PSI - Issue 17

Pavel Steinbauer et al. / Procedia Structural Integrity 17 (2019) 799–805 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

804

6

frequency spectra as follows

)

if

PSD f

min PSD pro f

0.1,  − + f f

  

1 0

( )

0.1



1



b SD f T i

T

i

i

P

( )

=

N

jinak

PSD

T

)

if

psd f

min PSD pro f

0.1,  − + f f

  

1 0

( )

0.1



1



b sd f a i

(7)

a

i

i

p

( )

=

N

jinak

psd

a

Then entropy analogy for excitated spectra, excluding absolute signal power amplitudes follows ( ) ( ) 2 ( ) ( ) log ( ) b b b T i T i T i i H PSD f PSD f PSD f  = −   , ( ) ( ) 2 ( ) , ( ) ( ) log ( ) b b b b T i a i T i a i i H PSD f psd f PSD f psd f  = −   . Difference of reference and actual excited spectra is then obtained as ( ) ( ) ( )   ( ) , ( ) ( ) ( ) , ( ) b b b b b T i a i T i T i a i PSD f psd f H PSD f H PSD f psd f bit  = − (8)

6. Results

The method was tested on acceleration data measured in the field pole laboratory. Each pole was equipped by MEMS accelerometer ADXL330 glued on the top of the pole (Fig. 5 a ). The experimental modal analysis was performed to obtain reference modal data of each pole.

Fig. 5 Pole laboratory a , prototype software in Simulink b For acceleration measurement, the poles were excited by ambient wind only. Data were processed by prototype software developed in Matlab/Simulink environment, which generated C-code for target platform directly (Fig. 5 b ). Eigen frequencies detected by the algorithm are shown on the Fig. 6 b .