Steel-wood structure stability online evaluation method based on multi-source data combination
By using a multi-source data joint evaluation method, strain, interface slip, and acoustic emission signals of steel and wood are collected in real time. The stress ratio and stiffness degradation index of steel-wood joints are calculated, which solves the problem of multi-source data fusion in the stability assessment of steel-wood structures in the existing technology. It realizes full-coverage monitoring and multi-dimensional assessment of steel-wood structures, and improves the accuracy and reliability of the assessment results.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA RAILWAY INVESTMENT GRP CO LTD
- Filing Date
- 2026-05-22
- Publication Date
- 2026-06-19
AI Technical Summary
Existing technologies lack multi-source data fusion and multi-dimensional index coordination in the stability assessment of steel-wood structures, which makes it impossible to accurately capture the load distribution state, difficult to identify the causal relationship between stiffness degradation and acoustic emission events caused by interface slip, and the single index evaluation method cannot reflect the real degradation process under complex loads, which is prone to misjudging the structural stability.
By synchronously collecting steel strain, wood strain, interface relative slip, and acoustic emission signals in real time, the stress ratio of steel-wood joints, load distribution index, stiffness degradation index, and temporal coincidence index are calculated to comprehensively evaluate the overall stability of steel-wood structures. Dynamic online evaluation is carried out using multi-source data.
It achieves full-coverage monitoring of steel-wood structures, accurately quantifies load distribution and stiffness degradation, improves the accuracy and timeliness of abnormal damage identification, provides multi-dimensional structural stability assessment, and enhances the reliability and early warning capabilities of assessment results.
Smart Images
Figure CN122241487A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of online monitoring technology, specifically, it relates to an online stability assessment method for steel-wood structures based on multi-source data fusion. Background Technology
[0002] The rapid development of the Internet of Things, sensor technology, and artificial intelligence has provided technical support for the intelligent monitoring of building structures, and promoted the paradigm shift of structural health monitoring from periodic inspections to real-time online assessments.
[0003] Existing technologies for stability assessment of steel-wood structures have significant limitations. They typically treat steel and wood as independent components and test them separately, neglecting the inherent mechanical relationship between the two materials working together at the joints. This makes it difficult to accurately capture the load distribution between steel and wood and to quantify the stiffness degradation effect caused by interface slip. Furthermore, the use of asynchronous, discrete testing makes it difficult to establish accurate temporal logic relationships during subsequent correlation analysis. Especially when local damage occurs, the causal relationship between the suddenness of acoustic emission events and the gradual nature of interface slip is often severed. In addition, existing assessment methods rely heavily on single indicators and static thresholds. This single-dimensional evaluation approach cannot reflect the true degradation process of steel-wood structures under complex loads, easily leading to missed early damage or misjudgments of stability. Using static thresholds makes it difficult to distinguish between intermittent anomalies and continuous deterioration, and the judgment of structural stability lacks multi-dimensional cross-validation support.
[0004] To address the aforementioned issues, this invention proposes an online stability assessment method for steel-wood structures based on multi-source data fusion. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides an online stability assessment method for steel-wood structures based on multi-source data fusion, solving the problem of the lack of dynamic online assessment with multi-source data fusion and multi-dimensional index synergy in existing technologies.
[0006] The objective of this invention can be achieved through the following technical solutions: A method for online stability assessment of steel-wood structures based on multi-source data ensemble, the method comprising: Step 1: Locate all steel and wood nodes in the target steel-wood structural frame, and collect steel strain, wood strain, relative interface slip, and acoustic emission signals in real time based on a pre-deployed set of sensors. Step 2: Use the steel strain and wood strain of any steel-wood node to evaluate the stress ratio of the corresponding steel-wood node. Based on the numerical performance of the stress ratio of the steel-wood node, lock the load distribution state, summarize and evaluate the load distribution index of the target steel-wood structural frame, and execute the first stability index output. Step 3: In real time, lock the relative slip of the interface of any steel-wood node in the target steel-wood structure frame within the preset monitoring period and integrate it with the stress ratio of the corresponding steel-wood node, calculate the stiffness degradation index, and execute the second stability index output. Step 4: Lock the occurrence time of acoustic emission signals generated by any steel-wood node, perform time-series correlation verification with the interface relative slippage within the monitoring period, lock the time-series overlap index, execute the third stability index output, and evaluate the overall stability index of the target steel-wood structural frame together with the first and second stability indices, and output the stability level.
[0007] As a further embodiment of the present invention, the target steel-wood structure frame in step one is a target steel-wood structure frame that has been pre-constructed by the operator and is in a state of waiting to be tested. The target steel-wood structure frame is equipped with a set of sensors pre-installed by the operator. The sensors are all located at the steel-wood joints and include resistance strain gauges, vibrating wire strain gauges, draw wire displacement gauges, and resonant acoustic emission sensors.
[0008] As a further aspect of the present invention, the specific method for real-time synchronous acquisition of steel strain, wood strain, interface relative slip, and acoustic emission signals in step one is as follows: The statistical analysis includes all steel-wood nodes inside the target steel-wood structural frame, denoted as j; Randomly sort j steel-wood nodes, denoted as the steel-wood node sequence Q1, Q2, ..., Qj; Lock any steel-wood node Qi, where i is the counting index, 1≤i≤j; Determine the start time of monitoring, adjust the acquisition frequency of all sensors in the sensor set associated with the steel-wood node Qi to be uniform, and start data acquisition from the start time of monitoring; The steel strain, wood strain, relative interface slip, and acoustic emission signals at each time point are locked. The steel strain sequence S1, S2, ..., Sm, the wood strain sequence W1, W2, ..., Wm, the relative interface slip sequence R1, R2, ..., Rm, and the acoustic emission signal sequence E1, E2, ..., Em are constructed in chronological order, where m is the total number of time points, which increases with time.
[0009] As a further aspect of the present invention, the specific method for evaluating the steel-wood joint stress ratio of the corresponding steel-wood joint in step two is as follows: Lock the steel strain Sn and wood strain Wn of any steel-wood node Qi at any time n; The steel stress FSn and the wood stress FWn at the steel-wood node Qi at time n are calculated using the elastic modulus of steel and wood. The elastic modulus of steel and wood are based on the types of steel and wood and are considered as known values. The stress ratio SWn of any steel-wood node Qi at time n is calculated using FSWn=|FSn| / (|FWn|+Ɛ), where Ɛ is a very small positive number to prevent division by zero.
[0010] As a further aspect of the present invention, in step two, the specific method for locking the load distribution state based on the numerical representation of the stress ratio of the steel-wood joint is as follows: Extract the preset monitoring period T, determine the total number of moments within the monitoring period T, and denote it as o; At o times within the monitoring period T, the stress ratio of steel-wood node Qi is locked and arranged in chronological order, and is denoted as the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo; A two-dimensional coordinate system is constructed with the time line as the horizontal axis and the value of the steel-wood node stress ratio as the vertical axis. The o steel-wood node stress ratios in the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo are plotted in the two-dimensional coordinate system as data points aligned with the horizontal axis time line, resulting in o data points. The data points are then fitted to a curve and denoted as the steel-wood node stress ratio fluctuation curve V1. Obtain the operator's preset stress ratio range for steel-wood joints [FSW_min, FSW_max]; Lock the corresponding scales of FSW_min and FSW_max on the vertical axis, and construct straight lines perpendicular to the vertical axis and parallel to the horizontal axis through FSW_min and FSW_max respectively, denoted as line L_min and line L_max respectively; Continuously monitor the stress ratio fluctuation curve V1 of the steel-wood joint and the straight lines L_min and L_max. If there are p consecutive time periods when the stress ratio fluctuation curve V1 of the steel-wood joint is outside the straight lines L_min and L_max, then the load distribution state is determined to be an abnormal load distribution state. Conversely, it is a normal load distribution state.
[0011] As a further aspect of the present invention, in step two, the specific method for evaluating the load distribution index of the target steel-wood structural frame and executing the first stability index output is as follows: Within the monitoring period T, the load distribution status of j steel-wood nodes in the target steel-wood structure frame is locked, the total number of abnormal steel-wood nodes with abnormal load distribution status is counted and denoted as k, and k abnormal steel-wood nodes Q1', Q2', ..., Qk' are randomly arranged. The total number of steel-wood nodes with normal load distribution status is h=jk. For abnormal steel-wood node Qb', extract the maximum number of consecutive durations in abnormal load distribution state within the monitoring period T, denoted as t_abn_Qb', where 1≤b≤k; Obtain the force transmission level r1 and the number of connecting members r2 of the abnormal steel-wood node Qb' in the target steel-wood structural frame, and calculate the importance index dr_b of the abnormal steel-wood node Qb' using dr_b=ω1×r1+ω2×r2, where ω1 and ω2 are preset importance calculation weights; Similarly, the importance indices of all abnormal steel-wood nodes are determined, resulting in dr_1, dr_2, ..., dr_k; Using CV=k / j×((1-1 / j)∑ b∈[1,k] (dr_b×(t_abn_Qb' / o))) calculates the load distribution index CV of the target steel-wood frame, where CV is set to be always equal to 1 when there are no abnormal steel-wood nodes; (1-1 / j) is the scaling factor, making ∑ b∈[1,k] The effect of (dr_b×(t_abn_Qb' / o) on the load distribution index is matched with the total number of steel-wood nodes; The load distribution index CV is mapped to the [0,1] numerical range and output as the first stability index F1.
[0012] As a further aspect of the present invention, the specific method for calculating the stiffness degradation index and executing the second stability index output in step three is as follows: Lock the periodic interface relative slip sequence R1, R2, ..., Ro and the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo for any abnormal steel-wood node Qb' within the monitoring period T; Calculate the relative slip velocity of the interface of the abnormal steel-wood node Qb' within the monitoring period T: vR = (R1 + R2 + ... + Ro) / o; Similarly, the average steel-wood joint stress ratio vSW is calculated as follows: vSW = (FSW1 + FSW2 + ... + FSWo) / o for abnormal steel-wood joints Qb'. The node degradation coefficient Yb' of the abnormal steel-wood node Qb' is calculated using Yb'=vR / (vSW+Ɛ); Take the initial degradation coefficient Y0b' of the abnormal steel-wood node Qb' in the initial state, and calculate the stiffness degradation component Db'=Y0b' / Yb'; Similarly, the stiffness degradation index K = ∑(dr_b × Db') is calculated based on the importance index dr_b of any abnormal steel-wood node Qb', and K ∈ [0,1] interval. When there is no abnormal steel-wood node, K = 1. The stiffness degradation index K is output as the second stability index F2.
[0013] As a further aspect of the present invention, the specific method for locking the timing overlap index and executing the third stability index output in step four is as follows: Lock the periodic acoustic emission signal sequence E1, E2, ..., Eo and the periodic interface relative slip sequence R1, R2, ..., Ro of any abnormal steel-wood node Qb' within the monitoring period T; Obtain the preset acoustic emission event threshold E_th. When the amplitude of the acoustic emission signal is greater than E_th, mark the corresponding time as the time when the acoustic emission event occurs. Statistically count the times when all acoustic emission events occur within the monitoring period T to form the acoustic emission event time set AE_b. Obtain the preset slip event threshold R_th, and calculate the slip change ΔRn=|Rn-Rn+1| at any time n and its adjacent times; ΔR_n>R_th, mark time n as the time when the slip event occurs, and count the times when all slip events occur within the monitoring period T to form the slip event time set AR_b; The total number of times in the intersection of the acoustic emission event time set AE_b and the slip event time set AR_b is denoted as sum; The temporal overlap C_b of the abnormal steel-wood node Qb' is calculated using C_b=sum / max(|AE_b|,|AR_b|), where |AE_b| and |AR_b| are the total number of acoustic emission event moments and the total number of slip event moments, respectively. When both are 0, C_b=1. And so on, processing all abnormal steel-wood nodes; Using C=∑ b∈[1,k] (dr_b×C_b) calculates the time series overlap index C, maps the time series overlap index C to the interval [0,1], and outputs it as the third stability index F3.
[0014] As a further aspect of the present invention, in step four, the specific method for evaluating the overall stability index of the target steel-wood structure frame and outputting the stability level is as follows: Obtain the first stability index F1, the second stability index F2, and the third stability index F3; The overall stability index Φ is calculated using Φ=α×F1+β×F2+γ×F3, where α, β, and γ are preset weighting coefficients, α+β+γ=1, and α, β, and γ are all greater than 0. Map the overall stability index Φ to the interval [0,1]; Based on the preset overall stability index rating range and corresponding stability level, combined with the overall stability index Φ, the stability level of the target steel-wood structure frame within the monitoring period T is determined and continuously monitored.
[0015] The beneficial effects of this invention are: This invention achieves comprehensive monitoring of material stress, nodal load distribution, and interface damage evolution by simultaneously acquiring steel strain, wood strain, relative interface slip, and acoustic emission signals, thus solving the problem of one-sided assessment by traditional single indicators. It accurately quantifies the collaborative stress state of the composite structure using the steel-wood nodal stress ratio and load distribution index; it calculates the stiffness degradation index by fusing interface relative slip and stress ratio, enabling timely detection of nodal stiffness decay trends; and it improves the accuracy and timeliness of abnormal damage identification by temporally correlating acoustic emission events with interface slip. Finally, it enhances the reliability and early warning capability of the assessment results by comprehensively evaluating the overall stability through multi-source indices. This invention quantifies the load distribution state of steel-wood joints by precisely locking the strain of steel and wood and calculating the stress ratio of steel-wood joints using the elastic modulus. By constructing a stress ratio fluctuation curve of steel-wood joints and continuously monitoring it within a preset range, abnormal load distribution states can be accurately identified. Based on this, the importance index is calculated by considering the force transmission level of abnormal nodes, the number of connecting components, and the duration of the abnormality. Combined with the proportion of normal nodes for comprehensive evaluation, the final output load distribution index and first stability index not only reflect the stress state of the nodes themselves but also the structural weight and influence depth of the nodes in the overall frame. This achieves a systematic evaluation from local stress to overall stability, providing an objective and quantifiable basis for the safety monitoring and performance evaluation of steel-wood structural frames. This invention locks the interface relative slip sequence and stress ratio sequence of abnormal steel-wood nodes, calculates the average slip rate and average stress ratio respectively, and constructs a node degradation coefficient based on the ratio of the two. This couples slip behavior and stress state into a single index, eliminating the influence of dimensions and making the characterization of node degradation degree more direct and objective. It introduces an importance index to perform weighted summation of node stiffness degradation components, effectively integrating local damage into an overall stiffness degradation index, reflecting the differences in the contribution of different nodes in the structural system, improving the engineering rationality of the evaluation results, and realizing continuous quantification and intuitive output of structural stability. This invention constructs a complete evaluation chain from node anomaly identification to overall framework stability rating through a dual monitoring mechanism of acoustic emission signals and interface relative slippage. It introduces a temporal overlap index to quantify the correlation between acoustic emission events and slippage events in the time dimension, avoiding the risk of misjudgment based on a single parameter. On this basis, it generates an overall stability index by weighted fusion of the first, second, and third stability indices and maps it to a preset rating range to output a level. This realizes the comprehensive quantification of multi-dimensional monitoring data and the visualization of results, enhancing the dynamism and precision of structural health monitoring and providing objective and continuous data support for engineering maintenance decisions. Attached Figure Description
[0016] The invention will now be further described with reference to the accompanying drawings.
[0017] Figure 1 This is a flowchart illustrating the method described in this invention; Figure 2 This is a flowchart illustrating the method described in Embodiment 3 of the present invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] like Figure 1 As shown, this application provides an online stability assessment method for steel-wood structures based on multi-source data fusion; As an embodiment 1 of this application, it specifically includes: Step 1: Locate all steel and wood nodes in the target steel-wood structural frame, and collect steel strain, wood strain, relative interface slip, and acoustic emission signals in real time based on a pre-deployed set of sensors. Step 2: Use the steel strain and wood strain of any steel-wood node to evaluate the stress ratio of the corresponding steel-wood node. Based on the numerical performance of the stress ratio of the steel-wood node, lock the load distribution state, summarize and evaluate the load distribution index of the target steel-wood structural frame, and execute the first stability index output. Step 3: In real time, lock the relative slip of the interface of any steel-wood node in the target steel-wood structure frame within the preset monitoring period and integrate it with the stress ratio of the corresponding steel-wood node, calculate the stiffness degradation index, and execute the second stability index output. Step 4: Lock the occurrence time of acoustic emission signals generated by any steel-wood node, perform time-series correlation verification with the interface relative slippage within the monitoring period, lock the time-series overlap index, execute the third stability index output, and evaluate the overall stability index of the target steel-wood structural frame together with the first and second stability indices, and output the stability level.
[0020] Example 2
[0021] This embodiment provides a detailed explanation and description of step one based on embodiment 1, specifically including the following: First, it is necessary to locate the target steel-wood structural frame that the operators have pre-built, that is, the target steel-wood structural frame in the state of being awaiting inspection. It should be noted that being in the state of being awaiting inspection means that the target steel-wood structural frame has been completed or assembled but has not yet been put into use. Therefore, the initial state of each steel-wood node in the target steel-wood structural frame is known and traceable. However, the current state of the target steel-wood structural frame can be in any state. For example, it may be that 3 months have passed since the completion of construction and it is still in the state of being awaiting inspection.
[0022] The steel-wood joint refers to the contact interface between steel and wood, the connectors, and the adjacent area. The target steel-wood structure frame is internally configured with a sensor set pre-deployed by the operators. The sensors in the sensor set are all deployed at the steel-wood joints, and the specific locations are determined by the operators based on experience and actual conditions. The sensor set includes resistance strain gauges, vibrating wire strain gauges, draw wire displacement gauges, and resonant acoustic emission sensors. Resistance strain gauges are used for steel monitoring, vibrating wire strain gauges are used for wood monitoring, draw wire displacement gauges directly measure the relative slippage of the interface between steel and wood, and resonant acoustic emission sensors acquire acoustic emission signals at the steel-wood joints, such as elastic waves generated by interface debonding. The following describes a method for real-time synchronous acquisition of steel strain, wood strain, relative interface slip, and acoustic emission signals: First, before data acquisition, the acquisition frequency of all sensors in the sensor set needs to be uniformly processed to ensure that the acquired data is aligned with time. Next, the total number of all steel-wood nodes inside the target steel-wood structure frame is counted and denoted as j; Then, sort the j steel-wood nodes in a random order to obtain the steel-wood node sequence Q1, Q2, ..., Qj, in order to avoid systematic bias in subsequent data processing caused by the original physical arrangement order and improve the robustness of data processing; Next, lock any steel-wood node from the steel-wood node sequence Q1, Q2, ..., Qj, and denote it as steel-wood node Qi, where i is the counting index, 1≤i≤j. The following is an example of processing steel-wood node Qi. Other steel-wood nodes are processed in the same synchronous way as steel-wood node Qi. The monitoring start time is determined, which is generally the moment when the operator begins monitoring the target steel-wood structure frame. From this moment (monitoring start time), data acquisition is performed on the steel-wood nodes Qi using sensors in the sensor set. Starting from the moment of monitoring, determine the steel strain, wood strain, relative interface slip, and acoustic emission signal associated with the steel-wood node Qi at each moment; All collected steel strains are arranged in chronological order to obtain the steel strain sequence S1, S2, ..., Sm. Similarly, the wood strain, relative interface slip, and acoustic emission signals are processed to obtain the wood strain sequence W1, W2, ..., Wm, the relative interface slip sequence R1, R2, ..., Rm, and the acoustic emission signal sequence E1, E2, ..., Em associated with the steel-wood node Qi. Here, m is the total number of time points, which increases over time, and the end time is determined by the operator based on actual needs. Similarly, other steel-wood nodes are processed to determine the steel strain sequence, wood strain sequence, interface relative slip sequence, and acoustic emission signal sequence associated with other steel-wood nodes.
[0023] Example 3
[0024] This embodiment, based on Embodiment 1, further discloses a method for monitoring the load distribution state of a steel-wood structural frame and outputting a first stability index, such as... Figure 2 As shown, it specifically includes the following: In this embodiment, the content described in Embodiment 2 is continued, and the steel-wood node Qi is processed in the same way. Other steel-wood nodes are processed in the same and synchronous way as the steel-wood node Qi. Lock the steel strain Sn and wood strain Wn of the steel-wood node Qi at any time n; Based on the types of steel and wood, determine the elastic modulus of the steel and the elastic modulus of the wood in the steel-wood joint Qi. Thus, by combining the elastic modulus of steel and the elastic modulus of wood, the steel strain Sn and the wood strain Wn are used to calculate the steel stress FSn and the wood stress FWn associated with the steel-wood node Qi at time n. The stress ratio SWn of the steel-wood joint Qi at time n is calculated by using FSWn=|FSn| / (|FWn|+Ɛ), which is used to characterize the stress sharing relationship between steel and wood in the steel-wood joint Qi at time n. It should be noted that, The value is a very small positive number to avoid calculation errors caused by the denominator being 0. If the stress ratio of the steel-wood joint is greater than 1, it means that the steel bears more stress. If the stress ratio of the steel-wood joint is less than 1, it means that the wood bears more stress. If the stress ratio of the steel-wood joint is close to 1, it means that the two are relatively balanced in terms of stress. Next, the monitoring period T preset by the operator is obtained, and the total number of moments within the monitoring period T is further determined and denoted as o; At o times within the monitoring period T, the stress ratio of the steel-wood node Qi is locked in real time based on the steel stress and wood stress of the steel-wood node Qi. Finally, the o stress ratios of the steel-wood node corresponding to o times can be obtained. The o stress ratios of the steel-wood node are sorted in chronological order and denoted as the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo. Next, a two-dimensional coordinate system is constructed with the timeline as the horizontal axis and the value of the steel-wood node stress ratio as the vertical axis. The o steel-wood node stress ratios in the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo are regarded as o data points. The horizontal and vertical coordinates are aligned and plotted in the two-dimensional coordinate system to obtain o data points. The o data points are fitted with a curve to obtain a complete curve, which is denoted as the steel-wood node stress ratio fluctuation curve V1. Then obtain the stress ratio range [FSW_min,FSW_max] of the steel-wood joint preset by the operator; The stress ratio range [FSW_min, FSW_max] of the steel-wood joint is determined by the stress ratio distribution under normal service load using a finite element model, which is within the allowable deviation range for the coordinated work of steel and wood. Next, lock the corresponding scales of FSW_min and FSW_max on the vertical axis of the two-dimensional coordinate system, and construct straight lines perpendicular to the vertical axis and parallel to the horizontal axis through the corresponding scales of FSW_min and FSW_max on the vertical axis, denoted as line L_min and line L_max. Thus, the interval between line L_min and line L_max corresponds to the stress ratio interval of steel-wood joint [FSW_min, FSW_max]. Monitor the positional relationship between the stress ratio fluctuation curve V1 of the steel-wood node and the straight lines L_min and L_max at o time points. If there are p consecutive time points in which the stress ratio fluctuation curve V1 of the steel-wood node is outside the interval formed by the straight lines L_min and L_max (excluding the time points on the straight lines L_min and L_max), then the load distribution state of the steel-wood node Qi is determined to be an abnormal load distribution state. Otherwise, the load distribution state of the steel-wood node Qi is determined to be a normal load distribution state, and monitoring continues.
[0025] Following the above method, the same processing is performed on j steel-wood nodes within the target steel-wood structure frame within the monitoring period T to determine the load distribution status of each of the j steel-wood nodes. Lock all steel-wood nodes in abnormal load distribution states and mark them as abnormal steel-wood nodes. Count the total number of abnormal steel-wood nodes, denoted as k, and randomly arrange k abnormal steel-wood nodes, represented as: Q1', Q2', ..., Qk'. In this way, the total number of steel-wood nodes in normal load distribution states is h = jk.
[0026] From the k abnormal steel-wood nodes, lock any one abnormal steel-wood node Qb' and extract the maximum number of consecutive durations of abnormal load distribution state of the abnormal steel-wood node Qb' within the monitoring period T, denoted as t_abn_Qb', which represents the proportion of abnormal duration of the abnormal steel-wood node Qb' within the monitoring period T, where 1≤b≤k; Then extract the force transmission level r1 and the number of connecting members r2 of the abnormal steel-wood node Qb' within the target steel-wood structural frame, where the force transmission level r1 and the number of connecting members r2 are both considered as known values; The importance index dr_b of the steel-wood node Qi is calculated using the formula dr_b = ω1 × r1 + ω2 × r2. Here, ω1 and ω2 are preset importance calculation weights by the operator, and ω1 + ω2 = 1. Both ω1 and ω2 are greater than 0. The larger the value of the importance index dr_b, the more critical the steel-wood node Qi is in the force transmission path of the target steel-wood structural frame. This also indicates that anomalies in the steel-wood node Qi have a greater impact on the overall structure. This method can determine the importance index of all abnormal or non-abnormal steel-wood nodes. Here, only the importance index of abnormal steel-wood nodes is processed to obtain dr_1, dr_2, ..., dr_k corresponding to Q1', Q2', ..., Qk'.
[0027] Then, by using CV=k / j×((1-1 / j)∑ b∈[1,k] (dr_b×(t_abn_Qb' / o))) calculates the load distribution index CV of the target steel-wood structure frame, where CV is set to be always equal to 1 when there are no abnormal steel-wood nodes; It should be noted that (1-1 / j) is a scaling factor, the purpose of which is to make ∑ b∈[1,k] The effect of (dr_b×(t_abn_Qb' / o) on the load distribution index matches the total number of steel-wood nodes, and can simultaneously reflect the proportion of abnormal steel-wood nodes and the severity of abnormal steel-wood nodes, and is comparable in structures with different total numbers of steel-wood nodes. Finally, the load distribution index CV is mapped to the [0,1] value range and used as the first stability index F1 output.
[0028] Example 4
[0029] This embodiment further discloses a method for evaluating the stability of a steel-wood structural frame, based on embodiment 3, specifically including the following: Based on the content described in Example 3, the first stability index F1 of the target steel-wood structure frame within the evaluation period T is extracted for subsequent processing. First, obtain the periodic interface relative slip sequence R1, R2, ..., Ro and the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo for any abnormal steel-wood node Qb' within the monitoring period T; The relative slip rate vR of the interface of the abnormal steel-wood node Qb' during the monitoring period T is calculated by using vR=(R1+R2+...+Ro) / o, which characterizes the cumulative deformation rate of the abnormal steel-wood node Qb'. Then, by using vSW=(FSW1+FSW2+...+FSWo) / o, the average stress ratio vSW of the abnormal steel-wood node Qb' within the monitoring period T is calculated, which characterizes the average stress level of the abnormal steel-wood node Qb'. Combining the relative slip rate vR and the average steel-wood joint stress ratio vSW, the node degradation coefficient Yb' of the abnormal steel-wood joint Qb' is calculated using Yb'=vR / (vSW+Ɛ). The larger the relative slip rate and the smaller the average steel-wood joint stress ratio, the more significant the slip rate and the more severe the stiffness degradation of the abnormal steel-wood joint Qb'. Here, Ɛ is a very small positive number, such as 10 to the power of -6, as described in Example 3, to prevent division by zero. Next, the initial degradation coefficient Y0b' of the abnormal steel-wood node Qb' in the initial state is obtained, which corresponds to the initial state of each steel-wood node in the target steel-wood structure frame described in Example 2; The stiffness degradation component Db' of the abnormal steel-wood node Qb' is calculated based on Db'=Y0b' / Yb', which is the ratio of the initial degradation coefficient to the current degradation coefficient. The larger the value of the stiffness degradation component Db', the more severe the degradation of the abnormal steel-wood node Qb'. It should be added that if Yb' is less than Y0b', the stiffness degradation component Db' will be limited to 1 to avoid abnormal results of stiffness enhancement; By repeating the above steps, the stiffness degradation components of all abnormal steel-wood joints within the monitoring period T can be calculated. The importance index dr_b of the abnormal steel-wood node Qb' is then extracted. The stiffness degradation index K of the target steel-wood structural frame is calculated by using K=∑(dr_b×Db'), where ∑(dr_b×Db') represents the product of the importance index and stiffness degradation component of all abnormal steel-wood nodes and the summation, so as to reflect the differentiated contribution of different abnormal steel-wood nodes to the overall structural stability. The calculated stiffness degradation index K is then mapped to the interval [0,1]. When there are no abnormal steel-wood nodes within the monitoring period T, K is set to be equal to 1, and the stiffness degradation index K is executed as the second stability index F2.
[0030] Then extract the periodic acoustic emission signal sequence E1, E2, ..., Eo and the periodic interface relative slip sequence R1, R2, ..., Ro of the abnormal steel-wood node Qb' within the monitoring period T; Obtain the acoustic emission event threshold E_th preset by the operator, and then traverse the periodic acoustic emission signal sequence E1, E2, ..., Eo. When the amplitude of any acoustic emission signal is greater than the acoustic emission event threshold E_th, mark the corresponding time as the time when the acoustic emission event occurs, and count all the times when the acoustic emission event occurs within the monitoring period T, and summarize them to form the acoustic emission event time set AE_b associated with the abnormal steel-wood node Qb'. Then, obtain the operator's preset slip event threshold R_th, and calculate the slip change ΔRn=|Rn-Rn+1| at any time n and its adjacent times. When n is equal to time 0, then ΔRn=0. The calculated slip change ΔRn is compared with the slip event threshold R_th. If ΔR_n > R_th, it means that time n is the slip event occurrence time. In this way, all slip event occurrence times within the monitoring period T are counted and summarized to form the slip event time set AR_b. Determine the intersection of the acoustic emission event time set AE_b and the slip event time set AR_b in terms of time, and count the total number of times in the intersection, denoted as sum; The temporal overlap C_b of the abnormal steel-wood node Qb' is calculated by using C_b=sum / max(|AE_b|,|AR_b|), where |AE_b| and |AR_b| are the total number of acoustic emission event moments and the total number of slip event moments, respectively. If both |AE_b| and |AR_b| are 0, then C_b=1 is calibrated.
[0031] Following the above method, all abnormal steel-wood nodes are processed, the temporal overlap of all abnormal steel-wood nodes is determined, and then C=∑ b∈[1,k] (dr_b×C_b) calculates the temporal overlap index C of the target steel-wood structure frame within the monitoring period T, and maps the temporal overlap index C to the [0,1] interval as the third stability index F3 for execution output.
[0032] Based on the above, the calculated first stability index F1, second stability index F2, and third stability index F3 are extracted. The first stability index F1 reflects the overall stiffness degradation of the target steel-wood frame, the second stability index F2 reflects the degree of local damage to the target steel-wood frame, and the third stability index F3 reflects the damage synergy of the target steel-wood frame. Finally, the overall stability index Φ of the target steel-wood structure frame is calculated by using Φ=α×F1+β×F2+γ×F3, where α, β, and γ are preset weighting coefficients, α+β+γ=1, and α, β, and γ are all greater than 0. Then map the overall stability index Φ to the [0,1] interval; Based on the preset overall stability index rating range and corresponding stability level, combined with the overall stability index Φ, the stability level of the target steel-wood structure frame within the monitoring period T is determined, and the target steel-wood structure frame is continuously monitored. For example, an overall stability index Φ ≥ 0.85 indicates a stable level, an overall stability index Φ ∈ [0.6, 0.85] indicates a basically stable level, an overall stability index Φ ∈ [0.3, 0.6] indicates a critically stable level, and an overall stability index Φ < 0.3 indicates an unstable level.
[0033] All data in the formulas described above have been calculated with dimensions removed. Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.
[0034] The above description is merely an example and illustration of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.
[0035] It should be stated that all user data collected in this application was collected with the user's consent and authorization. Furthermore, the uses of user data are legal and compliant, and the use and processing of user data comply with the relevant laws, regulations, and standards of the relevant regions.
Claims
1. A method for online stability assessment of steel-wood structures based on multi-source data fusion, characterized in that, The method includes: Step 1: Locate all steel and wood nodes in the target steel-wood structural frame, and collect steel strain, wood strain, relative interface slip, and acoustic emission signals in real time based on a pre-deployed set of sensors. Step 2: Use the steel strain and wood strain of any steel-wood node to evaluate the stress ratio of the corresponding steel-wood node. Based on the numerical performance of the stress ratio of the steel-wood node, lock the load distribution state, summarize and evaluate the load distribution index of the target steel-wood structural frame, and execute the first stability index output. Step 3: In real time, lock the relative slip of the interface of any steel-wood node in the target steel-wood structure frame within the preset monitoring period and integrate it with the stress ratio of the corresponding steel-wood node, calculate the stiffness degradation index, and execute the second stability index output. Step 4: Lock the occurrence time of acoustic emission signals generated by any steel-wood node, perform time-series correlation verification with the interface relative slippage within the monitoring period, lock the time-series overlap index, execute the third stability index output, and evaluate the overall stability index of the target steel-wood structural frame together with the first and second stability indices, and output the stability level.
2. The method according to claim 1, characterized in that, The target steel-wood structure frame in step one is a target steel-wood structure frame that has been pre-constructed by the operator and is in a state of waiting for inspection. The target steel-wood structure frame is equipped with a set of sensors pre-installed by the operator. The sensors are all located at the steel-wood joints and include resistance strain gauges, vibrating wire strain gauges, draw wire displacement gauges, and resonant acoustic emission sensors.
3. The method according to claim 2, characterized in that, In step one, the specific method for real-time synchronous acquisition of steel strain, wood strain, relative interface slip, and acoustic emission signals is as follows: The statistical analysis includes all steel-wood nodes inside the target steel-wood structural frame, denoted as j; Randomly sort j steel-wood nodes, denoted as the steel-wood node sequence Q1, Q2, ..., Qj; Lock any steel-wood node Qi, where i is the counting index, 1≤i≤j; Determine the start time of monitoring, adjust the acquisition frequency of all sensors in the sensor set associated with the steel-wood node Qi to be uniform, and start data acquisition from the start time of monitoring; The steel strain, wood strain, relative interface slip, and acoustic emission signals at each time point are locked. The steel strain sequence S1, S2, ..., Sm, the wood strain sequence W1, W2, ..., Wm, the relative interface slip sequence R1, R2, ..., Rm, and the acoustic emission signal sequence E1, E2, ..., Em are constructed in chronological order, where m is the total number of time points, which increases with time.
4. The method according to claim 3, characterized in that, In step two, the specific method for evaluating the stress ratio of the corresponding steel-wood joint is as follows: Lock the steel strain Sn and wood strain Wn at any time n for any steel-wood node Qi; The steel stress FSn and the wood stress FWn at the steel-wood node Qi at time n are calculated using the elastic modulus of steel and wood. The elastic modulus of steel and wood are based on the types of steel and wood and are considered as known values. The stress ratio SWn of any steel-wood node Qi at time n is calculated using FSWn=|FSn| / (|FWn|+Ɛ), where Ɛ is a very small positive number to prevent division by zero.
5. The method according to claim 4, characterized in that, In step two, the specific method for locking the load distribution state based on the numerical representation of the stress ratio of steel-wood joints is as follows: Extract the preset monitoring period T, determine the total number of moments within the monitoring period T, and denote it as o; At o times within the monitoring period T, the stress ratio of steel-wood node Qi is locked and arranged in chronological order, and is denoted as the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo; A two-dimensional coordinate system is constructed with the time line as the horizontal axis and the value of the steel-wood node stress ratio as the vertical axis. The o steel-wood node stress ratios in the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo are plotted in the two-dimensional coordinate system as data points aligned with the horizontal axis time line, resulting in o data points. The data points are then fitted to a curve and denoted as the steel-wood node stress ratio fluctuation curve V1. Obtain the operator's preset stress ratio range for steel-wood joints [FSW_min, FSW_max]; Lock the corresponding scales of FSW_min and FSW_max on the vertical axis, and construct straight lines perpendicular to the vertical axis and parallel to the horizontal axis through FSW_min and FSW_max respectively, denoted as line L_min and line L_max respectively; Continuously monitor the stress ratio fluctuation curve V1 of the steel-wood joint and the straight lines L_min and L_max. If there are p consecutive time periods when the stress ratio fluctuation curve V1 of the steel-wood joint is outside the straight lines L_min and L_max, then the load distribution state is determined to be an abnormal load distribution state. Conversely, it is a normal load distribution state.
6. The method according to claim 5, characterized in that, In step two, the load distribution index of the target steel-wood structural frame is evaluated, and the specific method for executing the first stability index output is as follows: Within the monitoring period T, the load distribution status of j steel-wood nodes in the target steel-wood structure frame is locked, the total number of abnormal steel-wood nodes with abnormal load distribution status is counted and denoted as k, and k abnormal steel-wood nodes Q1', Q2', ..., Qk' are randomly arranged. The total number of steel-wood nodes with normal load distribution status is h=jk. For abnormal steel-wood node Qb', extract the maximum number of consecutive durations in abnormal load distribution state within the monitoring period T, denoted as t_abn_Qb', where 1≤b≤k; Obtain the force transmission level r1 and the number of connecting members r2 of the abnormal steel-wood node Qb' in the target steel-wood structural frame, and calculate the importance index dr_b of the abnormal steel-wood node Qb' using dr_b=ω1×r1+ω2×r2, where ω1 and ω2 are preset importance calculation weights; Similarly, the importance indices of all abnormal steel-wood nodes are determined, resulting in dr_1, dr_2, ..., dr_k; The load distribution index CV of the target steel-wood structure frame is calculated by CV = k / j x ((1-1 / j)∑ b∈[1,k] (dr_b x (t_abn_Qb' / o)) where, when there is no abnormal steel-wood joint, CV is set to be equal to 1. (1-1 / j) is a scaling factor such that∑ b∈[1,k] The influence of the load distribution index (dr_b x (t_abn_Qb' / o) matches the total number of steel-wood joints. The load distribution index CV is mapped to the [0,1] numerical range and output as the first stability index F1.
7. The method according to claim 6, characterized in that, In step three, the specific method for calculating the stiffness degradation index and executing the second stability index output is as follows: Lock the periodic interface relative slip sequence R1, R2, ..., Ro and the periodic steel-wood node stress ratio sequence FSW1, FSW2, ..., FSWo for any abnormal steel-wood node Qb' within the monitoring period T; Calculate the relative slip velocity of the interface of the abnormal steel-wood node Qb' within the monitoring period T: vR = (R1 + R2 + ... + Ro) / o; Similarly, the average steel-wood joint stress ratio vSW is calculated as follows: vSW = (FSW1 + FSW2 + ... + FSWo) / o for abnormal steel-wood joints Qb'. The node degradation coefficient Yb' of the abnormal steel-wood node Qb' is calculated using Yb'=vR / (vSW+Ɛ); Take the initial degradation coefficient Y0b' of the abnormal steel-wood node Qb' in the initial state, and calculate the stiffness degradation component Db'=Y0b' / Yb'; Similarly, the stiffness degradation index K = ∑(dr_b × Db') is calculated based on the importance index dr_b of any abnormal steel-wood node Qb', and K ∈ [0,1] interval. When there is no abnormal steel-wood node, K = 1. The stiffness degradation index K is output as the second stability index F2.
8. The method according to claim 7, characterized in that, In step four, the specific method for locking the temporal overlap index and executing the third stability index output is as follows: Lock the periodic acoustic emission signal sequence E1, E2, ..., Eo and the periodic interface relative slip sequence R1, R2, ..., Ro of any abnormal steel-wood node Qb' within the monitoring period T; Obtain the preset acoustic emission event threshold E_th. When the amplitude of the acoustic emission signal is greater than E_th, mark the corresponding time as the time when the acoustic emission event occurs. Statistically count the times when all acoustic emission events occur within the monitoring period T to form the acoustic emission event time set AE_b. Obtain the preset slip event threshold R_th, and calculate the slip change ΔRn=|Rn-Rn+1| at any time n and its adjacent times; ΔR_n>R_th, mark time n as the time when the slip event occurs, and count the times when all slip events occur within the monitoring period T to form the slip event time set AR_b; The total number of times in the intersection of the acoustic emission event time set AE_b and the slip event time set AR_b is denoted as sum; The temporal overlap C_b of the abnormal steel-wood node Qb' is calculated using C_b=sum / max(|AE_b|,|AR_b|), where |AE_b| and |AR_b| are the total number of acoustic emission event moments and the total number of slip event moments, respectively. When both are 0, C_b=1. And so on, processing all abnormal steel-wood nodes; C = ∑ b∈[1,k] The (dr_b x C_b) calculation timing coincidence index C is mapped to the [0, 1] interval, and output as the third stability index F3.
9. The method according to claim 8, characterized in that, In step four, the specific method for evaluating the overall stability index of the target steel-wood structure frame and outputting the stability level is as follows: Obtain the first stability index F1, the second stability index F2, and the third stability index F3; The overall stability index Φ is calculated using Φ=α×F1+β×F2+γ×F3, where α, β, and γ are preset weighting coefficients, α+β+γ=1, and α, β, and γ are all greater than 0. Map the overall stability index Φ to the interval [0,1]; Based on the preset overall stability index rating range and corresponding stability level, combined with the overall stability index Φ, the stability level of the target steel-wood structure frame within the monitoring period T is determined and continuously monitored.