Design And Airworthiness Of Human Factors Based On Harmonic Measure Of Flight Safety And Scientific Computing

It is flight safety that makes the major criterion which expects the advanced airborne system as well as the core target in human factors design and airworthiness.Man-induced contingency is the leading cause of aviation accidents,which is inevitable to avoid since the ergonomic state space that we know is much more inadequate.Various uncertainties with regard to aviation systems and flight missions make the ergonomic state space becoming into a vast and complex one.Therefore,a complete set of scientific solution is in demand to sufficiently accumulate research samples in the ergonomic state space on which scientific advance and computational application in human factors are to be based,for the sake of design and airworthiness,and for that of automation and safety.The problem of human factors design and airworthiness can be transmuted into the one of fact and logic in ergonomic state space via modeling and simulation of aviation systems oriented to scientific computing.Dynamic evolution of conditions,states,and performance relating to the crew,the mission,the environment,and the aircraft by itself are to be transcribed as a multidimensional curve and a research sample in ergonomic state space.Configurations about the aviation system and interactions in the flight process consist the first informative dimension,and quantized measures of flight safety,aircraft performance,human performance,and general evaluations consist the second informative dimension,of the research samples.Statistic modeling of such data samples,casual inference based on statistic data model,human factors inverse based on statistic data model,and orthogonal experiment design based on statistic data model are information algorithms at four levels in computational human factors and ergonomics.Prospective studies in aviation human factors design and airworthiness will provide scientific bases for coordination and decision-making in flight system control and aeronautical systems engineering.This dissertation pursued comprehensive and in-depth theoretical and computational studies over the above topics,and the specific work is divided into four parts:First of all,we methodologically discussed the solution for human factors design and airworthiness based on harmonic safety measure,and we systematically elaborated the profoundness of modeling and simulation of aviation systems oriented to scientific computing from the aspects of approach,theory,technology,and tool development.Upon the scientific problem of human factors design and airworthiness,we explored the heterogeneity of human-aircraft interactions,the applicability of scientific computing and virtual engineering,the human centered prototyping and manufacturing,the unified representation of multi-granularity synthetic flight sphere,the multi-level heterogeneous interaction,and the multi-objective assessment.Upon the breadth,depth,and span of computational human factors and ergonomics,we proposed the feasible solution that the problem of human factors design and airworthiness will be transmuted into the one of fact and logic.Then we discussed the sample generation approach which was elaborated from eight aspects of human strategy modeling,aircraft invariant tensor modeling,flight scenario,intrinsic determinants and harmonic measure of flight safety,dimensionless reduction of aircraft performance,objective assessment of flying quality and human performance,transform domain measure of aviation general evaluation,and distributed deployment of digital fast-time computational platform.In accomplishing our tasks,we would keep hoping for the best,being prepared for the worst,and ensuring for the average.Such capabilities and traits will be simulated from a unique perspective of multiple characteristic pattern and structural realization via human strategy modeling,according to complexity,adaptability,and uncertainty,which are to be acknowledged by computational sequences with appropriate logic and arithmetic depth based on optimal control,robust control,and adaptive decision.Aircraft is the object to be manipulated,whose invariant modeling and matrix coding are suitable for computer simulation with adequate fidelity and credibility.Flight scenario are parametric configurations about missions and conditions in aviation,where mission segments along with transition descriptions are targeted at human strategy modeling,condition schedules along with data windows and data items are targeted at airworthiness tests.Awareness of the approaching hazards and consistent assessment of the potential risks are indispensable to continually defend the safety of flight.Hazard identification and risk management involve various aviation events and assessment criteria,in which both the objective quantitative analysis and the results can hardly be applied.A unified mechanism where safety factors are summed up in three intrinsic determinants and integrated into a transforming domain is proposed so that safety information shall be measurable,evaluable,and predictable in the form of probabilistic quantification.Dimensionless reduction of aircraft performance can be exploited in deriving intrinsic determinants for complicated flight conditions.Performance parameters with proper standardizing and grouping enable the detection and accuracy of dimensionless performance.Subjective assessment of man-aircraft system in virtual environment is pursued as the objective one for flying quality and human performance.Since the assessment of flying quality and human performance is a process of synergistic verifications,multiple decisions,and feedback controls,the introspection capabilities in human performance modeling are achieved accordingly in the framework of cognition based on synergistic learning and predictive control.Subjective evaluations for lots of abstract properties are involved in human factors.Relative probabilistic measures and quantitative analysis for such properties are made possible,facilitated by the analogy to transport phenomena and transport properties in fluid dynamics.Distributed deployment of computational platform based on service oriented architecture fully supports the modeling and simulation of man-aircraft-environment intelligent systems,which provides a scientific computing and research platform for human factors design and airworthiness with flexible computational complexity,distributivity,parallelism,openness,and extendibility.And we discussed the prospective research of aviation human factors based on ergonomic state space from a standpoint of big data computing.A data driven and problem oriented big data approach was proposed from five aspects of multi-level longitudinal data layout,statistic data model,causal inference,human factors inverse problem,and orthogonal experiment design.Mixed-effect and causal-covariate model,Bootstrapping stochastic algorithm for parameter estimation,causal inference algorithm for existence of causal relationship in the framework of goodness of fit,causal inference algorithm for significance evaluation of causal effect in the framework of potential outcome,optimal estimation algorithm for human factors response kernel,and algorithm for asymptotic approximation of ergonomic subspace push ahead aviation human factors research.Conclusively,we demonstrated from the standpoints of system control and systems engineering that human factors design and airworthiness based on scientific computing have good prospects for application in automation extension and safety enhancement.Special attentions were payed on the feedback of harmonic measure of flight safety in airborne system control,and on the feed forward of human factors design and airworthiness in aeronautical systems engineering.