A method of aircraft automatic landing safety compliance verification at the stage of mathematical simulation A.G. Kuznetsov, L.N. Aleksandrovskaya, V. N. Masur JSC Moscow Institute of Elektromechanics and Automatics, 5, Aviatzionny Pereulok, Moscow, 125319, Russia Phone: (495) 223-2709 (add. 3405), fax: (495) 152-2631, e-mail: [email protected] Objective: The ultralow risk parameters estimation procedures building for the aircraft safety management. Research direction: The probabilities distribution empirical functions analysis of the automatic landing safety main performance, achieved at the stage of statistical simulation. Main results: 1. The selection of analytical dependence probabilities distribution empirical functions approximation methods using • Pareto distribution; • mixture distribution; • spline functions. 2. The analytical dependence extrapolation to the unobservable distributions “tails”. Application: The rough analysis procedure development for the statistical simulations volume contraction and the reasonable engineering decision making. Validation: Automatic Landing System for certified An-148 aircraft. Статист. Statistical simulation моделир. nn11 = = 300,000 300тыс The results of the touchdown distance extrapolation to the unobservable probability distributions “tails” Xmax < Xadm Xmax<Xдоп Прогноз на Prediction on Pзад=0,999999 Pdr = 0.999999 Xpr < Xadm Xпрог<Xдоп Статист. Statistical simulation моделир. nn22 = = 600,000 600тыс Xmax<Xдоп Xmax < Xadm Прогноз на Prediction on Pзад=0,999999 Pdr = 0.999999 Xpr < Xadm Xпрог<Xдоп Статист. Statistical simulation моделир. n3n=3 1,000,000 = 1млн The touchdown distance prediction results Xmax<Xдоп Xmax < Xadm Прогноз на Prediction on P = 0.999999 Р=0,9999996 dr (n = 2,300,000) (при n=2,3млн) Коррекция Control law закона adjustment управления Xpr < Xadm Xпрог<Xдоп Resumption of Продолжение static tests стат.моделир. (to n 2,300,000) до n= = 2,3млн Система System meets удовлетворяет the requirements требованиям Ограничение на FCS limit, alert СУ, сигнализация Number of realizations 300,000 600,000 1,000,000 Approximation method Pareto Mixture Splines Pareto Mixture Splines Pareto Mixture Splines Prediction result (m) 839.673 857.746 826.03 830.435 848.35 825.08 815.865 829.842 818.67 The experimental value per 1 million of realizations 814,9 m Bibliography: L.N. Aleksandrovskaya, V.G. Borisov, V.N. Mazur, S.V. Hlgatyan, A.E. Ardalionova Non-Parametric Methods for Assessing Compliance with the Safety Requirements of Automatic Landing Aircraft Airworthiness // Mechatronics, automation, control №3(156), March 2014.
© Copyright 2024 ExpyDoc
