Linear Prediction Coding Used for Verification of the Quality of Steel Anchor Adhesion in Rock Mass
Krzysztof Gadek
Abstract:
The paper presents a method of determining the quality of adhesion based on the analysis of the supersonic wave driven by the Dirac potential impulse. Analysis of the response signal by means of linear prediction coding assumes similarity between generating speech in the vocal tract of the larynx and generating the response of the steel anchor to Dirac impulse function. The model of the vocal tract used in the analysis of voiced sounds by LPC was used for the analysis of the response signal of the steel anchor driven by a train of impulses. The purpose of this research is the analysis of reflection coefficient and residue and their relations with physical parameters of anchor adhesion. Detection of clear interrelations between these two values will confirm the thesis that this method of signal analysis by means of LPC is useful in verification of the quality of steel anchor adhesion in rock mass.
