A performance testing device, testing method and system for a steel frame

Abstract

The present invention relates to the technical field of vibration testing of structural components, and specifically to a performance testing device, testing method and system for a steel frame. The present invention first obtains all vibration signal components of the vibration signal of each measuring point of the bridge steel frame under each external excitation, and all vibration peaks therein; then clusters the vibration peaks in each vibration signal component, obtains the periodic index of the vibration peak in each cluster, and divides each vibration signal component to obtain all component sub-segments, and then obtains the representative coefficient of each vibration signal component, and then screens out the vibration characteristic components of the vibration signal of each external excitation at each measuring point, so as to facilitate the final evaluation of the performance of the bridge steel frame. Based on the characteristic that the vibration signal under external excitation has periodic vibration characteristics, the present invention analyzes and screens out the corresponding vibration characteristic components, reduces the interference of other random vibration factors such as the environment or traffic, and accurately evaluates the performance of the bridge steel frame.

Description

Technical Field

[0001] The present invention relates to the technical field of vibration testing of structural components, and in particular to a performance testing device, a testing method and a system for a steel frame. Background Art

[0002] In bridge design, the steel frame is a structural system made of steel. It bears loads through steel beams, steel columns, trusses and other components to enhance the stability and safety of the bridge. However, due to the influence of various factors such as long-term outdoor use and traffic loads, the steel frame of the bridge may suffer from defects such as fatigue damage, deformation corrosion or cracks, which in turn affect the stability of the bridge; therefore, it is very important to perform performance testing on the steel frame of the bridge.

[0003] At present, external excitation is usually applied to the steel structure of the bridge for vibration testing. Then, by comparing and analyzing the difference in vibration characteristics between the vibration signal of the bridge steel structure and the vibration signal of the external excitation, it is evaluated whether the bridge steel structure has defects such as fatigue damage, so as to determine the performance of the bridge steel structure. However, when performing vibration testing on the bridge, the bridge will be affected by the vibration caused by the traffic load of vehicles passing by the bridge deck and other random environmental factors, which will affect the vibration test results and further affect the accuracy of the performance test of the bridge steel structure. Summary of the invention

[0004] In order to solve the technical problem of poor accuracy of performance testing of bridge steel frames in the prior art, the purpose of the present invention is to provide a performance testing device, testing method and system for steel frames. The technical solutions adopted are as follows:

[0005] The present invention provides a performance testing method for a steel frame, the method comprising:

[0006] Obtain all vibration signal components of the vibration signal of the bridge steel structure at each measuring point under each external excitation, and obtain all vibration peaks in each of the vibration signal components;

[0007] In each of the vibration signal components, all the vibration peaks are clustered to obtain all clusters; in each cluster, according to the occurrence position of each vibration peak and the waveform difference between different vibration peaks, the periodicity index of the vibration peak in each cluster is obtained;

[0008] Based on the clustering results of the vibration peaks in the vibration signal components, each of the vibration signal components is segmented to obtain all component subsegments; in each of the vibration signal components, according to the centralized features of all the component subsegments and the fluctuation differences between different component subsegments, combined with the number differences of vibration peaks in different clusters and the periodicity indexes of all clusters, the representative coefficient of each vibration signal component is obtained;

[0009] At each measuring point, based on the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, a vibration characteristic component is screened out from all the vibration signal components of the vibration signal of each external excitation; the performance of the bridge steel frame is evaluated by comprehensively analyzing the similarity between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components.

[0010] Furthermore, the cluster acquisition method includes:

[0011] In each of the vibration signal components, a metric distance is obtained according to the waveform difference between different vibration peaks, and based on the metric distance between different vibration peaks and a preset K value, K-means clustering is performed on all the vibration peaks to obtain all clusters.

[0012] Furthermore, the method for obtaining the periodicity index includes:

[0013] In each cluster, the concentrated features of the DTW distances between all different vibration peaks and the discrete features of the time intervals between all two vibration peaks that are adjacent in time sequence are integrated to obtain the first periodic parameter of the vibration peaks in each cluster; the concentrated features of the DTW distances and the discrete features of the time intervals are both negatively correlated with the first periodic parameter;

[0014] In each cluster, the time interval between the first vibration peak and the last vibration peak in the time sequence of the corresponding vibration signal component is used as the period duration of the vibration peak in the corresponding cluster; the ratio of the period duration of each cluster to the total duration of the corresponding vibration signal component is used as the second periodic parameter of the vibration peak in each cluster;

[0015] The first periodicity parameter and the second periodicity parameter are fused to obtain a periodicity index of a vibration peak in each cluster.

[0016] Furthermore, the method for obtaining the component sub-segments includes:

[0017] In each of the vibration signal components, assigning a label to each of the vibration peaks, wherein the labels of vibration peaks in the same cluster are the same, and the labels of vibration peaks in different clusters are different;

[0018] Based on the labels of all the vibration peaks in the vibration signal component, a label sequence corresponding to the vibration signal component is obtained; in the vibration signal component, the vibration peak corresponding to the label with the highest frequency of occurrence in the label sequence is used as the tail vibration peak of each component sub-segment, and the adjacent vibration peak after the tail vibration peak of each component sub-segment is used as the head vibration peak of the next adjacent component sub-segment; the component sub-segment is determined using the head vibration peak and the tail vibration peak.

[0019] Furthermore, the method for obtaining the representative coefficient includes:

[0020] Obtaining a label subsequence corresponding to each component subsegment in each vibration signal component, and taking the label subsequence with the highest occurrence frequency as a representative subsequence; taking the ratio of the total number of the representative subsequences to the total number of all label subsequences as a first representative parameter;

[0021] In the label sequence corresponding to each vibration signal component, the similarities between all different label subsequences are integrated to obtain a second representative parameter; the similarity is positively correlated with the second representative parameter;

[0022] In each of the vibration signal components, multiplying the negative correlation mapping result of the variance of the total number of the vibration peaks in all clusters by the sum of the periodicity indexes of all clusters to obtain the regularity index of each of the vibration signal components;

[0023] The first representative parameter, the second representative parameter and the regularity index are integrated to obtain a representative coefficient of each vibration signal component.

[0024] Furthermore, the method for obtaining the vibration characteristic component includes:

[0025] At any measuring point, any externally excited vibration signal is taken as the target vibration signal, any vibration signal component in the target vibration signal is taken as the target vibration component, and all the vibration signal components of all non-target vibration signals are taken as reference vibration components of the target vibration component;

[0026] Between the target vibration component and each reference vibration component, the product of the mean of the representative coefficients of the corresponding components and the Pearson correlation coefficient between the corresponding components is used as the correlation matching coefficient between the corresponding components;

[0027] According to all the correlation matching coefficients between each vibration component signal and all corresponding reference vibration components, the vibration characteristic component of the target vibration signal is screened out from all the vibration signal components of the target vibration signal.

[0028] Further, the method of screening out the vibration characteristic component of the target vibration signal from all vibration signal components of the target vibration signal includes:

[0029] The maximum correlation matching coefficient is screened out from all the correlation matching coefficients, and the vibration signal component corresponding to the maximum correlation matching coefficient is used as the vibration characteristic component of the target vibration signal.

[0030] Furthermore, the method for evaluating the performance of a bridge steel frame includes:

[0031] At each measuring point, the DTW distance between the vibration signal of each external excitation and the corresponding vibration characteristic component is accumulated to obtain a vibration difference parameter;

[0032] At each measuring point, the DTW distance between each externally excited vibration signal and the corresponding vibration characteristic component is used as a data point to fit a data curve; the absolute value of the slope of the data curve is used as a vibration change parameter;

[0033] According to the vibration change parameters and the difference parameters at all measuring points, the damage degree of the bridge steel frame is obtained and the performance of the bridge steel frame is evaluated; the vibration change parameters and the difference parameters are both positively correlated with the damage degree.

[0034] The present invention also proposes a performance testing device for a steel structure, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the performance testing method for a steel structure are implemented.

[0035] The present invention also provides a performance testing system for a steel frame, the system comprising:

[0036] Steel frame vibration data acquisition module: obtains all vibration signal components of the vibration signal at each measuring point of the bridge steel frame under each external excitation, and obtains all vibration peaks in each of the vibration signal components;

[0037] Steel frame vibration data analysis module: in each vibration signal component, all the vibration peaks are clustered to obtain all clusters; in each cluster, according to the waveform characteristics of each vibration peak and the waveform differences between different vibration peaks, the periodicity index of the vibration peak in each cluster is obtained; based on the clustering results of the vibration peaks in the vibration signal component, each vibration signal component is divided to obtain all component sub-segments; in each vibration signal component, according to the centralized characteristics of all the component sub-segments and the fluctuation differences between different component sub-segments, combined with the number differences of vibration peaks in different clusters and the periodicity index of all clusters, the representative coefficient of each vibration signal component is obtained; at each measuring point, according to the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, the vibration characteristic component is screened out from all the vibration signal components of each externally excited vibration signal;

[0038] Steel frame performance evaluation module: The performance of the bridge steel frame is evaluated by comprehensively analyzing the similarities between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components.

[0039] The present invention has the following beneficial effects:

[0040] The present invention first obtains all vibration signal components of the vibration signal of the bridge steel frame at each measuring point under each external excitation, and obtains all vibration peaks in each vibration signal component, so as to prepare for the subsequent analysis of the periodic vibration characteristics of the vibration signal components; then, in each vibration signal component, all vibration peaks are clustered to obtain all clusters; in each cluster, according to the occurrence position of each vibration peak and the waveform difference between different vibration peaks, the periodicity index of the vibration peaks in each cluster is obtained; based on the clustering results of the vibration peaks in the vibration signal component, each vibration signal component is segmented to obtain all component sub-segments, so as to facilitate the subsequent evaluation of the corresponding vibration signal component based on the component sub-segments to determine whether it is regular and periodic, so as to evaluate whether it is a vibration characteristic component; in each vibration signal component, according to the concentrated characteristics of all component sub-segments, The representative coefficient of each vibration signal component is obtained by combining the difference in the number of vibration peaks in different clusters and the periodic index of all clusters. The representative coefficient quantifies the possibility that each vibration signal component is a vibration characteristic component, and provides a basis for the subsequent screening of vibration characteristic components; at each measuring point, according to the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, the vibration characteristic component with the least external interference and the closest to the external excitation vibration characteristics is screened out from all vibration signal components of each externally excited vibration signal; and then, under the premise of reducing other external vibration interferences, the similarity between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components is comprehensively considered to accurately evaluate the performance of the bridge steel frame. Based on the characteristic that the vibration signal under external excitation has periodic vibration characteristics, the present invention analyzes and screens out the corresponding vibration characteristic components with the least external interference and the closest to the external excitation vibration characteristics, reduces the interference of other random vibration factors such as environment or traffic, and accurately evaluates the performance of the bridge steel frame. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 A flow chart of a method for testing the performance of a steel frame provided by one embodiment of the present invention;

[0043] Figure 2 A flow chart of a method for obtaining a representative coefficient provided by an embodiment of the present invention;

[0044] Figure 3 A system block diagram of a steel frame performance testing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a performance testing device, testing method and system of a steel frame proposed by the present invention, its specific implementation method, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0046] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0047] The following is a detailed description of a performance testing device, a testing method and a specific scheme of a steel frame provided by the present invention in conjunction with the accompanying drawings.

[0048] See also Figure 1 , which shows a flow chart of a performance testing method of a steel frame provided by an embodiment of the present invention, specifically comprising:

[0049] Step S1, obtaining all vibration signal components of the vibration signal of the bridge steel frame at each measuring point under each external excitation, and obtaining all vibration peaks in each vibration signal component.

[0050] It should be noted that the embodiment of the present invention is directed to a bridge steel frame, and implementers may also use it to evaluate the performance of other steel frames, which is consistent with the method for evaluating the performance of a bridge steel frame and will not be described in detail here.

[0051] In order to evaluate the performance of the steel frame of a bridge, one embodiment of the present invention first installs a vibration exciter at the center of the bridge deck to stimulate the vibration response of the steel frame of the bridge, and installs vibration sensors at various measuring points of the steel frame of the bridge to collect vibration signals, so as to subsequently analyze the vibration signals and determine whether there is any abnormality or damage in the bridge, thereby evaluating the performance of the steel frame of the bridge.

[0052] As an example, the measuring points are located at the joints of the steel frame of the bridge. The implementer can also arrange the measuring points and install vibration sensors according to the structure of the steel frame. The sampling frequency of the vibration sensor is set to 500Hz. The implementer can also set it by himself, but the sampling frequency of the vibration sensors at all measuring points must be consistent;

[0053] During the performance test of the bridge steel structure, the vibration exciter sends different excitations in different time periods, for example, once an hour, the vibration exciter sends 200N excitation in the first hour, 400N excitation in the second hour, 600N excitation in the third hour, and so on. The sensor can be used to collect the vibration signal at each measuring point under each external excitation; then, each vibration signal is decomposed, specifically using the Empirical Mode Decomposition (EMD) algorithm, to obtain all vibration signal components of each vibration signal;

[0054] Then, all the maximum and minimum values ​​in each vibration signal component are obtained, and the maximum value is taken as the peak value of a vibration peak. The two minimum values ​​adjacent to the peak value on both sides of the maximum value are taken as the valley values ​​of the vibration peak. The vibration signal component subsegment corresponding to the valley value-peak value-valley value is taken as a complete vibration peak, thereby obtaining all vibration peaks.

[0055] It should be noted that the implementer may also design an external excitation application scheme according to the structural bearing capacity of the bridge steel frame, and may also adopt other signal decomposition algorithms such as VMD or principal component analysis algorithm to obtain vibration signal components. These algorithms, the EMD algorithm and the acquisition of extreme values ​​are all existing technologies well known to those skilled in the art and will not be elaborated here.

[0056] Step S2, in each vibration signal component, clustering all vibration peaks to obtain all clusters; in each cluster, according to the occurrence position of each vibration peak and the waveform difference between different vibration peaks, obtaining the periodicity index of the vibration peak in each cluster.

[0057] Considering that the bridge may be affected by environmental factors such as wind and wave impact, and traffic load factors such as vehicles running on the bridge deck, the vibration signal at each measuring point does not only reflect the vibration characteristics caused by external excitation, but may also include the random vibration influence caused by such external factors. Therefore, the embodiment of the present invention needs to evaluate the regularity and periodicity of each vibration characteristic according to the fluctuation characteristics of the decomposed vibration signal components, so as to screen out the vibration characteristic component with the least external interference and closest to the external excitation vibration characteristic from each vibration signal at each measuring point; when the steel frame of the bridge has defects such as cracks or damage, the vibration characteristics of the steel frame of the bridge will break the normal regularity and periodicity, and then the performance of the steel frame of the bridge can be evaluated by comparing the vibration characteristic components with the vibration characteristics caused by external excitation.

[0058] It should be noted that the analysis and processing method for each vibration signal at each measuring point is consistent, and the vibration signal analysis method at any measuring point is only taken as an example for illustration.

[0059] Considering that after the vibration signal is decomposed, each vibration signal component may include vibration characteristics of external excitation and may also include random vibration interference, which is the superposition result of multiple vibration characteristics; and considering that clustering can gather similar data characteristics into a cluster, the embodiment of the present invention can cluster the vibration peaks in each vibration signal component based on the clustering algorithm, wherein the vibration peaks in the same cluster represent the same vibration characteristic, and all clusters reflect the superposition of multiple vibration characteristics.

[0060] Preferably, in one embodiment of the present invention, considering that the purpose of clustering is to cluster vibration peaks with similar vibration characteristics, and the waveform of the vibration peak reflects its vibration characteristics; therefore, the method for obtaining clusters includes:

[0061] In each vibration signal component, the metric distance is obtained according to the waveform difference between different vibration peaks. Based on the metric distance between different vibration peaks and the preset K value, K-means clustering is performed on all vibration peaks to obtain all clusters.

[0062] As an example, considering that the DTW distance can evaluate the waveform differences between different vibration peaks, the DTW distance between any two vibration peaks is used as the metric distance, and the elbow method is used to obtain the preset K value. Then, based on the K-means clustering algorithm, all vibration peaks in each vibration signal component are clustered to obtain all clusters.

[0063] In other examples, implementers may also use the product of the peak height and peak width of each vibration peak as a waveform characteristic parameter, and use the absolute value of the difference between the waveform characteristic parameters of different vibration peaks as a metric distance for clustering; implementers may also evaluate the waveform characteristic parameters based on other waveform characteristics of the vibration peak, such as the rate of change from valley to peak value, to obtain clusters.

[0064] It should be noted that DTW distance, elbow method and K-means clustering algorithm are all existing technologies, and implementers may also use other distance clustering algorithms, which will not be described in detail here.

[0065] After obtaining all clusters of all vibration peaks in each vibration signal component, the periodicity index of the vibration peaks in each cluster can be obtained according to the occurrence position of each vibration peak and the waveform difference between different vibration peaks. The periodicity index reflects the periodicity of the vibration characteristics of the corresponding species of the vibration peaks in each cluster. The larger the periodicity index, the more similar the periodic vibration characteristics of the vibration peaks in the cluster, which makes it easier to subsequently evaluate whether the vibration signal component has periodic vibration characteristics by combining the periodicity indexes of all clusters, so as to screen out the vibration characteristic components from each vibration signal.

[0066] Preferably, in one embodiment of the present invention, the method for obtaining the periodicity index includes:

[0067] In each cluster, the concentrated features of the DTW distances between all different vibration peaks and the discrete features of the time intervals between all two adjacent vibration peaks are integrated to obtain the first periodic parameter of the vibration peaks in each cluster; the concentrated features of the DTW distances and the discrete features of the time intervals are negatively correlated with the first periodic parameter;

[0068] In each cluster, the time interval between the first vibration peak and the last vibration peak in the time sequence of the corresponding vibration signal component is used as the period duration of the vibration peak in the corresponding cluster; the ratio of the period duration of each cluster to the total duration of the corresponding vibration signal component is used as the second periodicity parameter of the vibration peak in each cluster;

[0069] The first periodicity parameter and the second periodicity parameter are fused to obtain the periodicity index of the vibration peak in each cluster.

[0070] As an example, taking any vibration signal component as an example, the calculation formula of the periodicity index of each cluster is:

[0071] ; Where i is the cluster number; is the periodicity index of the ith cluster; is the mean of the DTW distances between all different vibration peaks in the i-th cluster; is the variance of the time intervals between all two sequentially adjacent vibration peaks in the i-th cluster; exp() is an exponential function with the natural constant e as the base; is the first periodic parameter of the vibration peak in the i-th cluster; is the duration of the period of the vibration peak in the i-th cluster; is the total duration of the vibration signal component; is the second periodic parameter of the vibration peak within the i-th cluster.

[0072] It should be noted that the time interval between two vibration peaks that are adjacent in time sequence in a cluster is the time interval between the corresponding moments of the peak values ​​of the two vibration peaks; the cycle duration is the time interval between the corresponding moments of the peak values ​​of the first vibration peak and the last vibration peak in the time sequence in the corresponding vibration signal component; in other examples, the time interval between the corresponding left valley values ​​or the time interval between the corresponding right valley values ​​can also be used instead of the time interval between the peak values, which will not be elaborated here.

[0073] In the calculation formula of the periodicity index, the smaller the mean of the DTW distances between all different vibration peaks in the cluster, the more similar all the vibration peaks in each cluster are, and the more periodic vibration characteristics they have; the smaller the variance of the time intervals between all two sequentially adjacent vibration peaks in the cluster, the more periodic vibration characteristics the vibration peaks have; the mean and variance are multiplied and merged to perform negative correlation mapping, adjust the logic and normalize, the smaller the product, the larger the first periodicity parameter; the greater the proportion of the cycle duration in the total duration, the greater the second periodicity; and then the first periodicity parameter is multiplied and merged with the second periodicity parameter to obtain the periodicity index.

[0074] In another embodiment of the present invention, the implementer may also use the absolute value of the difference between the waveform characteristic parameters of different vibration peaks in each cluster instead of the DTW distance to obtain the first periodic parameter; the implementer may also use other negative correlation mapping methods such as inverse, or may use other basic mathematical operations to combine the first periodic parameter and the second periodic parameter, which will not be repeated here.

[0075] Step S3, based on the clustering results of the vibration peaks in the vibration signal components, each vibration signal component is divided to obtain all component subsegments; in each vibration signal component, according to the centralized characteristics of all component subsegments and the fluctuation differences between different component subsegments, combined with the difference in the number of vibration peaks in different clusters and the periodic index of all clusters, the representative coefficient of each vibration signal component is obtained.

[0076] Considering that each vibration signal component can be regarded as the superposition result of multiple vibration characteristics, and the vibration signal component will also have a certain periodic vibration characteristic under the premise of being affected by other random vibration interference; considering that all clusters of each vibration signal component reflect the superposition of multiple vibration characteristics; therefore, the embodiment of the present invention first divides each vibration signal component based on the clustering result of the vibration peak in the vibration signal component, and obtains all component sub-segments. Each component sub-segment can be regarded as a vibration characteristic sub-segment after superposition, and then the corresponding vibration signal component can be evaluated based on the component sub-segment to determine whether it has regularity and periodicity, so as to evaluate whether it is a vibration characteristic component.

[0077] Preferably, in one embodiment of the present invention, the method for obtaining component sub-segments includes:

[0078] In each vibration signal component, each vibration peak is assigned a label, wherein the labels of vibration peaks in the same cluster are the same, and the labels of vibration peaks in different clusters are different;

[0079] Based on the labels of all vibration peaks in the vibration signal component, the corresponding label sequence of the vibration signal component is obtained; in the vibration signal component, the vibration peak corresponding to the label with the highest frequency of occurrence in the corresponding label sequence is used as the tail vibration peak of each component sub-segment, and the adjacent subsequent vibration peak of the tail vibration peak of each component sub-segment is used as the head vibration peak of the next adjacent component sub-segment; the component sub-segment is determined using the head vibration peak and the tail vibration peak.

[0080] As an example, in each vibration signal component, each vibration peak is assigned a label. For example, there are four clusters in total, and the cluster labels of the clusters are a, b, c, and d respectively. Then the label sequence corresponding to the vibration signal component is abcdabcdabdabcdbcdacd. The label with the highest frequency in the label sequence is d, so that the vibration peak at the beginning and the vibration peak at the end can be determined. Then the division results of the label sequence corresponding to the vibration signal component are: abcd, abcd, abd, abcd, bcd, acd. Based on the division results of the label sequence, the corresponding component sub-segments can be obtained.

[0081] It should be noted that, if there are multiple tags with the highest frequency of occurrence, the vibration peak corresponding to the tag that appears latest in timing is taken as the tail vibration peak, so as to obtain a complete component sub-segment of the periodic vibration.

[0082] Considering that if the component sub-segments in the vibration signal component are more concentrated and tend to have similar fluctuation characteristics, it means that it has more periodic vibration characteristics and is more likely to be a vibration characteristic component that is least affected by external interference and closest to the external excitation vibration characteristics; considering that the closer the number of vibration peaks in all clusters corresponding to the vibration signal component is, and the higher the periodicity index of each cluster is, it means that the vibration signal component also has more periodic vibration characteristics and is more likely to be a vibration characteristic component;

[0083] Therefore, after obtaining all component subsegments, the embodiment of the present invention can obtain the representative coefficient of each vibration signal component in each vibration signal component according to the centralized characteristics of all component subsegments and the fluctuation differences between different component subsegments, combined with the difference in the number of vibration peaks in different clusters and the periodic index of all clusters; the representative coefficient quantifies the possibility that each vibration signal component is a vibration characteristic component, providing a basis for the subsequent screening of vibration characteristic components.

[0084] Preferably, in one embodiment of the present invention, the method for obtaining the representative coefficient includes:

[0085] See also Figure 2 , which shows a flow chart of a method for obtaining a representative coefficient provided by an embodiment of the present invention, specifically comprising:

[0086] Step S301, obtaining the label subsequence corresponding to each component subsegment in each vibration signal component, and taking the label subsequence with the highest occurrence frequency as the representative subsequence; taking the ratio of the total number of representative subsequences to the total number of all label subsequences as the first representative parameter.

[0087] Considering that the higher the frequency of component sub-segments appearing in the vibration signal component, and the greater the proportion of such component sub-segments in the vibration signal component, the more likely the vibration signal component is to show the vibration characteristics of the component sub-segment as a periodic wave;

[0088] Therefore, one embodiment of the present invention first obtains the label subsequence corresponding to each component subsegment in each vibration signal component, selects the label subsequence with the highest occurrence frequency as the representative subsequence, and then takes the total number of representative subsequences as the numerator and the total number of all label subsequences as the denominator, and the obtained quantity ratio is taken as the first representative parameter; the larger the first representative parameter is, the more periodic vibration characteristics the vibration component signal has, and the greater the possibility of being a vibration characteristic component.

[0089] Step S302, in the label sequence corresponding to each vibration signal component, the similarities between all different label subsequences are integrated to obtain the second representative parameter; the similarity is positively correlated with the second representative parameter.

[0090] Considering that the more similar the component sub-segments in the vibration signal component are, the more periodic vibration characteristics the vibration component signal has, and the label can better help evaluate the similarity;

[0091] Therefore, as an example, in the label sequence corresponding to each vibration signal component, the label subsequences corresponding to different component subsegments are compared. For example, the two label subsequences are abcd and abd, and the total number of repeated labels is 3. Then the similarity parameter between the label subsequences corresponding to the different component subsegments is 3. The larger the similarity parameter, the greater the similarity. Then, the similarity parameters between all different label subsequences are obtained, and the mean of the similarity parameters is used as the second representative parameter. The larger the mean, the larger the second representative parameter, and the greater the possibility that the vibration component signal is a vibration characteristic component.

[0092] In other embodiments, the implementer may also calculate the DTW distance of the label subsequences corresponding to different component subsegments, and perform negative correlation mapping on the DTW distance, such as inverse or exponential negative correlation mapping, to compare the differences and evaluate the similarity, thereby obtaining the second representative parameter; the similarity may also be evaluated by using the ratio of the similarity parameter to the longest length of the two label subsequences, and other similarity comparison methods may also be used, which are all existing technologies and will not be described in detail here.

[0093] Step S303: In each vibration signal component, multiply the negative correlation mapping result of the variance of the total number of vibration peaks in all clusters by the sum of the periodicity indexes of all clusters to obtain the regularity index of each vibration signal component.

[0094] Since the variance can reflect whether the data has a large volatility, it can also reflect whether the data is close or similar. Since the larger the variance, the more dissimilar the data, the variance can be negatively correlated to adjust the logic.

[0095] As an example, the regularity index is calculated as: ;in, is the serial number of the vibration signal component; is the regularity index of the mth vibration signal component; exp() is an exponential function with the natural constant e as the base; is the variance of the total number of vibration peaks in all clusters corresponding to the mth vibration signal component; i is the cluster number; n is the total number of clusters corresponding to the mth vibration signal component; is the periodicity index of the ith cluster.

[0096] In this example, the variance of the total number of vibration peaks in all clusters is taken as the x in the exponential function exp(-x) with the natural constant e as the base, and the negative correlation mapping adjustment logic has been performed; the larger the negative correlation mapping value, the closer the number of vibration peaks in each cluster, and the larger the cumulative value of the periodic index of all clusters, the more regular the periodic vibration characteristics of the vibration signal component are, and the greater the possibility of being a vibration characteristic component.

[0097] In other examples, implementers may also use other negative correlation mapping means, and then combine the negative correlation mapping value and the accumulated value through other basic mathematical operations or related mapping means such as addition or weighted summation, which will not be described in detail here.

[0098] Step S304: The first representative parameter, the second representative parameter and the regularity index are integrated to obtain a representative coefficient of each vibration signal component.

[0099] As an example, the three are multiplied and combined, and then the product is linearly normalized, and the normalized result is used as the representative coefficient of the corresponding vibration signal component; in other examples, the implementer may also use other basic mathematical operations such as addition or weighted summation or related mapping means to combine the three, and may also use other normalization methods, which will not be repeated here.

[0100] Step S4, at each measuring point, based on the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, the vibration characteristic component is screened out from all the vibration signal components of the vibration signal of each external excitation; the performance of the bridge steel frame is evaluated by comprehensively analyzing the similarity between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components.

[0101] After obtaining the table coefficients of each vibration signal component, at each measuring point, combined with the similarities between the vibration signal components in the vibration signals with different external excitations, the vibration characteristic components in each vibration signal that are least affected by external interference and closest to the external excitation vibration characteristics can be evaluated and screened out.

[0102] Preferably, in one embodiment of the present invention, considering that at the same measuring point, the vibration characteristic components in the vibration signals under different external excitations should have similar vibration characteristics, that is, the vibration amplitudes are different, but the vibration periodicities are similar; and considering that the vibration characteristic components should have higher periodic vibration characteristics, the corresponding representative coefficients should also be higher; therefore, the method for obtaining the vibration characteristic components includes:

[0103] At any measuring point, any externally excited vibration signal is taken as the target vibration signal, any vibration signal component in the target vibration signal is taken as the target vibration component, and all vibration signal components of all non-target vibration signals are taken as reference vibration components of the target vibration component;

[0104] Between the target vibration component and each reference vibration component, the product of the mean of the representative coefficients of the corresponding components and the Pearson correlation coefficient between the corresponding components is used as the correlation matching coefficient between the corresponding components;

[0105] According to all relevant matching coefficients between each vibration component signal and all corresponding reference vibration components, the vibration characteristic component of the target vibration signal is screened out from all vibration signal components of the target vibration signal.

[0106] In a preferred embodiment of the present invention, the maximum correlation matching coefficient is screened out from all correlation matching coefficients, and the vibration signal component corresponding to the maximum correlation matching coefficient is used as the vibration characteristic component of the target vibration signal.

[0107] As an example, first, at any measuring point, a target vibration signal is determined, and any vibration signal component in the target vibration signal is taken as the target vibration component for analysis, and each vibration signal component in the non-target vibration signal under other external excitations is taken as a reference vibration component of the target vibration component; then, a correlation matching coefficient between the target vibration component and each reference vibration component is calculated; the larger the correlation matching coefficient, the more likely it is that the two components have similar vibration characteristics, and are more likely to be vibration characteristic components under different external excitations;

[0108] The calculation formula of the correlation matching coefficient is: ; Where A is the serial number of the target vibration component; B is the serial number of the reference vibration component; is the correlation matching coefficient between the target vibration component A and the Bth reference vibration component; is the representative coefficient of the target vibration component A; is the representative coefficient of the Bth reference vibration component; is the Pearson correlation coefficient between the target vibration component A and the Bth reference vibration component; is the mean of the representative coefficients of the target vibration component A and the Bth reference vibration component.

[0109] In this example, the larger the mean value of the representative coefficient is, the more likely it is that the target vibration component and the reference vibration component are vibration characteristic components under different external excitations; the larger the Pearson correlation coefficient is, the more likely it is that the two components have similar vibration characteristics, which further indicates that the target vibration component and the reference vibration component are more likely to be vibration characteristic components under different external excitations; the two are then multiplied and combined to obtain the correlation matching coefficient.

[0110] In other examples, implementers may also use other similarity measurement methods, such as negative correlation mapping of DTW distance, or combine the mean and Pearson correlation coefficient using other basic mathematical operations or related mapping methods such as addition or weighted summation, which are all existing technologies and will not be described in detail here.

[0111] By changing the target vibration component, all the correlation matching coefficients between each vibration component signal and all the corresponding reference vibration components can be obtained, and then a maximum correlation matching coefficient is screened out from all the corresponding correlation matching coefficients of all vibration component signals in the target vibration signal and other reference vibration components, and the vibration signal component corresponding to the maximum correlation matching coefficient is used as the vibration characteristic component of the target vibration signal.

[0112] In another embodiment of the present invention, it is assumed that the target vibration signal is Ax, which is decomposed into three vibration signal components, namely A1, A2, and A3; the non-target vibration signals are Bx, Cx, and Dx, and the corresponding vibration signal components are B1, B2, B3, C1, C2, C3, D1, D2, and D3, respectively, and each vibration signal component in the non-target vibration signal is used as a reference vibration component; under the target vibration signal, the correlation matching coefficient between A1 and each reference vibration component is calculated, the correlation matching coefficient between A2 and each reference vibration component is calculated, and the correlation matching coefficient between A3 and each reference vibration component is calculated;

[0113] Among them, in each calculation process, for example, the largest correlation matching coefficients are screened out as the correlation matching coefficients between A1-B1, A1-C1, and A1-D1, and the three correlation matching coefficients are averaged to obtain the characteristic evaluation parameter. The larger the characteristic evaluation parameter, the more likely it is that A1 has similar periodic vibration characteristics to the vibration signal components in the vibration signal under other different external excitations, and they are more likely to be vibration characteristic components under different external excitations. Similarly, the characteristic evaluation parameters of A2 and A3 are calculated, and then the characteristic evaluation parameters of the three vibration signal components in the target vibration signal are compared, and the vibration signal component with the largest characteristic evaluation parameter is taken as the vibration characteristic component of the target vibration signal.

[0114] By changing the target vibration signal, the vibration characteristic component of the vibration signal of each external excitation can be obtained. Furthermore, the similarity between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components can be integrated to evaluate the performance of the bridge steel structure.

[0115] Preferably, in one embodiment of the present invention, the method for evaluating the performance of a bridge steel frame comprises:

[0116] At each measuring point, the DTW distance between the vibration signal of each external excitation and the corresponding vibration characteristic component is accumulated to obtain the vibration difference parameter;

[0117] At each measuring point, the DTW distance between each externally excited vibration signal and the corresponding vibration characteristic component is used as a data point to fit the data curve; the absolute value of the slope of the data curve is used as the vibration change parameter;

[0118] According to the vibration change parameters and difference parameters at all measuring points, the damage degree of the bridge steel frame is obtained and the performance of the bridge steel frame is evaluated; the vibration change parameters and difference parameters are positively correlated with the damage degree.

[0119] As an example, the damage level is calculated as: ; Where S is the damage degree of the bridge steel frame; is the vibration variation parameter at the pth measuring point; q is the serial number of the vibration signal under external excitation; is the total number of vibration signals at the pth measuring point; p is the serial number of the measuring point; P is the total number of measuring points; is the DTW distance between the vibration signal of the qth external excitation at the pth measuring point and the corresponding vibration characteristic component; is the vibration difference parameter at the pth measuring point.

[0120] In this example, the slope of the fitted data curve at each measuring point is obtained based on a two-point calculation formula. When the slope is closer to 0, it means that the vibration signals under different external excitations are more similar to the corresponding vibration characteristic components, and the more consistent with the vibration characteristics of the bridge under external excitation; but when the absolute value of the slope is larger, it means that the similarity between the vibration signals under different external excitations and the corresponding vibration characteristic components has changed. The larger the vibration change parameter, the more likely it is caused by defects such as cracks; the larger the DTW distance between the vibration signals of all external excitations and the corresponding vibration characteristic components, it also indirectly indicates that the similarity between the vibration signals under different external excitations and the corresponding vibration characteristic components has changed, and the larger the vibration difference parameter; then the vibration change parameter at each measuring point is multiplied and fused with the vibration difference parameter to obtain the average, and the vibration characteristics at all measuring points are comprehensively used to evaluate the damage degree of the bridge steel frame. The damage degree reflects the performance. The greater the damage degree, the lower the performance.

[0121] In another embodiment of the present invention, the implementer may also normalize the degree of damage and set a damage threshold. When the degree of damage is greater than a preset threshold, such as 0.8, the bridge is judged to be severely damaged and requires inspection and maintenance to ensure its safety and stability.

[0122] The present invention also provides a performance testing device for a steel frame, the device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the steps of a performance testing method for a steel frame when executing the computer program.

[0123] The present invention also provides a performance testing system for a steel frame, see Figure 3 , which shows a system block diagram of a steel frame performance testing system provided by an embodiment of the present invention, the system includes a steel frame vibration data acquisition module 101, a steel frame vibration data analysis module 102 and a steel frame performance evaluation module 103; each module in the system can execute the steps in the above-mentioned steel frame performance testing method to perform performance evaluation of the bridge steel frame.

[0124] Steel frame vibration data acquisition module 101: obtains all vibration signal components of the vibration signal at each measuring point of the bridge steel frame under each external excitation, and obtains all vibration peaks in each vibration signal component;

[0125] Steel frame vibration data analysis module 102: In each vibration signal component, all vibration peaks are clustered to obtain all clusters; in each cluster, according to the waveform characteristics of each vibration peak and the waveform differences between different vibration peaks, the periodicity index of the vibration peaks in each cluster is obtained; based on the clustering results of the vibration peaks in the vibration signal component, each vibration signal component is segmented to obtain all component sub-segments; in each vibration signal component, according to the centralized characteristics of all component sub-segments and the fluctuation differences between different component sub-segments, combined with the number differences of vibration peaks in different clusters and the periodicity indexes of all clusters, the representative coefficient of each vibration signal component is obtained; at each measuring point, according to the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, the vibration characteristic component is screened out from all vibration signal components of each vibration signal of external excitation;

[0126] Steel frame performance evaluation module 103: The performance of the bridge steel frame is evaluated by comprehensively analyzing the similarities between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components.

[0127] In summary, the present invention first obtains all vibration signal components of the vibration signal at each measuring point of the bridge steel frame under each external excitation, and all vibration peaks therein; then clusters all vibration peaks in each vibration signal component to obtain all clusters; further obtains the periodic index of the vibration peak in each cluster; then based on the clustering result, segments each vibration signal component to obtain all component sub-segments, and then obtains the representative coefficient of each vibration signal component; then at each measuring point, the vibration characteristic component is screened out from all vibration signal components of the vibration signal of each external excitation; finally, the similarity between the vibration signal at all measuring points under each external excitation and the corresponding vibration characteristic component is comprehensively evaluated to evaluate the performance of the bridge steel frame. Based on the characteristic that the vibration signal under external excitation has periodic vibration characteristics, the present invention analyzes and screens out the corresponding vibration characteristic components, reduces the interference of other random vibration factors such as the environment or traffic, and accurately evaluates the performance of the bridge steel frame.

[0128] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0129] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A performance testing method for a steel frame, characterized in that: The method comprises: Obtain all vibration signal components of the vibration signal of the bridge steel structure at each measuring point under each external excitation, and obtain all vibration peaks in each of the vibration signal components; In each of the vibration signal components, all the vibration peaks are clustered to obtain all clusters; in each cluster, according to the occurrence position of each vibration peak and the waveform difference between different vibration peaks, the periodicity index of the vibration peak in each cluster is obtained; Based on the clustering results of the vibration peaks in the vibration signal components, each of the vibration signal components is segmented to obtain all component subsegments; in each of the vibration signal components, according to the centralized features of all the component subsegments and the fluctuation differences between different component subsegments, combined with the number differences of vibration peaks in different clusters and the periodicity indexes of all clusters, the representative coefficient of each vibration signal component is obtained; At each measuring point, based on the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, a vibration characteristic component is screened out from all the vibration signal components of the vibration signal of each external excitation; the performance of the bridge steel frame is evaluated by comprehensively analyzing the similarity between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components.

2. A steel frame performance testing method according to claim 1, characterized in that: The cluster acquisition method includes: In each of the vibration signal components, a metric distance is obtained according to the waveform difference between different vibration peaks, and based on the metric distance between different vibration peaks and a preset K value, K-means clustering is performed on all the vibration peaks to obtain all clusters.

3. A steel frame performance testing method according to claim 1, characterized in that: The method for obtaining the periodicity index includes: In each cluster, the concentrated features of the DTW distances between all different vibration peaks and the discrete features of the time intervals between all two vibration peaks that are adjacent in time sequence are integrated to obtain the first periodic parameter of the vibration peaks in each cluster; the concentrated features of the DTW distances and the discrete features of the time intervals are both negatively correlated with the first periodic parameter; In each cluster, the time interval between the first vibration peak and the last vibration peak in the time sequence of the corresponding vibration signal component is used as the period duration of the vibration peak in the corresponding cluster; the ratio of the period duration of each cluster to the total duration of the corresponding vibration signal component is used as the second periodic parameter of the vibration peak in each cluster; The first periodicity parameter and the second periodicity parameter are fused to obtain a periodicity index of a vibration peak in each cluster.

4. A steel frame performance testing method according to claim 1, characterized in that: The method for obtaining the component sub-segments includes: In each of the vibration signal components, assigning a label to each of the vibration peaks, wherein the labels of vibration peaks in the same cluster are the same, and the labels of vibration peaks in different clusters are different; Based on the labels of all the vibration peaks in the vibration signal component, a label sequence corresponding to the vibration signal component is obtained; in the vibration signal component, the vibration peak corresponding to the label with the highest frequency of occurrence in the label sequence is used as the tail vibration peak of each component sub-segment, and the adjacent vibration peak after the tail vibration peak of each component sub-segment is used as the head vibration peak of the next adjacent component sub-segment; the component sub-segment is determined using the head vibration peak and the tail vibration peak.

5. A steel frame performance testing method according to claim 4, characterized in that: The method for obtaining the representative coefficient includes: Obtaining a label subsequence corresponding to each component subsegment in each vibration signal component, and taking the label subsequence with the highest occurrence frequency as a representative subsequence; taking the ratio of the total number of the representative subsequences to the total number of all label subsequences as a first representative parameter; In the label sequence corresponding to each vibration signal component, the similarities between all different label subsequences are integrated to obtain a second representative parameter; the similarity is positively correlated with the second representative parameter; In each of the vibration signal components, multiplying the negative correlation mapping result of the variance of the total number of the vibration peaks in all clusters by the sum of the periodicity indexes of all clusters to obtain the regularity index of each of the vibration signal components; The first representative parameter, the second representative parameter and the regularity index are integrated to obtain a representative coefficient of each vibration signal component.

6. A steel frame performance testing method according to claim 1, characterized in that: The method for obtaining the vibration characteristic component comprises: At any measuring point, any externally excited vibration signal is taken as the target vibration signal, any vibration signal component in the target vibration signal is taken as the target vibration component, and all the vibration signal components of all non-target vibration signals are taken as reference vibration components of the target vibration component; Between the target vibration component and each reference vibration component, the product of the mean of the representative coefficients of the corresponding components and the Pearson correlation coefficient between the corresponding components is used as the correlation matching coefficient between the corresponding components; According to all the correlation matching coefficients between each vibration component signal and all corresponding reference vibration components, the vibration characteristic component of the target vibration signal is screened out from all the vibration signal components of the target vibration signal.

7. A steel frame performance testing method according to claim 6, characterized in that: The method for screening out the vibration characteristic component of the target vibration signal from all vibration signal components of the target vibration signal comprises: The maximum correlation matching coefficient is screened out from all the correlation matching coefficients, and the vibration signal component corresponding to the maximum correlation matching coefficient is used as the vibration characteristic component of the target vibration signal.

8. A steel frame performance testing method according to claim 1, characterized in that: The method for evaluating the performance of a bridge steel frame comprises: At each measuring point, the DTW distance between the vibration signal of each external excitation and the corresponding vibration characteristic component is accumulated to obtain a vibration difference parameter; At each measuring point, the DTW distance between each externally excited vibration signal and the corresponding vibration characteristic component is used as a data point to fit a data curve; the absolute value of the slope of the data curve is used as a vibration change parameter; According to the vibration change parameters and the difference parameters at all measuring points, the damage degree of the bridge steel frame is obtained and the performance of the bridge steel frame is evaluated; the vibration change parameters and the difference parameters are both positively correlated with the damage degree.

9. A performance testing device for a steel frame, characterized in that: The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of a method for testing the performance of a steel structure as described in any one of claims 1 to 8 are implemented.

10. A performance testing system for a steel frame, characterized in that: The system comprises: Steel frame vibration data acquisition module: obtains all vibration signal components of the vibration signal at each measuring point of the bridge steel frame under each external excitation, and obtains all vibration peaks in each of the vibration signal components; Steel frame vibration data analysis module: in each vibration signal component, all the vibration peaks are clustered to obtain all clusters; in each cluster, according to the waveform characteristics of each vibration peak and the waveform differences between different vibration peaks, the periodicity index of the vibration peak in each cluster is obtained; based on the clustering results of the vibration peaks in the vibration signal component, each vibration signal component is divided to obtain all component sub-segments; in each vibration signal component, according to the centralized characteristics of all the component sub-segments and the fluctuation differences between different component sub-segments, combined with the number differences of vibration peaks in different clusters and the periodicity index of all clusters, the representative coefficient of each vibration signal component is obtained; at each measuring point, according to the similarity between the vibration signal components in the vibration signals of different external excitations and the representative coefficient of each vibration signal component, the vibration characteristic component is screened out from all the vibration signal components of each externally excited vibration signal; Steel frame performance evaluation module: The performance of the bridge steel frame is evaluated by comprehensively analyzing the similarities between the vibration signals at all measuring points under each external excitation and the corresponding vibration characteristic components.

Images