Neural network technologies in metrology: automation of calculations of standard deviation and uncertainty of measurements

The article is devoted to the application of neural network technologies for automating calculations of the standard deviation and uncertainty of measurements in metrology. The evolution of data processing methods is considered, the limitations of traditional approaches are identified, and the need for innovative solutions is justified. The development of a TensorFlow-based neural network system is described, which provides 99.3% accuracy and reduces processing time from 2 hours to 24 seconds.