Design and development of the instant rotational speed analysis system

Abstract

This thesis investigates the development of a Python code-based system for correction using signals generated by zebra stripes to analyze instantaneous angular velocities (IAS). It addresses the data errors mainly due to butt joint faults. It classifies the data acquired from different types of motion of rotating machines and applies different correction algorithms correspondingly. For constant speed conditions, LOWESS smoothing combined with Z-scores and the 3 σ rule was used to identify outliers and linear interpolation was used to reconstruct the data. For sinusoidal speed conditions, the system uses Gaussian filtering and the 3 σ rule to detect outliers and Fourier series fitting to reconstruct the sample point data. Experimental verification under different conditions shows that this method effectively improves the accuracy and reliability of IAS measurements.