USE OF THE METHOD OF DECOMPOSITION OF SINGULAR VALUES FOR ESTIMATION OF NOISE LEVEL AND DETECTION OF SENSOR CALIBRATION VIOLATIONS IN INFORMATION AND MEASUREMENT SYSTEMS
The Singular Value Decomposition (SVD) is a powerful tool for data analysis in information and measurement systems (IMS). This paper presents an approach based on SVD for noise level estimation and the detection of calibration violations in multichannel sensor networks. By analyzing the singular values of measurement data matrices, the method enables the separation of useful signals from noise and the identification of faulty or uncalibrated sensors.