Human performance modeling and its uncertainty factors affecting decision making: a survery

Abstract

This paper introduces the background and connotation of human performance modeling (HPM), HPM models, and the application of artificial intelligence algorithms in HPM. It deeply analyzes the connotation and uncertainty of each model and finally puts forward its military application. The aim is to provide relevant researchers in the field with an in-depth understanding of domain knowledge and related uncertainties and to indicate future research directions. The first part is a general overview of human factors engineering, where the definition, origin, research field, importance, and general problems of HPM are elaborated. The composition of the man–machine system and its corresponding relationship with the observe–orient–decide–act loop are described. The second part reviews the models of perception, cognition, understanding, and decision making. Among them, models of cognition consist of visual search, visual sampling, mental workload, and goals, operators, methods, and selection rules; models of action consist of Hick–Hyman law, Fitts’s law, and manual control theory. The third part is a review of the source and importance of the integrated models and focuses on the principles, composition, and successful application cases of the three models, namely SAINT, IMPRINT, and ACT-R. The fourth part is a review of the application of the algorithms and models in the fields of artificial intelligence, deep learning, and data mining in analyzing multivariate datasets in HPM. In addition, future HPM military applications are presented.