人的可靠性对航天任务的成败起关键作用。基于认知、人格等个体特征的筛选是选拔航天员的重要环节。但筛选标准大都来自专家判断，缺乏实证研究。本论文致力于研究个体特征对操作类航天任务绩效的影响。通过文献调研，总结航天任务的分类、特点和绩效影响因素，以及认知特征和人格特质的研究现状。同时，根据国际空间站内典型的机械臂遥操作任务，抽取出“点对准”、“线对准”和“面对准”三种基本任务。最后，建立研究框架，设计实验任务和场景，开展四项实验研究。 研究一基于飞船仪表盘训练台，以应急规程操作为对象，探索了认知能力和认知风格对操作绩效的影响。实验结果显示，认知能力显著影响完成时间和操作失误次数。参试者越倾向于序列型，学习得越快，操作失误越少。此外，随着经验增加，认知能力对绩效的预测减弱，而序列/整体认知风格维度仍能显著预测绩效。 研究二以“点对准”、“线对准”和避障三种机械臂遥操作为对象，探索了空间能力、认知风格以及“大五”人格特质对遥操作绩效的影响。实验结果显示，空间能力显著影响遥操作绩效。参试者越倾向于表象型，任务绩效和安全绩效越好。参试者越倾向于整体型，安全绩效越优。此外，外向特质显著提高任务绩效但会降低安全绩效，高情绪稳定性和亲和力在遥操作中体现出优势。 研究三以“面对准”的机械臂遥操作为对象，探索了认知特征与辅助信息显示方式对遥操作绩效的交互影响。实验结果显示，空间能力和辅助信息显著提高任务绩效和安全绩效。表象型个体的工作负荷显著低于言语型个体。此外，图形式辅助信息对表象型个体的帮助更大，而数值型辅助信息对言语型个体的帮助更大。数值型辅助信息能够降低言语型个体的工作负荷而增加表象型个体的工作负荷。 研究四以机器车遥操作为对象，探索了认知能力、人格特质的绩效影响机制。实验结果显示，空间能力或能通过空间认知间接影响遥操作绩效，且对遥操作绩效的直接影响显著；同时，经验开放性或能通过空间认知对遥操作绩效产生影响。此外，发现在完全陌生的环境中，情绪敏感性对个体的空间认知发挥主要作用；而在对陌生环境的熟悉过程中，经验开放性对个体的空间认知发挥主要作用。 本论文的一系列研究展示了个体特征对规程操作和遥操作绩效的影响，部分验证了认知能力、认知风格和人格特质的绩效影响模型，对航天员选拔和培训具有重要指导作用，对自适应性操作界面设计具有启发意义。

Human reliability is critical to the success of spaceflight tasks. Screening based on individual characteristics is an important part for astronauts’ selection. However, the screening criteria on cognitive characteristics and personality traits are mainly based on expert judgment, with the lack of support from empirical studies. Therefore, this dissertation aimed to explore the influence of cognitive characteristics and personality traits on the performance of different spaceflight tasks. The literature was reviewed from the classification, features, and performance influencing factors of spaceflight tasks, and the state of the art of researches on cognitive characteristics and personality traits. Base on the review a research framework was proposed. Besides, three basic tasks (point aiming, line alignment, cross aiming) were constructed based on the characteristic analysis of several typical robotic arm tasks in the International Space Station. Finally four experimental studies were designed and conducted. Based on a Chinese spaceflight instrument board, the first study explored the effects of cognitive ability and cognitive style on task performance of simulated space-flight emergency operations, and the possible interaction effects with training experience. The results revealed that cognitive ability significantly influenced task completion time and number of operation errors. Compared participants with a wholist cognitive style, those with a sequential cognitive style learned faster, and made fewer errors. Besides, with the increasing of training experience, the predictive capability of cognitive ability on performance decreased, whereas the predictive capability of the sequential/global cognitive style on performance increased. The relationship between individual characteristics (spatial ability, cognitive style, personality traits) and performance of three teleoperation tasks (point aiming, line alignment, and obstacle avoidance) was investigated in the second study. The results revealed that spatial ability significantly influenced teleoperation performance. Imagery participants performed better than verbal participants regarding both task and safety performance. Compared with the analytic participants, the wholist participants had better safety performance. Additionally, extraverted participants had better task performance but worse safety performance. Low neuroticism and high agreeableness are beneficial to teleoperation performance. In the third study, two types of additional spatial information were designed for a cross aiming task, and the interaction effects between cognitive characteristics and additional spatial information on teleoperation performance were examined. The results indicated that spatial ability significantly enhanced task and safety performance. Additional spatial information significantly improved safety performance for both imagery and verbal participants. Compared to verbal participants, imagery participants made fewer collisions under interface with graphic additional spatial information but more under interface with numerical additional spatial information. Besides, imagery participants’ workload was lower than verbal participants. Numerical additional spatial information decreased verbal participants’ workload but increased imagery participants’ workload. The fourth study aimed to explore how cognitive ability and personality traits could affect teleoperation performance. The results showed that spatial ability might affect teleoperation performance through spatial cognition, and significantly directly affected teleoperation performance. Openness to experience might affect teleoperation performance through spatial cognition. In addition, it was found that neuroticism played a key role for the participants to build up spatial cognition of a completely strange environment, and openness to experience played a key role for the participants to improve spatial cognition during the process of becoming familiar with a strange environment. The aforementioned studies demonstrate that individual characteristics can significantly influence procedure operation and teleoperation performance. They partially verify the performance influencing models of cognitive ability, cognitive style, and personality traits. The findings could be helpful for astronaut selection and training, and provide implications for adaptive interface design.