L.I. Dvoyris, D.V. Lutsenko, R.A. Mikhailov, V.S. Zavyalov

The authors have conducted experimental studies to test detection equipment, and the resulting statistical material on the basis of the mathematical apparatus of principal component analysis using the method based on support vector machines (support vector machine SVM), the so-called Lagrangian (iterative) SVM, showed signs of an impact on formation means of detection of false alarm messages in extreme operating conditions. Based on these results, it was constructed decision rule which allowed the classification of the false as false an alarm caused by the influence of meteorological factors. In an assessment of distributions of principal components normal operation mode of detection and of false alarm signal when exposed to weather disturbances.