人工智能的情境觉知:一种“在线”适应性表征

Situation Awareness in Artificial Intelligence:An Online Adaptive Representation

如何使机器人这种人工认知系统获得人类水平的智能,是目前人工智能发展的主要目标。要实现这个目标,人类智能的运作机制和模式无疑是人工智能模拟的最好模型。情境觉知理论的提出和发展,为实现这个目标提供了一种合理的框架和方法论。这就是关于情境觉知的通用模型、注意机制、情境觉知目标与心理模型的匹配要阐明的问题。“事件-控制-行动-适应性”模型是情境觉知理论的进一步发展,更突出了适应性表征特征。基于情境觉知理论开发的各种应用软件不仅是技术手段,更是方法论,体现在人工认知系统设计的方方面面。

How to make robot,an artificial cognitive system,acquire human-level intelligence is the main goal of artificial intelligence development at present.To achieve this goal,the operating mechanism and mode of human intelligence are undoubtedly the best model for artificial intelligence simulation.The development of situation awareness theory provides a reasonable framework and methodology for achieving this goal.This is the problem to be clarified about the matching of the general model,attention mechanism,situation awareness goals and psychological models.The event-control-action-adaptation model is the further development of situation awareness theory and highlights the characteristics of adaptive representation.All kinds of application software based on situational awareness theory are not only technical means,but also methodology,which are embodied in all aspects of artificial cognitive system design.