the analysis of the operation task of complex man-machine system, and establishes measurement index system of compatibility. According to the measurement index system, a fuzzy comprehensive evaluation model is constructed based on fuzzy theory, hence the task based measurement of compatibility is built. Uncertain task mode is studied subsequently, and the measurement model of uncertain task is established. The measurement model of uncertain task combines with the grid model and hyper geometric constraint, which constructs a multi-objective optimization model taking known tasks and uncertain tasks as optimization targets. Hybrid genetic simulated annealing algorithm is adopted to solve the optimization model to obtain the optimization layout scheme of complex man-machine interactive system.

高速列车、军用舰船等大规模的复杂工业与武器系统大大提升了人机交互操作的复杂程度，给人员的作业效率和系统的安全运行带来了新的挑战。人机交互适配性的提升可以有效的提高系统的安全性和高效性，复杂人机交互系统的操纵适配性度量方法与优化技术的研究是人机工程领域亟待开展的一项重要工作。本项目提出一种基于任务的复杂人机交互系统操作适配性度量方法和优化技术。该方法通过对复杂人机系统操作任务的分析，构建任务-器件-作业层级指标模型，建立适配性度量指标体系。在度量指标体系的基础上，基于模糊理论构建适配性模糊综合评价模型，建立基于任务的适配性度量方法。研究不确定任务的任务模式，建立不确定任务的度量模型，结合格点模型和超几何约束，建立以已知任务和不确定任务为优化目标的多目标优化模型，并结合混合遗传-模拟退火算法对模型进行求解，实现复杂人机交互系统的器件布局优化。

高速列车、军用舰船等大规模的复杂工业与武器系统大大提升了人机交互操作的复杂程度，给人员的作业效率和系统的安全运行带来了新的挑战。本项目以轨道交通列车驾驶界面、轨道交通行车调度界面为复杂人机交互系统操纵适配性的研究对象，开展了人机交互系统操纵适配性度量指标与评价模型研究，构建了基于任务的高速列车驾驶界面适配性评价指标体系，体系共分为任务层、器件层、作业类三个层级，基于模糊粗糙集理论提出一种主客观结合的器件层层级权重确定方法，在此基础上提出基于模糊综合评价的高速列车驾驶界面适配性评价模型；开展了复杂人机交互系统布局优化模型研究，提出了适用于器件面积不等的显控界面布局整数规划模型，提出了一种结合邻域搜索的多目标粒子群算法；在上述研究工作中，为了更深入的理解复杂人机交互系统任务特征与规律，开展了基于认知工程的认知工作分析理论研究；为了更精确描述人机交互过程中作业人员状态，开展了高速列车司机控制器手柄工效、基于Petri网的轨道交通驾驶人员行为/资源分配-工作负荷度量与预测的研究；针对理论应用过程中可能存在的操作者注意分配问题，开展了广视野域监控任务监控人员注意力分配模型构建的研究。本项目研究成果可为复杂人机交互系统界面设计提供理论依据，对提高高速列车、轨道交通行车调度界面、舰船指挥系统等复杂人机交互系统的安全性和操作绩效具有重要的意义。