A time positioning method for bridge health monitoring data gradual change noise

Abstract

The application discloses a time positioning method for gradual change noise of bridge health monitoring data, and comprises the following steps: 1) obtaining bridge health monitoring time series data without gradual change noise, and calculating the difference of original data; 2) calculating the mean square deviation of the difference sequence with 30s as a time unit, and obtaining the mean square deviation sequence; 3) calculating the change rate of the mean square deviation sequence, and calculating the mean and variance of the change rate sequence, and defining the mean square deviation change rate index I; 4) determining the upper and lower limits of the control chart, and the sample exceeding the upper and lower limits of the control chart indicates that the gradual change noise appears, so that the positioning of the gradual change noise is realized. The original complete monitoring data set containing short-time strong noise is obtained through the bridge health monitoring system, the positioning of the gradual change noise is realized based on the characteristics that the fluctuation of the noise is larger than that of normal data, and the gradual change noise data can be effectively positioned for subsequent cleaning.

Description

Technical Field

[0001] This invention belongs to the field of bridge engineering technology, and specifically relates to a time-based method for locating gradual noise in bridge health monitoring data. Background Technology

[0002] The completeness and availability of bridge health monitoring data have a significant impact on bridge damage identification and early warning. Due to improper sensor installation, aging, severe weather, and other uncontrollable factors, gradual noise often appears in monitoring sensors. Therefore, accurately locating data exhibiting gradual noise is crucial. Summary of the Invention

[0003] In view of this, the present invention provides a time-based localization method for gradual noise in bridge health monitoring data. Based on the characteristic that the noise itself fluctuates more than normal data, the method can locate the gradual noise for subsequent cleaning.

[0004] The technical solution adopted by this invention to solve its technical problem is:

[0005] A method for temporal localization of gradually changing noise in bridge health monitoring data, comprising the following steps:

[0006] 1) Obtain bridge health monitoring time history data without gradual noise and calculate the difference of the raw data;

[0007] 2) Calculate the mean square error of the difference sequence using 30s as the time unit, and obtain the mean square error sequence;

[0008] 3) Calculate the rate of change of the mean squared error series, and calculate the mean and variance of the rate of change series, and define the mean squared error rate of change index I;

[0009] 4) Determine the upper and lower limits of the control chart. If the sample exceeds the upper and lower limits of the control chart, it indicates the presence of gradual noise, thus enabling the localization of gradual noise.

[0010] Furthermore, the process of step 1) is as follows:

[0011] Consider a segment of historical data Y from a sensor that does not contain gradual noise, and calculate its difference sequence:

[0012] Y d =Y(2:end)-Y(1:end-1) (1)

[0013] Among them, Y d It is the difference sequence of the original data, that is, the difference between the previous data point and the next data point in the original data, so that the baseline of the data is 0, so that the mean square error can be calculated later.

[0014] Furthermore, the process of step 2) is as follows:

[0015] Using the moving window method, the mean squared error of only 30 seconds of data is calculated each time, with a step size of 30 seconds, thereby obtaining the mean squared error sequence Y of the difference sequence. s ∈R n .

[0016] Furthermore, the process of step 3) is as follows:

[0017] Calculate the mean squared error sequence Y s Rate of change Y′:

[0018] Y′=diff(Y s (2)

[0019] Where diff represents the expression for Y s After obtaining Y′ by differentiation, calculate the mean μ0 and variance σ0 of Y′:

[0020] μ0=mean(Y) (3)

[0021]

[0022] Where y′ i Let be the i-th element in Y′. Define the rate of change of the root mean square (RMS) index I according to the 3σ principle:

[0023]

[0024] Furthermore, the process of step 4) is as follows:

[0025] Determine the upper and lower limits of the control chart. If a sample exceeds these limits, it indicates the presence of gradual noise, thus enabling the localization of gradual noise. Based on the form of the mean squared error rate of change index I defined in 3), the upper and lower limits of the control chart should be 1 and -1, respectively. Once a sample value exceeds these limits, it indicates that gradual noise has occurred at the time represented by that sample value.

[0026] t=i×30(s) (6)

[0027] Where t is the time when the gradual noise appears, calculated with the time represented by the first sample value as 0, and i is the sample number that exceeds the upper and lower limits of the control chart. This completes the localization of the gradual noise.

[0028] The technical concept of this invention is as follows: by acquiring the original complete monitoring dataset containing short-term strong noise through a bridge health monitoring system, and based on the characteristic that the noise itself fluctuates more than normal data, the gradual noise is located for subsequent cleaning.

[0029] The beneficial effects of this invention are: it can efficiently and accurately locate gradual noise based on the characteristic that the noise itself fluctuates more than normal data. Attached Figure Description

[0030] Figure 1 Flowchart for locating gradual noise.

[0031] Figure 2 This is a graph of deflectometer data.

[0032] Figure 3 This is a graph showing the results of the data processing. Detailed Implementation

[0033] To enable those skilled in the art to better understand the technical solution of the present invention, a time-based localization method for gradual noise in bridge health monitoring data provided by the present invention is described in detail below with reference to embodiments. The following embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0034] Example 1

[0035] refer to Figures 1-2 A method for temporal localization of gradual noise in bridge health monitoring data, comprising the following steps:

[0036] 1) Obtain bridge health monitoring time history data without gradual noise, and calculate the difference between the original data:

[0037] Consider a segment of historical data Y from a sensor that does not contain gradual noise, and calculate its difference sequence:

[0038] Y d =Y(2:end)-Y(1:end-1) (1)

[0039] Among them, Y d It is the difference sequence of the original data, that is, the difference between the previous data point and the next data point in the original data, so that the baseline of the data is 0, so that the mean square error can be calculated later.

[0040] 2) Calculate the mean square error of the difference sequence using 30-second time units to obtain the mean square error sequence:

[0041] Using the moving window method, the mean squared error of only 30 seconds of data is calculated each time, with a step size of 30 seconds, thereby obtaining the mean squared error sequence Y of the difference sequence. s ∈R n .

[0042] 3) Calculate the rate of change of the mean squared error series, and calculate the mean and variance of the rate of change series, and define the mean squared error rate of change index I:

[0043] Calculate the mean squared error sequence Y s Rate of change Y′:

[0044] Y′=diff(Y s (2)

[0045] Where diff represents the expression for Y s After obtaining Y′ with derivative, calculate the mean μ0 and variance σ0 of Y′:

[0046] μ0=mean(Y) (3)

[0047]

[0048] Where y′ i Let be the i-th element in Y′. Define the rate of change of the root mean square (RMS) index I according to the 3σ principle:

[0049]

[0050] 4) Determine the upper and lower limits of the control chart. If the sample exceeds the upper or lower limit of the control chart, it indicates the presence of gradual noise, thus enabling the localization of gradual noise:

[0051] Determine the upper and lower limits of the control chart. If a sample exceeds these limits, it indicates the presence of gradual noise, thus enabling the localization of gradual noise. Based on the form of the mean squared error rate of change index I defined in 3), the upper and lower limits of the control chart should be 1 and -1, respectively. Once a sample value exceeds these limits, it indicates that gradual noise has occurred at the time represented by that sample value.

[0052] t=i×30(s) (6)

[0053] Where t is the time when the gradual noise appears, calculated with the time represented by the first sample value as 0, and i is the sample number that exceeds the upper and lower limits of the control chart. This completes the localization of the gradual noise.

[0054] The implementation process of the case is as follows:

[0055] (1) Experimental preparation

[0056] Data set Y is formed by selecting a week's worth of data without gradual noise, and Y is calculated. d Y s Y′, μ0, and σ0. Then, data containing gradually varying noise are selected to construct control charts.

[0057] (2) Experimental Results

[0058] The moment when the first sample point exceeding the control limits is taken as the moment when the gradual noise appears is calculated. The moment when the noise occurs is at the 54,000th point, while the actual human eye observation is at the 53,756th point, with an error of 0.454%.

[0059] The present invention provides a temporal localization method for gradual noise in bridge health monitoring data. This method acquires a complete original monitoring dataset containing short-duration strong noise from a bridge health monitoring system. Based on the characteristic that the noise itself fluctuates significantly more than normal data, it locates the gradual noise for subsequent cleaning. This invention can very effectively locate gradually changing noise data.

[0060] The embodiments of the present invention have been described in detail above. However, the present invention is not limited to the above embodiments. Various changes that can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for temporal localization of gradually changing noise in bridge health monitoring data, characterized in that, Includes the following steps: 1) Obtain bridge health monitoring time history data without gradual noise and calculate the difference of the raw data; 2) Calculate the mean square error of the difference sequence in 30-second time units to obtain the mean square error sequence; 3) Calculate the rate of change of the mean squared error series, and calculate the mean and variance of the rate of change series, and define the mean squared error rate of change index I; 4) Determine the upper and lower limits of the control chart. If the sample exceeds the upper and lower limits of the control chart, it indicates the presence of gradual noise, thus enabling the localization of gradual noise. The process of step 1) is as follows: Consider a segment of historical data Y from a sensor that does not contain gradual noise, and calculate its difference sequence: (1) Among them, Y d It is the difference sequence of the original data, that is, the difference between the previous data point and the next data point in the original data, so that the baseline of the data is 0, so that the mean square error can be calculated later. The process of step 3) is as follows: Calculate the mean squared error sequence rate of change : (2) Where diff represents the... Find the derivative and obtain Then, calculate The mean μ0 and variance σ0: (3) (4) in for The i-th element is used to define the rate of change of the root mean square error (I) according to the 3σ principle. (5) ； The process of step 4) is as follows: Determine the upper and lower limits of the control chart. If a sample exceeds these limits, it indicates the presence of gradual noise, thus allowing for the localization of this noise. Based on the form of the mean squared error rate of change index I defined in section 3), the upper and lower limits of the control chart should be 1 and -1, respectively. Once a sample value exceeds these limits, the moment represented by that sample value indicates the presence of gradual noise. (6) Where t is the time when the gradual noise appears, calculated with the time represented by the first sample value as 0, and i is the sample number that exceeds the upper and lower limits of the control chart, thus realizing the localization of the gradual noise.

2. The time-based localization method for gradual noise in bridge health monitoring data according to claim 1, characterized in that, The process of step 2) is as follows: Using the moving window method, the mean squared error of only 30 seconds of data is calculated each time, with a step size of 30 seconds, thereby obtaining the mean squared error sequence of the difference sequence. .

Images