Presented is a human factor risk model when piloting an aircraft. This model is based on comparing representations of the evaluated crew actions with the comparable action representations of various types and performance quality, which form a representative sample and are contained in a pre-formed specialized database. The risk in question is represented by probabilistic estimates, which result from consistent applications of the Principal Component Analysis, Multidimensional Scaling, and Cluster Analysis to three types of characteristics, viz.: parameters of flights and states of aircraft systems, gaze movement trajectories and time series of oculomotor activity primary indexes. These steps form the clusters of flight fragments for various types and performance quality, including abnormal ones. The Discriminant Analysis provides calculating the probabilistic profile for belonging to certain target clusters, with a final conclusion being derived from this structure. Key elements of the approach presented are three new metrics used to compare crew actions and to ensure significant discrimination for flight fragments of various types and performance quality. Detailing flight parameters contributions in differences of the flight fragments in a given metric is carried out to provide meaningful analysis of the detected abnormality causes. With sufficient computational performance, the flight data analysis under consideration can be implemented in real time automatic mode.
Mathematical models and methods for crew training level assessing based on video oculography data are presented. The results obtained are based on comparing the studied fragments of oculomotor activity of pilots with comparable patterns of video oculography data of various types and performance quality contained in a pre-formed specialized database. To obtain estimates, a complex combination of random process analysis and multivariate statistical analysis is used. The “intelligence” of diagnostic tools is contained in empirical data and can flexibly change as they accumulate. The considered example of determining the flight mode and pilot qualification based on video oculography data allows us to talk about the possibility of significant discrimination of the gaze movement trajectories of pilots at different flight phases and significant discrimination of the gaze movement trajectories of experienced and inexperienced pilots at certain phases of flight. An important new component of the presented results is a discriminant analysis for solving the problem of flight exercises classification, based on the principles of quantum computing. The scope of the considered approach is not limited to aviation applications and can be extended to tasks that are similar in content.
Presented are new approaches for supporting the outcome grading for activities of operators of complex technical systems, which are based on comparisons of current exercises with the activity database patterns in both the wavelet representation metric associated with time series of activity parameters and the likelihood metric of eigenvalue trajectories for these parameters transforms as well as on probabilistic assessments of skill class recognition using sample distribution functions of exercise distances to cluster centers in a scaling space and Bayesian likelihood estimations with the aid of probabilistic profile of staying in activity parameter ranges. These techniques have demonstrated the capabilities of recognizing sets of abnormal exercises and detection of parameters characterizing operator mistakes to reveal the causes of abnormality. The techniques in question overcome limitations of existing methods and provide advantages over manual data analysis since they greatly reduce the combinatorial enumeration of the options considered.
Estimating the infl uence of human factor on the activity of operators of complex technical systems is an important problem for condition monitoring, personnel training and diagnostics. Presented are both an overview and mutual comparisons of the approaches which are useful to reveal the effect of human factor and have already shown their performances in practical applications. Under consideration are: the structural equation modeling, the Bayesian estimations for probabilistic models represented by Markov random processes, the multivariate statistical techniques including the discriminant and cluster analysis as well as wavelet transforms.
The H.R. Wilson and J.D. Cowan model of human brain neural activity is considered, which connects the latent (non-measurable) parameters of external excitation with the brain neuronal activity parameters which are available for observation, with alpha rhythm frequencies being in use in this capacity. By means of computational experiments, revealed are the ranges of latent parameters, which ensure the occurrence of the Poincaré—Andronov—Hopf bifurcation (transforming the spiral sink to the spiral source equilibrium point with a limit cycle, or vice versa) that leads, in the first approximation, to the appearance of alpha rhythms. Relative changes of the latent parameters representing external excitation are estimated through the primary indices of oculomotor activity (gaze movement entropies and gaze fixation durations), which makes it possible to assess the pilot proximity to falling into a low efficiency state, including the real time mode.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.