Title :
Improved Algorithm for Correlation Dimension in Vibration Signal Fault Diagnosis
Author :
Du, Biqiang ; Tang, Guiji ; Shi, Junjie
Author_Institution :
Mech. Eng. Dept., North China Electr. Power Univ., Baoding, China
Abstract :
Abstract-Based on the fact that the fractal scaleless band and correlation dimension will gradually reach saturation and consensus along with the increase of the embedding dimension, an improved correlation dimension algorithm is proposed. The proposed algorithm, which is composed of complex auto-correlation method and curve-line-curve fitting method, can automate the calculation of reconstructing phase space parameters and fractal scaleless band, indeed correlation dimension of vibration signals. Case analysis results showed that the proposed algorithm can be used to calculate the correlation dimension of vibration signals automatically and accurately, without any prior knowledge of vibration signals.
Keywords :
acoustic signal processing; curve fitting; fault diagnosis; fractals; rotors; vibrations; complex auto-correlation method; correlation dimension algorithm; curve-line-curve fitting method; fractal scaleless band; mechanical equipments; reconstructing phase space parameter; rotor; vibration signal fault diagnosis; Algorithm design and analysis; Correlation; Delay; Delay effects; Fractals; Time series analysis; Vibrations;
Conference_Titel :
Power and Energy Engineering Conference (APPEEC), 2011 Asia-Pacific
Conference_Location :
Wuhan
Print_ISBN :
978-1-4244-6253-7
DOI :
10.1109/APPEEC.2011.5748476