Yu. G. Byluchev, I. A. Babushkin, L. I. Borodin, V. A. Golovskoy, A. A. Mozol’
The problem of data selection (identification) in the observation for some targets (T) is a charasteristics of the multipositional angular-measuring system (MAMS) along with the traditional problems of the T finding and tracking. Under selection it is meant the partition of the finding values measured in different points (P) into the classes according to the character of the belonging to the given T.
In one of his early works Byluchev Yu. G. has developed a solution method of the given problem with the use of one invariant (the slope angle of some plane related to the T and two MAMS finders) providing the processing decentering of the measurement data).
In the given article we give a further development of the given idea based on building all the independent invarians that permit to realize the decentering idea providing the ccomputational correctness of the formed selection procedures in this case.