CN104732097A - Correcting method for power spectrum in modal frequency identification of railroad bridge under strong signal interference - Google Patents

Abstract

The invention discloses a correcting method for a power spectrum in modal frequency identification of a railroad bridge under strong signal interference. An acceleration sensor is arranged on the railroad bridge, power spectrum analyzing is carried out on a time domain vibration signal acquired by the acceleration sensor, and a power spectrum curve is obtained; a finite element analyzing model of the railroad bridge is established, dynamic property analyzing is carried out, a theoretical modal frequency of the railroad bridge is obtained, and according to the theoretical modal frequency, a to-be-corrected frequency interval of the actual measured power spectrum curve is confirmed; by adopting the D-S theory, correcting is carried out on data, of the actual measured power spectrum curve, in the to-be-corrected frequency interval, and according to a peak value position of the corrected power spectrum curve, an actual measured modal frequency of the railroad bridge is identified. After correcting is carried out on actual measured power spectrum data of the railroad bridge, the modal frequency of the railroad bridge can be accurately identified, the unfavorable influence, caused by the strong signal interference, on the actual measure power spectrum data is effectively overcome, and it is certain that the correcting method receives a wide application and popularization.

Description

The modification method of power spectrum in railroad bridge identification of mode frequency under strong signal disturbing

Technical field

The present invention relates to the field of non destructive testing of railroad bridge engineering, particularly a kind of modification method being applied to power spectrum data in railroad bridge identification of mode frequency under strong signal disturbing.

Background technology

Larger vibration may be produced when bullet train runs on railroad bridge, must be paid attention to the impact of bullet train security of operation and bridge structure safe.Therefore, carry out vibration monitoring for railroad bridge, for guarantee railroad bridge operation security, there is important using value.The important content of railroad bridge vibration monitoring identifies bridge model frequency, by the vibrational state of the change reflection railroad bridge of model frequency.The common method of bridge identification of mode frequency is the peak picking method based on power spectrumanalysis, namely in bridge structure, acceleration transducer is installed, the time-domain signal of structural vibration is gathered by acceleration transducer, then power spectrumanalysis is carried out to vibration time-domain signal and obtain power spectrum curve, occur that the principle of peak value directly can identify bridge model frequency from power spectrum chart according to power spectrum curve at structural modal frequency place.Peak picking method is widely applied in the identification of mode frequency of highway bridge structure.

But, during railroad bridge employing peak picking method identification model frequency, there is the problem of strong signal disturbing.This has installed multiple instrument and communication apparatus in the line project due to railroad bridge, in order to transmit the information such as instruction and order about rolling stock service condition, travelling facility state and driving.These instrument and communication apparatus can produce significant high reject signal in railway bridge vibration of beam time-domain signal, cause power spectrum curve not only having bridge survey peak value corresponding to model frequency, also have " puppet " peak value caused by all kinds of undesired signal, " puppet " peak value that the latter causes has randomness simultaneously.Therefore, high reject signal can cause power spectrum chart comparatively disorderly, cannot directly according to the model frequency of spectrum peak determination railroad bridge.

Summary of the invention

Goal of the invention: in order to overcome the deficiencies in the prior art, the invention provides the modification method of power spectrum in railroad bridge identification of mode frequency under a kind of strong signal disturbing, there is more strong jamming during for solving the model frequency of Peak Intensity Method identification railroad bridge causes power spectrum chart comparatively disorderly, cannot directly according to the technical matters of the model frequency of spectrum peak determination railroad bridge.

Technical scheme: for achieving the above object, the technical solution used in the present invention is:

The modification method of power spectrum in railroad bridge identification of mode frequency under strong signal disturbing, comprises the following steps that order performs:

Step one, degree of will speed up sensor is arranged on railroad bridge, in order to measure the time domain vibration signal of railroad bridge;

Step 2, to acceleration transducer obtain time domain vibration signal carry out power spectrumanalysis, obtain measured power spectral curve;

Step 3, set up railroad bridge finite element analysis model, action edge specificity analysis of going forward side by side, obtain the theoretical model frequency f of railroad bridge _d; And according to theoretical model frequency f _dthat determines measured power spectral curve treats that frequency of amendment interval is for [f ₁, f ₂], wherein, the lower limit f of frequency separation ₁=f _d× 0.9, the higher limit f of frequency separation ₂=f _d× 1.1;

Step 4, employing D-S evidence theory are treating frequency of amendment interval [f to the measured power spectral curve that step 2 obtains ₁, f ₂] in data revise;

Step 5, identify the actual measurement model frequency f of railroad bridge according to the peak of power spectrum curve after revising _e.

Further, in the present invention, in step 2, within every 10 minutes, one group of measured power spectral curve is obtained; In step 4, the method for correction is as follows: at frequency separation [f ₁, f ₂] in frequency values f _i, m ₁(f _i), m ₂(f _i), m ₃(f _i), m ₄(f _i), m ₅(f _i), m ₆(f _i) be respectively respective frequencies value f in 6 groups of measured power spectral curves of 1 little interior acquisition _ipower spectral value; Frequency values f in this hour is calculated according to following (1) to (5) formula _icorresponding revised power spectral value M (f _i):

$T_1\left(f_i\right)=\frac{m_1\left(f_i\right)\×m_2\left(f_i\right)}{1-m_1\left(f_i\right)\×m_2\left(f_i\right)}---\left(1\right)$

$T_2\left(f_i\right)=\frac{T_1\left(f_i\right)\×m_3\left(f_i\right)}{1-T_1\left(f_i\right)\×m_3\left(f_i\right)}---\left(2\right)$

$T_3\left(f_i\right)=\frac{T_3\left(f_i\right)\×m_4\left(f_i\right)}{1-T_2\left(f_i\right)\×m_4\left(f_i\right)}---\left(3\right)$

$T_4\left(f_i\right)=\frac{T_3\left(f_i\right)\×m_5\left(f_i\right)}{1-T_3\left(f_i\right)\×m_5\left(f_i\right)}---\left(4\right)$

$M\left(f_i\right)=\frac{T_4\left(f_i\right)\×m_6\left(f_i\right)}{1-T_4\left(f_i\right)\×m_6\left(f_i\right)}---\left(5\right)$

To frequency separation [f ₁, f ₂] in measured power spectrum corresponding to all frequency values revise all as stated above, finally obtain revised power spectrum curve.

Beneficial effect: railroad bridge, owing to there is significant high reject signal in line project, causes power spectrum chart comparatively disorderly, cannot according to the model frequency of spectrum peak determination railroad bridge.The present invention adopts D-S evidence theory to revise measured power modal data, the power spectral value that can effectively suppress high reject signal to cause, and significantly amplifies power spectral value corresponding to actual measurement model frequency, thus accurately can identify the model frequency of railroad bridge.The method implements simple and convenient, can obtain extensive propagation and employment.

Accompanying drawing explanation

Fig. 1 is that in the embodiment of the present invention, the front view closing bridge is won in Beijing-Shanghai express railway Nanjing completely;

Fig. 2 be in the embodiment of the present invention 6 groups of measured power spectral curves treating the data in frequency of amendment interval;

Fig. 3 is treating revised power spectrum curve in frequency of amendment interval in the embodiment of the present invention.

Embodiment

Below in conjunction with accompanying drawing, the present invention is further described.

The modification method of power spectrum in railroad bridge identification of mode frequency under a kind of strong signal disturbing of the present invention, the method comprises the steps:

(1) degree of will speed up sensor is arranged on railroad bridge, in order to measure the time domain vibration signal of railroad bridge;

(2) with 10 minutes for computation interval, to acceleration transducer obtain time domain vibration signal carry out power spectrumanalysis, 6 groups of measured power spectral curves can be obtained in 1 hour;

(3) set up railroad bridge finite element analysis model, action edge specificity analysis of going forward side by side, obtain the theoretical model frequency f of railroad bridge _d.According to theoretical model frequency f _dthat determines measured power spectral curve treats that frequency of amendment interval is for [f ₁, f ₂], wherein, the lower limit f of frequency separation ₁for f ₁=f _d× 0.9, the higher limit f of frequency separation ₂for f ₂=f _d× 1.1.

(4) with 1 hour for computation interval, adopt D-S evidence theory frequency of amendment interval [f is being treated to 6 groups of measured power spectral curves that step (2) obtains ₁, f ₂] in data revise:

(4a) f is established _ifor frequency separation [f ₁, f ₂] in some frequency values, m ₁(f _i), m ₂(f _i), m ₃(f _i), m ₄(f _i), m ₅(f _i), m ₆(f _i) be respectively respective frequencies f in 6 groups of measured power spectral curves _ipower spectral value, then revised power spectral value M (f _i) by following formulae discovery (T ₁(f _i), T ₂(f _i), T ₃(f _i) and T ₄(f _i) be results of intermediate calculations):

$T_1\left(f_i\right)=\frac{m_1\left(f_i\right)\×m_2\left(f_i\right)}{1-m_1\left(f_i\right)\×m_2\left(f_i\right)},T_2\left(f_i\right)=\frac{T_1\left(f_i\right)\×m_3\left(f_i\right)}{1-T_1\left(f_i\right)\×m_3\left(f_i\right)},T_3\left(f_i\right)=\frac{T_2\left(f_i\right)\×m_4\left(f_i\right)}{1-T_2\left(f_i\right)\×m_4\left(f_i\right)}$

$T_4\left(f_i\right)=\frac{T_3\left(f_i\right)\×m_5\left(f_i\right)}{1-T_3\left(f_i\right)\×m_5\left(f_i\right)},M\left(f_i\right)=\frac{T_4\left(f_i\right)\×m_6\left(f_i\right)}{1-T_4\left(f_i\right)\×m_6\left(f_i\right)}$

(4b) to frequency separation [f ₁, f ₂] in measured power spectrum corresponding to all frequency values all revise by step (4a), finally obtain revised power spectrum curve.

(5) the actual measurement model frequency f of railroad bridge is identified according to the peak revising rear power spectrum curve _e.

Embodiment:

Win completely for Beijing-Shanghai express railway Nanjing below and close bridge identification of mode frequency, specific embodiment of the invention process is described:

(1) Nanjing wins the one-piece construction of pass bridge completely as shown in Figure 1, and in bridge span, a vertical acceleration transducer is installed in position, in order to measure the time domain vibration signal of railroad bridge.

(2) with 10 minutes for computation interval, to acceleration transducer obtain time domain vibration signal carry out power spectrumanalysis, 6 groups of measured power spectral curves can be obtained in 1 hour.

(3) set up Nanjing and win the finite element analysis model closing bridge completely, action edge specificity analysis of going forward side by side (giving the analysis result of single order vertical motion frequency in the present embodiment).Calculating shows, wins the theoretical model frequency f closing bridge single order vertical motion completely _dfor 0.3280Hz, therefore, measured power spectral curve treat that frequency of amendment interval is for [0.2952Hz, 0.3608Hz].

(4) 6 groups of measured power spectral curves that step (2) obtains are treating the data in frequency of amendment interval [0.2952Hz, 0.3608Hz] as shown in Figure 2.As can be seen from the figure, the power spectrum repeatedly identified under strong signal disturbing is comparatively disorderly, effectively cannot determine bridge single order vertical motion frequency.

(5) 6 groups of corresponding to all frequency values in frequency separation [0.2952Hz, 0.3608Hz] measured power modal data adopt D-S evidence theory to revise, and revised power spectrum curve as shown in Figure 3.The practical frequency f of single order vertical motion accurately can be identified from the peak figure _efor 0.3174Hz.Therefore, D-S evidence theory is adopted to revise measured power modal data, the power spectral value that can effectively suppress high reject signal to cause, and significantly amplify the power spectral value that actual measurement model frequency is corresponding, this shows that the method is applicable to the structural modal frequency identification of the strong signal disturbing of this kind of existence of railroad bridge.

The above is only the preferred embodiment of the present invention; be noted that for those skilled in the art; under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (2)

1. the modification method of power spectrum in railroad bridge identification of mode frequency under strong signal disturbing, is characterized in that: comprise the following steps that order performs:

Step one, degree of will speed up sensor is arranged on railroad bridge, in order to measure the time domain vibration signal of railroad bridge;

Step 2, to acceleration transducer obtain time domain vibration signal carry out power spectrumanalysis, obtain measured power spectral curve;

Step 3, set up railroad bridge finite element analysis model, action edge specificity analysis of going forward side by side, obtain the theoretical model frequency f of railroad bridge _d; And according to theoretical model frequency f _dthat determines measured power spectral curve treats that frequency of amendment interval is for [f ₁, f ₂], wherein, the lower limit f of frequency separation ₁=f _d× 0.9, the higher limit f of frequency separation ₂=f _d× 1.1;

Step 4, employing D-S evidence theory are treating frequency of amendment interval [f to the measured power spectral curve that step 2 obtains ₁, f ₂] in data revise;

Step 5, identify the actual measurement model frequency f of railroad bridge according to the peak of power spectrum curve after revising _e.

2. the modification method of power spectrum in railroad bridge identification of mode frequency under strong signal disturbing according to claim 1, is characterized in that: in step 2, within every 10 minutes, obtains one group of measured power spectral curve; In step 4, the method for correction is as follows: at frequency separation [f ₁, f ₂] in frequency values f _i, m ₁(f _i), m ₂(f _i), m ₃(f _i), m ₄(f _i), m ₅(f _i), m ₆(f _i) be respectively respective frequencies value f in the 6 groups of measured power spectral curves obtained in 1 hour _ipower spectral value; Frequency values f in this hour is calculated according to following (1) to (5) formula _icorresponding revised power spectral value M (f _i):

$T_1\left(f_i\right)=\frac{m_1\left(f_i\right)\×m_2\left(f_i\right)}{1-m_1\left(f_i\right)\×m_2\left(f_i\right)}---\left(1\right)$

$T_2\left(f_i\right)=\frac{T_1\left(f_i\right)\×m_3\left(f_i\right)}{1-T_1\left(f_i\right)\×m_3\left(f_i\right)}---\left(2\right)$

$T_3\left(f_i\right)=\frac{T_2\left(f_i\right)\×m_4\left(f_i\right)}{1-T_2\left(f_i\right)\×m_4\left(f_i\right)}---\left(3\right)$

$T_4\left(f_i\right)=\frac{T_3\left(f_i\right)\×m_5\left(f_i\right)}{1-T_3\left(f_i\right)\×m_5\left(f_i\right)}---\left(4\right)$

$M\left(f_i\right)=\frac{T_4\left(f_i\right)\×m_6\left(f_i\right)}{1-T_4\left(f_i\right)\×m_6\left(f_i\right)}---\left(5\right)$

To frequency separation [f ₁, f ₂] in measured power spectrum corresponding to all frequency values revise all as stated above, finally obtain revised power spectrum curve.