Detection method for judging quality of cable-stayed pipe

Abstract

The invention discloses a detection method for judging the quality of a cable-stayed pipe, relates to the technical field of cable-stayed pipe quality detection, and is used for solving the problem of poor cable-stayed pipe quality detection precision. The optical fiber sensors are arranged on the surface of the cable-stayed pipe to collect strain data, a continuous space-time strain field is constructed, local gradient changes are extracted, identification of a micro-damage area is achieved, a node response frequency spectrum is obtained in combination with a multi-frequency-band acoustic excitation means, nonlinear energy and frequency coupling characteristics are extracted, and the micro-damage area is identified. Calculating a nonlinear display probability to locate a potential nonlinear response concentration area of the structure, constructing a topological graph model of the structure, reflecting a mechanical connection relation between nodes and monitoring abnormal response, when an abnormal structure behavior is detected, carrying out topological structure state analysis on the area where the abnormal structure behavior is located, extracting key information such as structure coupling and state propagation capability, and carrying out state analysis on the abnormal structure behavior; and high and low risk signals are generated according to a judgment result, so that the accurate identification and intelligent regulation and control of the quality change of the cable-stayed pipe are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of stay cable detection, and more specifically, to a detection method for judging the quality of stay cables. Background Art

[0002] Stay cables are key load-bearing components in the cable-stayed bridge structure. They are usually in a tension state, connecting the bridge tower and the bridge deck, and are used to bear and transmit various actions such as vehicle loads, wind loads, self-weights, and bridge deck disturbances. As the main load-bearing structural elements, they directly affect the overall stability, dynamic response characteristics, and service life of the bridge. Stay cables usually adopt a composite structure of high-strength steel strands, metal sheaths, and polymer protective layers to ensure their tensile performance, corrosion resistance, and service stability. In terms of structural functions, stay cables not only need to provide sufficient strength and stiffness to resist the tensile effects caused by dead loads and live loads, but also must have good fatigue tolerance and seismic performance to cope with high-frequency dynamic disturbances under daily traffic, wind vibration, and earthquakes.

[0003] Deficiencies of the prior art: The current structural health monitoring technology for stay cables generally relies on traditional methods such as low-frequency modal parameter changes, single-point strain acquisition, or static threshold judgment. There are problems such as the inability to identify early micro-damages, ignoring the coupling relationship between multi-node states, and lacking a dynamic tracking and systematic early warning mechanism for abnormal behaviors. Especially in the initial stage of structural micro-crack initiation and fatigue degradation, there are often non-significant local disturbances. Traditional methods are difficult to achieve effective identification without frequency shift or visible deformation, and cannot conduct systematic modeling and dynamic grading response for potential abnormal propagation paths, key weak point nodes, and their linkage behaviors, severely limiting the intelligent maintenance ability and early risk intervention level of the key infrastructure of cable-stayed bridges. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, there are the following solutions to solve the problem of poor detection accuracy of stay cable quality in the above-mentioned background art.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A detection method for judging the quality of stay cables, comprising the following steps:

[0007] Arrange fiber optic sensors on the surface of the stay cable and collect signals, construct a continuous spatio-temporal strain field and extract local strain gradient information, identify the strain mutation region, perform time series modeling based on the strain field, and identify the structural regions with micro-damages;

[0008] Apply multi - frequency excitation signals to the stay - cable structure to determine the node response, obtain the frequency - domain response spectrum, identify the non - linear spectral characteristics, extract the non - linear energy index and the cross - band coupling path strength, calculate the non - linear manifestation probability to determine the potential abnormal excitation potential of the node, and locate the concentrated area of potential non - linear response of the structure;

[0009] Construct a structure diagram based on the stay - cable structure and the sensing nodes, simulate the connection and force - transmission relationship between the node structures, and monitor whether there are abnormal structural behaviors in the nodes;

[0010] If there are abnormal structural behaviors, obtain the structural state information generated during the topological analysis of the stay - cable area where the nodes with abnormal structural behaviors are located, analyze it, and perform regulation on the stay - cable according to different signals generated by the analysis results.

[0011] In a preferred embodiment, fiber optic sensors are arranged on the surface of the stay - cable to collect signals, a continuous spatio - temporal strain field is constructed, and local strain gradient information is extracted to identify the strain mutation area. The specific steps are as follows:

[0012] Arrange fiber Bragg grating sensors on the surface of the stay - cable at regular intervals, and arrange sensing nodes on each optical fiber. Each node independently measures the axial strain response at its location.

[0013] The fiber Bragg grating sensors collect and record the wavelength drift signals of all measurement points and perform strain conversion.

[0014] Perform spatial strain interpolation and local gradient extraction on the collected signals. For each time point, use cubic spline interpolation as the interpolation function within the measurement point interval through the known strain values of the measurement points for interpolation.

[0015] Perform interpolation for each time period during the monitoring period. Each time period corresponds to a strain spatial distribution curve. After combining all time periods, a two - dimensional strain field of the stay - cable during the entire monitoring period is obtained.

[0016] Calculate the first - order spatial derivative of the interpolation function at the position of the node for each time point, and identify the possible areas of local concentrated damage of the structure based on the calculated first - order spatial derivative.

[0017] In a preferred embodiment, perform time - series modeling based on the strain field to identify the structural areas with micro - damage. The specific steps include:

[0018] Based on the method of dynamic deviation modeling and statistical residual analysis, identify the early potential damage areas of the structure;

[0019] For each node x i , calculate the offset function between the average strain behavior within the current period and the reference state: ΔYB(x i , tk ) = YB(x i , t k ) - YB ref (x i ), where YB(x i , t k ) is the strain variable at the current time t k ; YB ref (x i ) is the reference value of the strain variable; the offset function represents the instantaneous offset value of the strain at the spatial position of the node relative to the healthy state at time t k .

[0020] For each node x i , construct the offset trajectory curve over the entire time period and define the damage criterion function: where is the time curvature of the strain offset; spatial gradient;

[0021] Set the abnormal intensity threshold QY. When SH i > QY, take the actual physical position collected by the sensor as the micro-damage area and summarize the micro-damage areas as the potential micro-damage point set.

[0022] In a preferred embodiment, apply multi-band excitation signals to the stay cable structure to determine the node response, obtain the frequency-domain response spectrum and identify the non-linear spectral characteristics, and extract the non-linear energy index and the cross-band coupling path strength. The specific steps are as follows:

[0023] Apply continuous frequency scanning excitation or step frequency pulse excitation to the stay cable structure;

[0024] Analyze the second harmonic components, third harmonics, and non-integer multiple frequency components existing in the spectral structure for each response point and define the non-linear energy index as: In the formula, Z harm represents all non-main frequency band sets, and R i (f) is the spectral response of the i-th node, and f is the frequency;

[0025] Construct the cross-coupling tensor for all node response spectra as the coupling path strength: where O(·) represents taking the real part; R i (f p ) is the response of node i at frequency f p ; is the conjugate response of node j at frequency f p .

[0026] In a preferred embodiment, the abnormal excitation potential of the nodes is determined by calculating the non - linear manifestation probability, and the potential non - linear response concentration area of the structure is located, including the following steps:

[0027] Fuse the non - linear energy index and the coupling path analysis results, construct a spatially distributed non - linear manifestation probability field, and perform local enhancement detection and subsequent path reconstruction;

[0028] Accumulate the non - linear energy index of each node and the intensity of the coupling paths involved. The expression for estimating the non - linear manifestation probability of the node is: Where: is the coupling path intensity between node i and adjacent node j in the cross - frequency band; Z is a normalization constant to ensure that ∑ i P i = 1; P i represents the relative probability that node i becomes a non - linear manifestation area, represents the non - linear energy index of node i;

[0029] The non - linear manifestation probability field is used to describe the probability that each node is a non - linear excitation point;

[0030] Locate the potential non - linear response concentration area at the structural level through the non - linear manifestation probability field.

[0031] In a preferred embodiment, according to the stay - cable structure and the sensing nodes, a structure diagram is constructed to simulate the connection and force - transmission relationship between the node structures, and whether there are abnormal structural behaviors of the nodes is monitored, including the following steps:

[0032] Abstract the relationship between the stay - cable structure and the sensing nodes into a graph structure. Each measurement point or monitoring unit is used as a node in the graph, and graph edges are established between every two physically adjacent, structurally interconnected or significantly response - coupled nodes;

[0033] The graph node set is that each node corresponds to a measurement point or a structural unit;

[0034] If there is a structural connection or energy coupling relationship between the edges in the graph edge set, an undirected edge is established, and the eigenvector of each node is constructed. The eigenvector of the node includes the non - linear manifestation probability of the node, the characteristic frequency position of the node in the frequency domain, and the coupling path intensity of the node;

[0035] Determine whether there are abnormal structural behaviors of the monitoring nodes according to the non - linear manifestation probability of the nodes.

[0036] In a preferred embodiment, if there are abnormal structural behaviors, the structural state information generated during the topological analysis of the stay - cable area where the nodes with abnormal structural behaviors are located is obtained, including the following steps:

[0037] Obtaining structural state information generated during the topological analysis of the cable-stayed tube region where the constant structural behavior node is located. The structural state information includes structural coupling information and state propagation information.

[0038] The structural coupling information includes the structural coupling sparsity index, and the state propagation information includes the abnormal state conduction index;

[0039] The obtained structural coupling sparsity index and abnormal state conduction index are combined to generate an abnormality determination coefficient;

[0040] The structural coupling sparse index and abnormal state conductivity index are positively correlated with the abnormality determination coefficient.

[0041] In a preferred embodiment, the following steps are included:

[0042] Compare the generated abnormality determination coefficient with the set abnormality determination threshold;

[0043] If the abnormality determination coefficient is greater than or equal to the abnormality determination threshold, a high-risk signal for the inclined-stayed tube structure is generated, indicating that there is a linkage risk in the inclined-stayed tube structure, and inspection and reinforcement intervention of the corresponding inclined-stayed tube are carried out;

[0044] If the abnormality determination coefficient is less than the abnormality determination threshold, a low-risk signal for the inclined-stayed tube structure is generated, and the inclined-stayed tube structure monitoring continues without triggering additional intervention measures.

[0045] The technical effects and advantages of the detection method for judging the quality of inclined-stayed pipes of the present invention are as follows:

[0046] The present invention collects strain data by deploying optical fiber sensors on the surface of the inclined-stayed tube, constructs a continuous spatiotemporal strain field, and extracts local gradient changes to identify micro-damaged areas. The method combines multi-band acoustic excitation to obtain the node response spectrum, extract nonlinear energy and frequency coupling characteristics, and calculate the probability of nonlinear manifestation to locate the potential nonlinear response concentration area of the structure. A structural topology model is further constructed to reflect the mechanical connection relationship between nodes and monitor abnormal responses. When abnormal structural behavior is detected, a topological structural state analysis is performed on the area where it is located to extract key information such as structural coupling and state propagation capability. Based on the judgment results, high- and low-risk signals are generated to achieve accurate identification and intelligent regulation of quality changes of the inclined-stayed tube, which can be used for early damage detection, nonlinear response positioning, and structural risk warning, thereby improving the accuracy of perception of the operating status of the inclined-stayed tube and the efficiency of maintenance decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 The figure is a flow chart of a detection method for judging the quality of an oblique-stayed pipe according to the present invention. DETAILED DESCRIPTION

[0048] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0049] To achieve the above object, Figure 1 a structural schematic diagram of a detection method for judging the quality of stay cables of the present invention is given, which specifically includes the following steps;

[0050] Arrange fiber optic sensors on the surface of the stay cable and collect signals, construct a continuous spatio-temporal strain field and extract local strain gradient information, identify the strain mutation region, perform time series modeling based on the strain field, and identify the structural regions with micro-damage;

[0051] Apply multi-band excitation signals to the stay cable structure to determine the node response, obtain the frequency-domain response spectrum and identify the non-linear spectral characteristics, extract the non-linear energy index and the cross-band coupling path strength, calculate the non-linear manifestation probability to determine the abnormal excitation potential of the node, and locate the potential non-linear response concentration area of the structure;

[0052] Construct a structure diagram according to the stay cable structure and sensing nodes, simulate the connection force transmission relationship between node structures, and monitor whether there are abnormal structural behaviors in the nodes;

[0053] If there are abnormal structural behaviors, obtain the structural state information generated during the topological analysis of the stay cable area where the nodes with abnormal structural behaviors are located, and perform analysis, and perform regulation of the stay cable according to different signals generated by the analysis results.

[0054] Step 1, Micro-deformation reconstruction of the surface of the stay cable based on the fiber optic array. During long-term service, the stay cable may be affected by wind vibration, vehicle impact load, temperature change or prestress relaxation, etc., resulting in small non-uniform deformations on its surface. Therefore, it is necessary to perform high-precision reconstruction of the strain field on the surface of the stay cable to capture the abnormal deformation area, so as to achieve early identification of quality deterioration. The specific steps are as follows:

[0055] Arrange multiple fiber Bragg grating sensors on the surface of the stay cable at regular intervals. Several sensing nodes are arranged on each fiber, and each node can independently measure the axial strain response at its location. The arrangement method needs to consider the stress characteristics of the stay cable and be arranged on multiple axial planes (such as the windward side and the leeward side) at the same time to comprehensively capture the strain behavior under local bending, shear or tensile-compressive combined actions; for example, set the fiber arrangement path along the axial direction of the stay cable, and arrange 2 to 4 fibers (symmetrically arranged up and down, front and back) to capture the bending / tensile-compressive combined stress characteristics;

[0056] The sensing nodes are spaced 10 to 20 centimeters apart. Let the total number of measurement points be N, and the corresponding set of measurement point coordinates be {x1, x2,..., x N};

[0057] The fiber Bragg grating sensor collects the drift signal of the Bragg wavelength in real time. The collection records the wavelength drift Δλ of all measurement points at a period t (such as 100 ms) i (t k ). The strain conversion is carried out by the following formula: where λ B is the initial Bragg wavelength; p e is the effective photoelastic coefficient, which can be taken as 0.22; YB i (t k ) is the strain value of the i-th measurement point at time t k ; t k is the k-th sampling moment.

[0058] Furthermore, construct the original spatio-temporal strain matrix E raw ∈R N×T , where the i-th row is the strain time series at the sensing node x i ; the k-th column is the spatial strain distribution of all measurement points at time t k .

[0059] Since the actual measurement points are discretely distributed and cannot directly reflect the continuous surface strain state, an interpolation method is needed to spatially complete the strain values and extract the local strain gradient for stress concentration area identification;

[0060] For spatial strain interpolation and local gradient extraction, for each time point t k , through the known strain values of the measurement points {YB1(t k ),..., YB N (t k ), N is the total number of measurement points, construct an interpolation function within the measurement interval [x1, x N ; Use cubic spline interpolation;

[0061] Interpolate each time period in the monitoring period (total measurement time) to obtain a continuous function of spatio-temporal bivariate. Each time period corresponds to a strain spatial distribution curve. After combining all time periods, a two-dimensional strain field of the stay cable during the entire monitoring period is constructed. This strain field can not only reflect the overall structural deformation trend but also depict the deformation mutation phenomenon occurring in the local area;

[0062] The interpolation coefficients a i , bi and c i and d i It is determined by the following conditions: whether the interpolation function is continuous in each segment, whether the first derivative and the second derivative are continuous at the interior points, and whether the boundary conditions (natural boundary or equal derivative boundary) are satisfied;

[0063] Extract the local gradient information from the spatial strain function. The local strain gradient represents the rate of change of strain per unit length, and its physical meaning is the reflection of internal stress concentration or material geometric discontinuity;

[0064] Under the ideal stress state, the strain gradient should tend to be smoothly distributed; while in the damaged or cracked area, it often shows sudden changes or severe oscillations. Therefore, by solving the first derivative of the strain function, the strain gradient function can be obtained, which is used to identify potential high-risk areas on the surface of the stay cable;

[0065] Take the interpolation function at each time point t k and calculate the first-order spatial derivative at any position x This gradient measure quantifies the severity of local strain changes and is used to identify areas where local concentrated damage in the structure is likely;

[0066] Construct a continuous spatial function to form tensor data with improved spatial resolution; output the gradient function tensor, which is used as the key input for damage identification, and compare it with the preset gradient anomaly threshold to record the index set of abnormal positions;

[0067] After completing the spatial reconstruction of the strain field on the surface of the stay cable, further explore the evolution law of the strain field in the time dimension;

[0068] The micro-damage behavior of the structure often does not change violently at one time, but shows cumulativeness, volatility and non-stationarity under multiple load cycles. Therefore, it is necessary to systematically model the changing trend of strain over time;

[0069] According to the obtained continuous spatial strain function, perform uniform sampling on it to structure the time series data into a three-dimensional tensor: ZL ∈ R X×T×D , where X is the number of spatial sampling points (obtained by discretization after interpolation), T is the number of time steps; D is the feature dimension, with a base of 1 (only the strain value), and can be extended to multiple dimensions, such as introducing the first-order spatial derivative (strain gradient) at the same time;

[0070] The tensor represents the structural response characteristics of the surface of the stay cable in the continuous space-time domain, and is the basis for identifying the deformation trend, detecting the change pattern and extracting statistical features;

[0071] To convert complex tensor data into discriminable analysis indicators, it is necessary to extract representative statistical features from the time series at each spatial position in the tensor, including calculating the mean (representing the central tendency of the structural state), standard deviation (measuring the amplitude of fluctuations), coefficient of variation (the ratio of the standard deviation to the mean, reflecting the intensity of fluctuations), or rate of change (indicating the dynamic activity level of the structure). These features will be encoded into a two-dimensional matrix F ∈ R X×M , where M is the number of statistical features and X is the number of spatial sampling points.

[0072] For microdamage identification, during the long-term monitoring of the cable-stayed pipes of a bridge, early microdamage often does not immediately cause frequency changes or sudden mutations in deformation, but will show non-stationary phenomena such as abnormal fluctuations, strain amplification, and inconsistent responses in local areas. Based on the method of dynamic deviation modeling and statistical residual analysis, identify the early potential damage areas of the structure. The specific steps are as follows:

[0073] For each node x i , calculate the offset function between its average strain behavior during the current time period and the reference state: ΔYB(x i , t k ) = YB(x i , t k ) - YB ref (x i ), where YB(x i , t k ) is the strain at the current time t k ; YB ref (x i ) is the reference value of the strain; this function reflects the instantaneous offset value of the strain at the spatial position of the node relative to the healthy state at time t k ;

[0074] For each node x i (the actual physical position collected by the sensor), construct its offset trajectory curve during the entire time period [t1, t T , and define the following damage criterion function: Where, is the time curvature of the strain offset (reflecting the degree of sudden change in response); spatial gradient (reflecting the intensity of local stress concentration at this point);

[0075] If SH i is significantly greater than other positions, it indicates that there may be microcracks, fatigue initiation, or other strain mutation characteristics at that place;

[0076] Set the abnormal intensity threshold QY. When SH iWhen it is >QY, the actual physical position collected by the sensor is used as the micro-damage area, and the micro-damage areas are summarized as the potential micro-damage point set.

[0077] Most structural defects exhibit micro-scale characteristics in the initial stage, such as: contact non-linearity between interfaces at the connection part, periodic opening and closing of internal cracks, local stiffness reduction caused by material fatigue, etc. These phenomena do not immediately cause changes in the structural frequency, but will trigger phenomena such as energy distortion, non-linear response, and spectral leakage under the action of external excitation, which are exactly what the linear analysis method cannot perceive.

[0078] Step 2, perform non-linear vibro-acoustic coupling response analysis and detect the manifestation of structural non-linearity. The specific steps are as follows:

[0079] Collect the multi-band excitation signals of controllable sound and the responses of the structure (the actual structure of the cable-stayed pipe to be detected). Apply continuous frequency scanning excitation or step frequency pulse excitation (e.g., gradually increasing from 5 kHz to 50 kHz) to the cable-stayed pipe structure to obtain the coupled response covering a wide band. Let the excitation source be the input signal and the structural response be the nodal response signal r i (t), and its spectrum is expressed as: R i (f) = F{r i (t)}, where F represents the Fourier transform operation to obtain the frequency-domain response spectrum. Take the spectral responses of each sampling point as the input and transfer them to the next step for non-linear spectral component analysis;

[0080] In the ideal elastic state, the output response of the structure should be linearly proportional to the excitation, and only the main frequency component should be included in the spectrum. However, if there is non-linear mismatch or damage in the structure, non-main frequency components such as second harmonics, third harmonics, and intermodulation frequencies will appear in its response spectrum, reflecting non-linear distortion;

[0081] For each response point (the monitored structural position), analyze whether there are second harmonic components, third harmonics, and non-integer multiple frequency components in its spectral structure, and define the non-linear energy index: In the formula, Z harm represents all non-main frequency band sets (excluding the excitation frequency band), and R i (f) is the spectral response of the i-th node;

[0082] The non-linear energy index reflects the proportion of non-linear response energy relative to the total response, and further reveals the degree of manifestation of non-linear effects in the structure.

[0083] Local damage or microcracks can induce energy leakage from the main modal frequency band to other frequency bands, which is related to the non - linear contact, repeated friction or local buckling at the damage site. To track the energy leakage path and identify abnormal areas, it is necessary to quantitatively analyze the energy coupling path between spectra, that is, to conduct non - linear cross - band energy leakage path analysis. The specific steps are as follows:

[0084] Construct a cross - coupling tensor for all node response spectra as the coupling path strength: where, O(·) represents the operation of taking the real part; R i (f p ) is the response of node i at frequency f p ; is the conjugate response of node j at frequency f p ;

[0085] This cross - coupling tensor represents the mutual coupling path strength of energy between different frequencies among spatial nodes. If there are obvious cross - energy coupling peaks and they are concentrated in a specific frequency range, it indicates that energy may leak under the action of non - linear mechanisms inside the structure, and the abnormal path propagates along the direction with a larger coupling path strength.

[0086] Non - linear characteristics often exist in the form of discrete points, discontinuous regions or patches. It is necessary to fuse the aforementioned non - linear energy index and the coupling path analysis results to construct a spatially distributed non - linear manifestation probability field, which is used to describe the probability of each node being a non - linear excitation point, so as to conduct local strengthening detection and subsequent path reconstruction. Accumulate the non - linear energy index of each node and the coupling path strength it participates in; Define the node non - linear manifestation probability: where, is the coupling path strength between node i and adjacent node j in the cross - frequency band; Z is a normalization constant to ensure that ∑ i P i = 1; P i represents the relative probability of node i becoming a non - linear manifestation area.

[0087] Through the non - linear manifestation probability field, the potential non - linear response concentration areas can be visually located at the structural level, that is, the possible fatigue hot spots, crack initiation zones or critical stiffness jump zones on the stay cables.

[0088] In a cable-stayed bridge structure, stay cables are usually not independent entities but a highly coupled, structurally continuous, and mechanically interconnected system. A change in the state of a local stay cable (such as abnormal stress, nonlinear vibration, etc.) often conducts to other nodes through the cable force transmission path and the structural topological relationship. If the abnormality is judged only based on a single node, it is easy to ignore the transfer mechanism and coupling relationship between structures. For example, the abnormality of a node may be due to its local damage or an abnormal response caused by the redistribution of loads due to damage to adjacent nodes. Such situations require joint modeling and reasoning judgment based on the overall topological characteristics of the structure and the state propagation path.

[0089] Step 3, perform topological weakness reasoning on the stay cables, and the specific steps are as follows:

[0090] Abstract the stay cable structure and its relationship with sensing nodes into a graph structure. Each measurement point or monitoring unit is regarded as a node in the graph, and an edge is established between every two physically adjacent, structurally interconnected, or significantly coupled nodes in terms of response. Through the node set and edge set G=(V, E) of the graph, the physical topological relationship of the entire structure is reflected.

[0091] Define the graph node set V, where each node corresponds to a measurement point or a structural unit; define the graph edge set E. If there is a structural connection or energy coupling relationship between edges, an undirected edge is established. Define the feature vector of each node as: where P i is the probability of nonlinear manifestation of node i; λ i is the characteristic frequency position of node i in the frequency domain (such as the main peak offset frequency); JL i is the simplified index after projecting the coupling path strength of node i;

[0092] Each node simultaneously bears the physical state information extracted in the previous two steps (such as the probability of nonlinear acoustic vibration manifestation, coupling path response, frequency perturbation characteristics, etc.) as its initial feature vector. The significance of constructing this structural graph is to physically simulate the connection and force transmission relationship between node structures and simultaneously build a state propagation network channel at the data layer;

[0093] If the probability of nonlinear response manifestation of a certain monitoring node or the energy density reflected in its frequency response significantly increases, it indicates that there may be abnormal structural behavior in the stay cable area where the node is located. Conduct a topological analysis at the structural level of the stay cable area where the node is located, and obtain the structural state information generated during the topological analysis of the stay cable area where the abnormal structural behavior node is located. The structural state information includes structural coupling information and state propagation information;

[0094] The structural coupling information includes the structural coupling sparse index and is calibrated as SCSI, and the state propagation information includes the abnormal state conduction index and is calibrated as ASPI;

[0095] The structural coupling sparsity index is used to indicate the degree of coupling connection sparsity of the monitoring node of the cable-stayed tube in the structural topology diagram, reflecting the connection redundancy capacity, force path density and local topological vulnerability of a node in the structural network.

[0096] The structural coupling sparsity index measures whether the coupling path strength between a node and its adjacent nodes is strong enough, for example, whether the structural path has sufficient stiffness, energy coupling capability, or vibration response correlation:

[0097] In other words, the higher the structural coupling sparsity index, the thinner the connection of the node, the weaker the coupling, and the sparser the topology. This indicates that the node is the area in the structure most likely to be isolated due to local fracture, looseness, or failure.

[0098] The logic for obtaining the structural coupling sparsity index is as follows:

[0099] Get the three-dimensional space coordinates of nodes i and j: x i Coordinates and x j The parameters in the coordinates represent the coordinates of nodes i and j in the X-axis direction, Y-axis direction, and Z-axis direction, respectively. The Euclidean distance is calculated as the structural path length:

[0100] Collect the acoustic vibration response signal r for nodes i and j i (t), r j (t), calculate the frequency domain response value of nodes i and j: RX i (f) = F{r i (t)}, RX j (f) = F{r j (t)}, where F represents the Fourier transform operation, and the node structure coupling path strength is calculated based on the frequency domain response value: Where f1 is the lower bound of the frequency integration interval, indicating the starting frequency; f2 is the upper bound of the frequency integration interval, indicating the ending frequency; the structural coupling sparsity index is calculated based on the structural path length and the node-structure coupling path strength: Among them, N(i) is the set of adjacent nodes of node i.

[0101] It should be noted that the nodes collected are the nodes with abnormal structural behaviors in the inclined-stayed tube area.

[0102] The abnormal state conductivity index is used to indicate the linkage, conduction capability, and diffusion risk of the abnormal response of a monitoring node in the inclined-stayed tube during the evolution of the structural abnormality. It is used to identify key nodes that may not be the initial damage source but are located at the convergence of abnormal paths or the amplification hub of the response chain.

[0103] The abnormal state conduction index measures the abnormal response strength of the node itself and the degree of response coupling between the node and adjacent nodes. In other words, it indicates whether the node is in resonance or response channel with other abnormal nodes. By combining structural distance, graph topology position, and frequency domain linkage, its ability to serve as an abnormal hub is judged.

[0104] The abnormal state conductivity index is used to identify core nodes with high-risk abnormal propagation and assist in determining systemic linkage effects. For example, in a nonlinear acoustic vibration disturbance field, the abnormal state conductivity index can reveal whether a node is a strong point of resonant coupling and an energy accumulation point in the system linkage chain, thereby assessing the risk of abnormal state propagation. In areas with high abnormal state conductivity indexes, redundant monitoring or local reinforcement should be prioritized during structural operation and maintenance to prevent single-point damage from triggering a chain reaction of multiple abnormalities.

[0105] The logic for obtaining the abnormal state conduction index is as follows:

[0106] Get the frequency response signal XY of node i i (f) Set the frequency band set outside the main band as ZP NL , calculate the energy ratio of abnormal frequency band: Where ZP total is the total frequency band set, f represents the frequency;

[0107] Extract the second-order and third-order frequency band energy from the frequency response signal of node i, normalize the energy to bandwidth, and calculate the spectrum distortion energy density as follows:

[0108] Get the physical path length d between node i and adjacent node j ij , calculate the abnormal state conduction index, the calculation expression is: Where M(i) is the set of adjacent nodes of node i, PZ j ,MD j are the abnormal frequency band energy proportion and spectrum distortion energy density of node j respectively.

[0109] It should be noted that the frequency response signal can be the node response signal r obtained from step 2 i (t) Performing Fourier transform to obtain a frequency response signal; the frequency band set outside the main frequency band includes, for example, the second-order and third-order harmonic regions.

[0110] The structural coupling information and state propagation information are combined to generate the abnormality determination coefficient, that is, the obtained structural coupling sparse index and abnormal state conduction index are combined to generate the abnormality determination coefficient. The abnormality determination coefficient D x The expression is: Wherein, α and β are preset proportionality coefficients of the structural coupling sparse index and the abnormal state conduction index, and both α and β are greater than 0.

[0111] Specific methods for jointly generating the abnormal determination coefficient may involve multiple algorithms and models, which depend on the actual situation and application requirements. In this embodiment, the weighted summation method can be used to combine the structural coupling sparse index and the abnormal state conduction index to generate a comprehensive abnormal determination coefficient. This abnormal determination coefficient can be used as an input for comprehensively evaluating the connection redundancy degree and topological vulnerability of a node in the structural topology to determine the final abnormal node phenomenon.

[0112] It should be noted that the magnitude of the preset proportionality coefficient is a specific value obtained by quantifying each parameter. It is for the convenience of subsequent comparison. Regarding the magnitude of the coefficient, it depends on the amount of sample data and the initial setting of the corresponding preset proportionality coefficient for each group of sample data by those skilled in the art; it is not unique, as long as it does not affect the proportional relationship between the parameter and the quantified value. For example, the structural coupling sparse index is directly proportional to the abnormal determination coefficient. The structural coupling sparse index and the abnormal state conduction index are normalized to have the same dimension and range, which can be achieved by subtracting the mean from the original data and dividing by the standard deviation, or by mapping the data to the range of [0, 1].

[0113] The larger the structural coupling sparse index and the larger the abnormal state conduction index, the larger the jointly generated abnormal determination coefficient, indicating that the monitoring node has both characteristics of structural connection vulnerability and abnormal response linkage risk in the stay cable structure. On the one hand, a higher value of the structural coupling sparse index reflects that the number of connection paths between this node and adjacent nodes is small and the coupling path strength is weak, meaning that it is at the mechanical boundary or redundant break point in the structural topology diagram. Once damage occurs, it will be difficult to be compensated by other paths; on the other hand, a higher value of the abnormal state conduction index indicates that this node not only has a significant abnormal response itself, but also has a strong non - linear coupling relationship with multiple abnormal adjacent nodes, and has the potential ability to quickly conduct and even amplify the local abnormal state to the surrounding structural area. The area where the stay cable is located is very likely to become the starting source of structural quality degradation or the key propagation channel of response instability, and needs to be used as the key target area for priority inspection, precise monitoring, and local reinforcement;

[0114] The smaller the structural coupling sparse index and the smaller the abnormal state conduction index, the smaller the jointly generated abnormal determination coefficient, indicating that the monitoring node is in an area with a stable structural connection and normal response behavior in the stay cable system. The node is currently in an area with good structural health and negligible systematic risk, and can be regarded as a stable node or reference point in the stay cable structure;

[0115] Compare the generated anomaly determination coefficient with the pre-set anomaly determination threshold to generate a high-risk signal for the stay cable structure and a low-risk signal for the stay cable structure. This indicates that the stay cable structure has good connection density and mechanical coupling, and has a certain redundancy and fault tolerance ability. Even if the local force changes, it can maintain the stability of the structure through the multi-path shunt mechanism. The current node of the stay cable structure is in an area with good structural health and negligible systematic risk, and can be regarded as a stable node or reference point in the stay cable structure.

[0116] After obtaining the anomaly determination coefficient, compare the anomaly determination coefficient with the anomaly determination threshold.

[0117] If the anomaly determination coefficient is greater than or equal to the anomaly determination threshold, a high-risk signal for the stay cable structure is generated at this time, indicating that the node is in a highly sensitive risk state in the current structural health assessment. Specifically, the node has both a high structural coupling sparsity and a strong abnormal state conduction ability, indicating that its connection paths in the stay cable system are sparse and the mechanical coupling is weak. Once local damage or failure occurs, it is difficult to share the load through redundant paths and is extremely likely to form a mechanical breakpoint. At the same time, there is a significant non-linear response coupling between the node and the surrounding nodes, and its abnormal state has the potential to spread to adjacent structural areas and trigger a linkage risk; it is necessary to prompt that this area may be the source of structural quality deterioration, the core node of the response chain, or a potential fracture area, and early warning, inspection, and reinforcement intervention should be given priority.

[0118] Mark the node as a high-risk point and overlay and display it through visual methods such as heat maps and risk maps. At the same time, push a risk escalation signal to the maintenance platform / central control system, and record the trigger time, threshold overstep amplitude, and historical trends of relevant indicators; and include this point in the first round of tasks in the next inspection cycle, and use the ground team or drone / robot platform to conduct on-site re-inspection, and carry out local reinforcement intervention and preventive measures, etc.

[0119] If the anomaly determination coefficient is less than the anomaly determination threshold, a low-risk signal for the stay cable structure is generated, indicating that the node is currently in an area with good structural health. At this time, its structural coupling sparsity index is small, indicating that there are rich connection paths and close coupling between the node and the surrounding structural units, and it has good mechanical redundancy ability; at the same time, its abnormal state conduction index is also low, indicating that the node does not show significant non-linear abnormal characteristics in the frequency domain response, and the linkage response with adjacent nodes is weak, and no systematic linkage risk is formed. No key behaviors affecting structural safety have been found for this node, and it can be used as a stable reference point or a low-priority monitoring target, and there is no need to trigger emergency intervention measures, which helps to improve the overall inspection and resource scheduling efficiency.

[0120] It should be noted that the threshold information related in this embodiment is pre-set by professionals and will not be explained in detail here. In the embodiment, there are some cases where the English letters of some parameters are the same, but different meanings are explained when they are used, and they will not be explained one by one here.

[0121] The present invention collects strain data by arranging fiber optic sensors on the surface of the stay cables, constructs a continuous spatio-temporal strain field and extracts local gradient changes to identify micro-damage areas; combines multi-band acoustic excitation means to obtain the node response spectrum, extracts non-linear energy and frequency coupling characteristics, and calculates the non-linear manifestation probability to locate the potential non-linear response concentration area of the structure; further constructs a structure topology model to reflect the mechanical connection relationship between nodes and monitor abnormal responses; when abnormal structural behavior is detected, perform topological structure state analysis on the area where it is located, extract key information such as the coupling property and state propagation ability of the structure, and generate high and low risk signals according to the judgment result to achieve precise identification and intelligent control of the quality change of the stay cables, which can be used for early damage detection, non-linear response location and structural risk warning, and improve the perception accuracy of the operating state of the stay cables and the efficiency of maintenance decision-making.

[0122] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0123] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0124] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0125] In addition, in each embodiment of the present application, the functional modules can be integrated into one processing module, or each module exists physically alone, or two or more modules can be integrated into one module.

[0126] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims described above.

[0127] Finally: The above description is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A detection method for judging the quality of stay cables, characterized in that: The steps are as follows: Deploy fiber optic sensors on the surface of the stay cable pipe and collect signals, construct a continuous spatio-temporal strain field and extract local strain gradient information, identify strain mutation regions, perform time series modeling based on the strain field, and identify structural regions with micro-damage; Apply multi-band excitation signals to the stay cable pipe structure to determine the node response, obtain the frequency domain response spectrum and identify non-linear spectral features, extract non-linear energy indicators and cross-band coupling path strengths, calculate the non-linear manifestation probability to determine the abnormal excitation potential of the nodes, and locate the potential non-linear response concentration region of the structure; Construct a structure diagram based on the stay cable pipe structure and sensing nodes, simulate the connection force transmission relationship between node structures, and monitor whether there are abnormal structural behaviors in the nodes; If there are abnormal structural behaviors, obtain the structural state information generated during the topological analysis of the stay cable pipe area where the nodes with abnormal structural behaviors are located, analyze it, and perform regulation of the stay cable pipe according to different signals generated by the analysis results.

2. The detection method for judging the quality of stay cables according to claim 1, characterized in that: Deploy fiber optic sensors on the surface of the stay cable pipe and collect signals, construct a continuous spatio-temporal strain field and extract local strain gradient information, and identify strain mutation regions. The specific steps are as follows: Deploy fiber Bragg grating sensors on the surface of the stay cable pipe at regular intervals, and arrange sensing nodes on each optical fiber. Each node independently measures the axial strain response at its location; The fiber Bragg grating sensors collect and record the wavelength drift signals of all measurement points and perform strain conversion; Perform spatial strain interpolation and local gradient extraction on the collected signals. For each time point, use cubic spline interpolation as the interpolation function within the measurement point interval through the known strain values of the measurement points for interpolation; Perform interpolation for each time period during the monitoring period. Each time period corresponds to a strain spatial distribution curve. After combining all time periods, a two-dimensional strain field of the stay cable pipe during the entire monitoring period is obtained; Calculate the first-order spatial derivative of the interpolation function at the position of the nodes for each time point, and identify the possible regions of local concentrated damage to the structure based on the calculated first-order spatial derivative.

3. The detection method for judging the quality of stay cables according to claim 2, characterized in that: Perform time series modeling based on the strain field to identify structural regions with micro-damage. The specific steps include: Identify potential early damage regions of the structure based on the method of dynamic deviation modeling and statistical residual analysis; For each node x i , calculate the offset function between the average strain behavior during the current period and the reference state: ΔYB(x i , t k ) = YB(x i , t k ) - YB ref (x i ), where YB(x i , t k ) is the strain at time t k ; YB ref (x i ) is the reference value of the strain; the offset function represents the instantaneous offset value of the strain at the spatial position of the node relative to the healthy state at time t k ; For each node x i , construct the offset trajectory curve over the entire time period and define the damage criterion function: where is the time curvature of the strain offset; spatial gradient Set the abnormal intensity threshold QY. When SH i >QY, take the actual physical position collected by the sensor as the micro-damage area, and summarize the micro-damage areas as the potential micro-damage point set.

4. The detection method for judging the quality of stay cables according to claim 3, wherein: Apply multi-band excitation signals to the stay cable pipe structure to determine the node response, obtain the frequency domain response spectrum and identify non-linear spectral features, extract non-linear energy indicators and cross-band coupling path strengths. The specific steps are as follows: Implement continuous frequency sweep excitation or step frequency pulse excitation on the stay cable pipe structure; Analyze the second harmonic component, third harmonic, and non-integer multiple frequency components existing in the spectral structure for each response point, and define a non-linear energy index as follows: where Z harm represents all non-dominant frequency band sets, and R i (f) is the spectral response of the i-th node, and f is the frequency; Construct a cross - coupling tensor as the coupling path strength for the response spectra of all nodes: where O(·) represents taking the real part; R i (f p ) is the response of node i at frequency f p ; is the conjugate response of node j at frequency f p .

5. The detection method for judging the quality of stay cables according to claim 4, characterized in that: Calculate the non-linear manifestation probability to determine the abnormal excitation potential of the nodes and locate the potential non-linear response concentration region of the structure, including the following steps: Fuse the non-linear energy indicator and the coupling path analysis results, construct a spatially distributed non-linear manifestation probability field, and perform local enhancement detection and subsequent path reconstruction; Accumulate the non - linear energy index of each node with the strength of the participating coupling paths. The expression for estimating the probability of node non - linear manifestation is as follows: Where, is the coupling path strength between node i and adjacent node j in the cross - frequency band; Z is a normalization constant to ensure that ∑ i P i = 1; P i represents the relative probability that node i becomes a non - linear manifestation region, represents the non - linear energy index of node i; The non-linear manifestation probability field is used to describe the probability of each node being a non-linear excitation point; Locate the potential non-linear response concentration region at the structural level through the non-linear manifestation probability field.

6. The detection method for judging the quality of stay cables according to claim 5, characterized in that: A structural diagram is constructed based on the inclined-stayed tube structure and sensor nodes, the connection force transmission relationship between the node structures is simulated, and the nodes are monitored for abnormal structural behavior, including the following steps: The relationship between the inclined-stayed tube structure and the sensor nodes is abstracted into a graph structure. Each measuring point or monitoring unit is regarded as a node in the graph, and a graph edge is established between every two physically adjacent, structurally interconnected, or significantly coupled nodes. The graph node set is that each node corresponds to a measurement point or structural unit; If there is a structural connection or energy coupling relationship between the edges in the graph edge set, an undirected edge is established and the feature vector of each node is constructed. The feature vector of the node includes the nonlinear manifestation probability of the node, the characteristic frequency position of the node in the frequency domain, and the coupling path strength of the node; The nonlinear manifestation probability of the node is used to determine whether the monitored node has abnormal structural behavior.

7. The detection method for judging the quality of stay cables according to claim 6, characterized in that: If there is abnormal structural behavior, the structural state information generated by the topological analysis process of the inclined-stayed tube area where the abnormal structural behavior node is located is obtained, including the following steps: Obtaining structural state information generated during the topological analysis of the cable-stayed tube region where the constant structural behavior node is located. The structural state information includes structural coupling information and state propagation information. The structural coupling information includes the structural coupling sparsity index, and the state propagation information includes the abnormal state conduction index; The obtained structural coupling sparsity index and abnormal state conduction index are combined to generate an abnormality determination coefficient; The structural coupling sparse index and abnormal state conductivity index are positively correlated with the abnormality determination coefficient.

8. The detection method for judging the quality of stay cables according to claim 7, wherein: The analysis is then performed, and the inclined-stayed tube is regulated based on different signals generated by the analysis results, including the following steps: Compare the generated abnormality determination coefficient with the set abnormality determination threshold; If the abnormality determination coefficient is greater than or equal to the abnormality determination threshold, a high-risk signal for the inclined-stayed tube structure is generated, indicating that there is a linkage risk in the inclined-stayed tube structure, and inspection and reinforcement intervention of the corresponding inclined-stayed tube are carried out; If the abnormality determination coefficient is less than the abnormality determination threshold, a low-risk signal for the inclined-stayed tube structure is generated, and the inclined-stayed tube structure monitoring continues without triggering additional intervention measures.

Images