V. A. Ufaev – Dr.Sc. (Eng.), Senior Research Scientist, Air Force Academy n.a. Prof. N.E. Zhukovsky and Yu.A. Gagarin (Voronezh). E-mail: Andreyuff@mail.ru
Objects coordinate identification and estimation according to bearing mixed sample algorithms are synthesized by statistical hypothesis control method.
Statistically optimal identification-estimation algorithm consists in hypothesis control. The number of these hypothesis increases according to factorial-degree rule in dependence on magnifying objects and direction finding cycles number. Degree dependence is eliminated by local optimality without correction principle usage. However, early accepted solutions extension is necessary. Factorial dependence is eliminated by extra criterion of distance minimum between standards and measurements. Distances analytical estimations are achieved by the sum of squared distances from source to bearings line.
Complex and autonomous identification efficiency is estimated by simulation. Necessary direction finding proximities are calculated. It is established, that transition to the specified criterion, and also from complex to independent identification on direction finding posts are accompanied by practically comprehensible losses of identification-evaluation efficiency.
True identification probabilities are relatively stable, but slow error decrease with observation time decrease is typical for coordinate estimation. True identification probability decreases and coordinates estimation error increases with the objects number increase. The saturation effect is occurred in estimation characteristics. It causes coordinate estimation errors limitation by the threshold level.
The object zone extension causes the true identification probabilities and saturation threshold levels increase.
It is need to provide direction finding errors not more than 0,3…0.7° for providing practically significant level of true identification probability, which is not less than 0,7 at 2..20 simultaneously functioning objects per km2. Average coordinates definition squared error is not more than 0,25..0,04 km at these conditions.
The article contains 8 pictures, 2 tables and 5 references.