A method and system for predicting the residual bearing capacity of an RC hollow slab beam bridge

Abstract

The application discloses a kind of RC hollow slab beam bridge's residual bearing capacity prediction method and system, it is related to highway bridge detection technical field, including obtaining bridge prediction section, collecting the dynamic load data generated by vehicle driving at bridge prediction section some time, dynamic load data evaluation is carried out in preset bearing assessment model, and dynamic load influence value is obtained;Synchronous static load data, according to the weight of dynamic load influence value and static load data to bridge prediction section obtained by grey correlation analysis;Similarity comparison is carried out to bridge prediction section based on weight by historical bridge section data, select the historical bridge prediction section data that exceeds preset similarity threshold as sample, and construct training set;The residual bearing capacity data of bridge prediction section is obtained according to the training set training preset bearing capacity prediction model, the residual bearing capacity prediction of certain bridge section can be realized by the application, the representativeness of training set is enhanced, and the accuracy of bearing capacity prediction is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of highway bridge detection, and more particularly to a residual bearing capacity prediction method and system for RC hollow slab beam bridges. BACKGROUND

[0002] After decades of large-scale bridge construction, highway bridges have developed to a certain scale, among which concrete bridges and prestressed concrete bridges account for a large proportion. With the rise of big data, artificial intelligence and image recognition technology, the evaluation of bridges has also turned to intelligence. With the help of software algorithms, subjective factors can be avoided in the evaluation of the technical condition of bridges, and the evaluation of the technical condition of bridges is more scientific and systematic.

[0003] The purpose of intelligent evaluation of a certain section of old concrete bridge is to more accurately evaluate and predict the performance of the bridge, so that the bridge management and maintenance system is systematic, scientific and intelligent. Currently, the detection of the bearing capacity of the bridge mainly relies on comprehensive evaluation algorithms and load test methods, including static load test method and dynamic load test method. The static load test method is the most intuitive and effective method for evaluating the bearing capacity of the bridge. However, the load test method requires vehicle loading and measures the deformation of the bridge, which consumes a lot of manpower and material resources; at the same time, the test process needs to interrupt the traffic, which brings great inconvenience to the traffic.

[0004] Therefore, how to realize the residual bearing capacity prediction of a certain section of bridge is a problem to be solved by those skilled in the art. SUMMARY

[0005] Therefore, the present application provides a residual bearing capacity prediction method and system for RC hollow slab beam bridges, which can realize the residual bearing capacity prediction of a certain section of bridge.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0007] A residual bearing capacity prediction method for RC hollow slab beam bridges comprises:

[0008] According to the length of the RC hollow slab beam, the bridge prediction section is obtained by segmenting the bridge, the dynamic load data generated by the vehicle driving at a certain time of the bridge prediction section is collected by the sensor, and the dynamic load data is evaluated in the preset bearing evaluation model by the fuzzy comprehensive evaluation method to obtain the dynamic load influence value of the bridge based on the dynamic load data;

[0009] Synchronously collect the static load data of the bridge changes caused by the vehicle driving, and obtain the weight of the dynamic load influence value and the static load data for the bridge prediction section according to the gray correlation degree analysis;

[0010] Obtain historical bridge section data, compare the historical bridge section data with the bridge prediction section based on weight similarity, select the historical bridge prediction section data exceeding the preset similarity threshold as a sample, and construct a training set;

[0011] Train a preset bearing capacity prediction model according to the training set to obtain residual bearing capacity data of the bridge prediction section.

[0012] Preferably, the obtained dynamic load influence value of the bridge based on dynamic load data specifically comprises:

[0013] According to the type and position of the dynamic load data generated during vehicle driving, a bearing evaluation model is constructed to determine the influence degree of the dynamic load data generated during vehicle driving on the bridge prediction section;

[0014] The dynamic load data generated during vehicle driving is collected by a sensor, and a current influence level vector is obtained by processing the data collected by the sensor according to the bearing evaluation model through a fuzzy comprehensive evaluation method; and the dynamic load influence value of the bridge based on dynamic load data is determined according to the current influence level vector.

[0015] Preferably, the determination of the influence degree of the dynamic load data generated during vehicle driving on the bridge prediction section specifically comprises: establishing a bearing evaluation model of the type of dynamic load data generated during vehicle driving and the position of the bridge; comparing the influence degree of each dynamic load data in the bearing evaluation model to form a judgment matrix of each dynamic load data; and performing consistency test on the judgment matrix, and obtaining a weight vector of the dynamic load data after normalization processing of the judgment matrix passing the consistency test.

[0016] Preferably, the establishment of the bearing evaluation model of the type of dynamic load data generated during vehicle driving and the monitoring position specifically comprises: the first layer in the bearing evaluation model is the type of dynamic load data generated during vehicle driving, and the second layer is the different vehicle positions of the dynamic load data generated during vehicle driving.

[0017] Preferably, the obtained current influence level vector specifically comprises: the dynamic load data generated during vehicle driving is collected by a sensor, and the collected data is subjected to membership vectorization processing to obtain a fuzzy relationship matrix corresponding to the data at the current time; the current influence level vector is determined by fuzzy comprehensive evaluation according to the fuzzy relationship matrix and the weight vector.

[0018] Preferably, the determination of the dynamic load influence value and the weight of the static load data on the bridge prediction section according to the gray correlation degree analysis specifically comprises:

[0019] Select c historical bridge sections corresponding to the RC hollow slab beam, and for the maximum deflection δ max , the average deflection δavg minimum deflection δ min average strain ε avg average bending resistance M vag dynamic load influence value F, linearly mapped to the interval [0, 1], and six attribute factor vectors z1 to z6 of c x 1 are constructed;

[0020] Taking the average deflection as a reference vector, linearly mapped to the interval [0, 1], the same influence vector z7 of c x 1 is constructed, and the correlation coefficients of the six attribute factor vectors and the influence vector in the e-th component are calculated by using the grey correlation coefficient method:

[0021]

[0022] where λ f (f) represents the correlation coefficient of the f-th influence characteristic vector z f (e) and the power vector z7(e) in the e-th component, and p ∈ [0, 1] is a resolution coefficient;

[0023] Since the weights of dynamic load data of the same type of bridge are the same, the correlation degree r of z f and z7 is calculated as follows:

[0024]

[0025] The weight of each factor is solved, and the formula is as follows:

[0026]

[0027] where σ f is the weight of the f-th influence characteristic vector.

[0028] Preferably, the similarity comparison specifically includes:

[0029]

[0030] wherein, sample L = (l1, l2, …, l e ), O = (o1, o2, …, o e ), l a , o a are the normalized values of the influence factors, S(L, O) represents the similarity between the sample L and the sample O, w a indicates the weight of the a-th factor of the sample, and m1 and m2 represent the weighted average values of the p influence factors, respectively.

[0031] After classifying the samples, the average values of the dynamic load influence values and the static load data are calculated as the clustering center points:

[0032]

[0033] Wherein, the d type historical bridge section has r samples, m represents the number of influence factors, the similarity of the new sample and the clustering center point is calculated by formula, and the category with larger similarity is taken as the category of the new sample.

[0034] Preferably, the dynamic load data includes natural frequency, damping ratio and dynamic strain; and the static load data includes deflection, strain and bending capacity.

[0035] A residual carrying capacity prediction system of an RC hollow slab beam bridge comprises:

[0036] A dynamic load evaluation module obtains a bridge prediction section by segmenting the bridge according to the length of the RC hollow slab beam, collects dynamic load data generated by vehicle driving at a certain time of the bridge prediction section through a sensor, evaluates the dynamic load data in a preset carrying capacity evaluation model by a fuzzy comprehensive evaluation method, and obtains a dynamic load influence value of the bridge based on the dynamic load data.

[0037] A weight analysis module synchronously collects static load data of the bridge caused by vehicle driving, and obtains the weight of the dynamic load influence value and the static load data for the bridge prediction section according to a grey correlation degree analysis.

[0038] A training set construction module obtains historical bridge section data, compares the historical bridge section data with the bridge prediction section based on the weight, selects historical bridge prediction section data exceeding a preset similarity threshold as samples, and constructs a training set.

[0039] A carrying capacity prediction module trains a preset carrying capacity prediction model according to the training set, and obtains residual carrying capacity data of the bridge prediction section.

[0040] Compared with the prior art, the residual carrying capacity prediction method and system of the RC hollow slab beam bridge provided by the technical solution can realize real-time collection of dynamic load and static load data by using a sensor, so that the evaluation of the residual carrying capacity of the bridge is more accurate, the dynamic load data can be processed by a fuzzy comprehensive evaluation method, the influence of various factors can be comprehensively considered, the flexibility and adaptability of the evaluation model can be improved, the grey correlation degree analysis can be introduced to quantitatively analyze the mutual relationship between the dynamic load and the static load data, and a scientific basis can be provided for the determination of the weight; the similarity comparison based on the historical bridge section data can effectively utilize the existing data, enhance the representativeness of the training set, and improve the accuracy of the carrying capacity prediction; the carrying capacity prediction model can be self-optimized with the increase of data by continuously updating the training set and the evaluation model, and the reliability of the prediction can be improved; the present application is applicable to various types of RC hollow slab beam bridges, and the model parameters can be adjusted to adapt to different bridge structures and use conditions. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below only illustrate a part of the embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor based on the provided drawings.

[0042] Figure 1 The method steps provided by the present application are shown in the figure.

[0043] Figure 2 The structural schematic diagram provided by the present application is shown in the figure. DETAILED DESCRIPTION

[0044] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0045] The embodiments of the present application disclose a residual carrying capacity prediction method of RC hollow slab beam bridge, as shown in the figure, comprising: Figure 1

[0046] According to the length of the RC hollow slab beam, the bridge is segmented to obtain a bridge prediction section, the dynamic load data generated by the vehicle driving at a certain time of the bridge prediction section is collected by the sensor, and the dynamic load data is evaluated in the preset carrying evaluation model by the fuzzy comprehensive evaluation method, to obtain the dynamic load influence value of the bridge based on the dynamic load data.

[0047] Synchronously collect the static load data of the bridge changes caused by the vehicle driving, and obtain the weight of the dynamic load influence value and the static load data for the bridge prediction section according to the gray correlation degree analysis;

[0048] Obtain historical bridge section data, compare the historical bridge section data with the bridge prediction section based on the weight, select the historical bridge prediction section data exceeding the preset similarity threshold as a sample, and construct a training set.

[0049] According to the training set, the preset carrying capacity prediction model is trained to obtain the residual carrying capacity data of the bridge prediction section.

[0050] The collection of dynamic load data can be performed by a plurality of sensors arranged at random positions of the bridge section to be measured or sensors arranged on the vehicle, and then the data is collected through signal processing technology, which is prior art and will not be described in detail here.

[0051] ​In one specific embodiment, obtaining the dynamic load influence value of the bridge based on the dynamic load data specifically comprises:

[0052] According to the type and position of the dynamic load data generated during the vehicle driving process, a bearing evaluation model is constructed to determine the influence degree of the dynamic load data generated during the vehicle driving process on the prediction section of the bridge;

[0053] The dynamic load data generated during the vehicle driving process is collected by the sensor, and the current influence level vector is obtained by processing the data collected by the sensor according to the bearing evaluation model through the fuzzy comprehensive evaluation method; and the dynamic load influence value of the bridge based on the dynamic load data is determined according to the current influence level vector.

[0054] In one specific embodiment, determining the influence degree of the dynamic load data generated during the vehicle driving process on the prediction section of the bridge specifically comprises: establishing a bearing evaluation model of the type of dynamic load data generated during the vehicle driving process and the position of the bridge; comparing the influence degrees of each dynamic load data in the bearing evaluation model to form a judgment matrix of each dynamic load data; and performing consistency test on the judgment matrix, and obtaining the weight vector of the dynamic load data after normalizing the judgment matrix that passes the consistency test.

[0055] In one specific embodiment, the bearing evaluation model applies the analytic hierarchy process (AHP) to determine the weight of each dynamic load data, and the steps for determining the weight of each dynamic load data are as follows:

[0056] Step 1: Determine the bearing evaluation model;

[0057] Establish a bridge evaluation structure: the final goal of the present application is to determine the dynamic load influence value of the bridge, the first layer is each dynamic load data generated during the vehicle driving process on the bridge, and the second layer is the influence of the same dynamic load data on the bearing capacity of the bridge at different vehicle positions. The evaluation hierarchy structure established in this way is logically rigorous and conforms to the evaluation characteristics of the bridge. The monitoring data can be obtained by the sensors corresponding to the dynamic load data affecting the bearing capacity of the bridge at different vehicle positions on the bridge, and then the dynamic load influence value of the bridge affected by the dynamic load can be obtained from the monitoring data.

[0058] The main evaluation set is the dynamic load data affecting the bearing capacity of the bridge. The main evaluation set is mainly composed of natural frequency, damping ratio and dynamic strain. Therefore, the main evaluation set of the bridge is {natural frequency, damping ratio, dynamic strain}.

[0059] The secondary evaluation set is the influence of the same dynamic load data on the bearing capacity of the bridge at different vehicle positions, i.e. the secondary evaluation set of the bridge is {monitoring point 1, monitoring point 2, …, monitoring point q}.

[0060] Step 2: Construct the judgment matrix: on the basis of the bearing evaluation model, the relative importance between each other is compared to obtain the weight of each layer, and then the judgment matrix is formed.

[0061] Construct judgment matrix A i for

[0062]

[0063] In the formula: r ij This indicates the relative importance of the i-th influencing factor to the j-th factor. The relative importance is measured using a scale of 1 to 9, with the importance level divided into 9 levels, where level nine is the most important and the importance gradually decreases from level to level one, with level one being equally important.

[0064] The relative importance of the same dynamic load data at different locations to the state of the steel bridge is analyzed to obtain the judgment matrix for natural frequency, damping ratio, and dynamic strain. The weight vector is then determined. For a judgment matrix that meets the consistency requirement, the eigenvector corresponding to its largest eigenvalue, after normalization, becomes the weight vector for that influencing factor. The second-layer weight vector is determined, and the maximum eigenvalues ​​and corresponding eigenvectors of the judgment matrices A1, A2, and A3 for the natural frequency, damping ratio, and dynamic strain monitoring point locations are determined. After normalization, the weight vectors of the same dynamic load data at different vehicle locations are obtained.

[0065] The weight vectors of the first layer of measured dynamic load data are determined. Based on the definition of the relative importance judgment scale in the analytic hierarchy process (AHP), the influence of the natural frequency, damping ratio, and dynamic strain of the measured dynamic load data on the remaining bearing capacity of this bridge section is compared pairwise. A judgment matrix A is constructed, and the result is obtained through the formula... Perform a consistency check to obtain the weight vector.

[0066] In one specific embodiment, the load assessment model for establishing the types of dynamic load data generated during vehicle operation and the monitoring location specifically includes: the first layer of the load assessment model is the types of dynamic load data generated during vehicle operation, and the second layer is the different vehicle locations that generate dynamic load data during vehicle operation.

[0067] In one specific embodiment, the obtained current influence level vector specifically includes: collecting dynamic load data generated during vehicle operation through sensors, performing membership vectorization processing on the collected data to obtain the fuzzy relation matrix corresponding to the data at the current moment; and performing fuzzy comprehensive evaluation based on the fuzzy relation matrix and weight vector to determine the current influence level vector.

[0068] In one embodiment, according to the General Code for Design of Highway Bridges and Culverts, the bridge carrying capacity is subdivided into five discrete states with 1 as a unit, i.e. bridge carrying capacity Q = {1, 2, 3, 4, 5}. The membership degree of each dynamic load data is determined by using fuzzy statistics method, the dynamic load data affecting the residual carrying capacity of the bridge section is collected by sensors, the threshold value of the single dynamic load data is determined after processing, K intervals (the examples shown in the present application are divided into 5 intervals, i.e. K = 5) are divided for evaluation, and the evaluation results of each interval are represented by membership vector X v (v = 1, 2, 3, 4, 5).

[0069] The interval to which the data measured by the sensor at each monitoring point (vehicle position) belongs is determined, and the data is represented by the membership vector of the interval. Further, L data measured in a time period are set, and the number of times that the data respectively belong to each interval is n p (p = 1, 2, 3...), which is recorded as the weight of the interval. The membership vector of the factor in the time period after weighted averaging is:

[0070]

[0071] The dynamic load influence value is determined based on the two-level fuzzy comprehensive evaluation method, and the specific process is as follows:

[0072] The factor set U is determined:

[0073] The factor set U is divided into n factor subsets (first level) under the current state U of the bridge:

[0074] U = {u1, u2, u3... u n};

[0075] Each factor subset u i has k factors (second level):

[0076] u i = {u i1 ,u i2 …u in}。

[0077] u ij (i = 1, 2... n; j = 1, 2... k) represents the jth factor in the ith factor subset. The dynamic load data generated during the vehicle driving process of the bridge section is taken as the factor set U, and u i represents the dynamic load data affecting the residual carrying capacity of the bridge. The corresponding sensor is used to measure u ij represents the influence of each measured dynamic load data on the residual carrying capacity of the bridge section at different vehicle positions.

[0078] Determination of evaluation set: the evaluation level set refers to a set of various results that the evaluator can make on the evaluation factor set, reflecting the degree to which the evaluated influencing factor belongs to each possible result, and the evaluation of the residual bearing capacity of the bridge in the present application refers to the Highway Bridge and Culvert Maintenance Specification, and the evaluation level set is set as V={good, better, worse, poor, dangerous} 5 levels.

[0079] Since the two-layer model structure is established, two-level fuzzy comprehensive evaluation is adopted.

[0080] First-level fuzzy comprehensive evaluation: the first-level fuzzy comprehensive evaluation is to evaluate the second layer, and the present application is to perform fuzzy comprehensive evaluation on the vehicle position, i.e., the monitoring point of each measured dynamic load data.

[0081] Establishment of the second layer weight set and determination of the second layer fuzzy relation matrix: according to the monitoring data of each dynamic load data on the bridge at each vehicle position, the result is composed into a fuzzy relation matrix R after membership vectorization processing. i .

[0082] First-level fuzzy comprehensive evaluation: the second layer fuzzy relation matrix R is subjected to first-level comprehensive evaluation by using the weighted average model M(·,+) with the weight vector of the second layer. i

[0083]

[0084] B is normalized again B i = [b i1, b i2, b i3, b i4, b i5], i = 0, 1, 2,..., n. i

[0085] Second-level fuzzy comprehensive evaluation: the second-level fuzzy comprehensive evaluation is to evaluate the first layer, and the measured dynamic load data is subjected to fuzzy comprehensive evaluation.

[0086] Establishment of the first layer weight set: the method for establishing the first layer weight set is the same as that of the second layer;

[0087] Determination of the first layer fuzzy relation matrix: the first layer fuzzy relation matrix is the element in the second layer evaluation result B i , i.e.

[0088] R = (B 1, B 2, B 3, B 4) T ;

[0089] Second-level fuzzy comprehensive evaluation: the first layer fuzzy relation matrix R is subjected to second-level comprehensive evaluation by using the weighted average model M(·,+) with the weight vector of the first layer.

[0090]

[0091] ​​​The dynamic load influence value is as follows:

[0092] F = B x Q T ;

[0093] The B vector is a current influence level vector obtained after fuzzy comprehensive evaluation, and Q is a bridge bearing capacity part obtained according to the membership degree.

[0094] In one embodiment, the weight of the dynamic load influence value and the static load data for the bridge prediction section is obtained according to the grey correlation degree analysis, and specifically includes:

[0095] A c number of historical bridge sections corresponding to RC hollow slab beams are selected, and the maximum deflection δ max , the average deflection δ avg , the minimum deflection δ min , the average strain ε avg , the average bending resistance M vag , and the dynamic load influence value F of the RC hollow slab beams in the historical bridge sections are linearly mapped to the [0, 1] interval to construct six c x 1 attribute factor vectors z1 to z6, respectively.

[0096] The average deflection is taken as a reference vector, linearly mapped to the [0, 1] interval, and the same c x 1 influence vector z7 is constructed, and the grey correlation coefficient of the six attribute factor vectors and the influence vector in the e-th component is obtained by using the grey correlation coefficient method:

[0097]

[0098] Wherein, λ f (e) represents the correlation coefficient of the f-th influence characteristic vector z f (e) and the power vector z7(e) in the e-th component, and ρ ∈ [0, 1] is a resolution coefficient.

[0099] Since the weights of the dynamic load data of the same type of bridge are the same, the correlation degree r of z f and z7 is calculated as follows:

[0100]

[0101] Solving the weight of each factor, the formula is as follows:

[0102]

[0103] Wherein, σ f is the weight of the f-th influence characteristic vector.

[0104] In one embodiment, the similarity comparison specifically includes:

[0105]

[0106] wherein, sample L = (l1, l2, …, lp), O = (o1, o2, …, op) e ), O = (o1, o2, …, op) e ), O = (o1, o2, …, op) a ), O = (o1, o2, …, op) a are normalized values of influence factors, S(L, O) represents the similarity between sample L and sample O, w a represents the weight of the a th influence factor of the sample, m1, m2 respectively represent the weighted average value of p influence factors;

[0107] In one embodiment, the method of the present application can also be used to calculate samples of other types of bridges. After classifying the samples, the average values of the dynamic load influence values and the static load data are calculated as the cluster center points:

[0108]

[0109] wherein, the d th class of historical bridge section has r samples, m represents the number of influence factors, the similarity between the new sample and the cluster center point is calculated by the formula, and the class with greater similarity is taken as the class of the new sample.

[0110] In one embodiment, the dynamic load data includes natural frequency, damping ratio, dynamic strain; the static load data includes deflection, strain, and bending resistance.

[0111] In one embodiment, the load capacity prediction model adopts a RBF-BP combined neural network optimized by a genetic algorithm, the transfer function of the hidden layer is Gaussian function radbas in the first layer and tangent type transfer function tansig in the second layer, and the output layer adopts linear function purelin. The number of hidden layer nodes is determined by a construction method, i.e., a small number of hidden nodes is first set, and if the network output error after learning does not meet the set requirement, the number of nodes is gradually increased until the network error no longer obviously decreases. The fitness function is constructed as follows:

[0112]

[0113] wherein, is the fitness, h represents the number of output nodes, a s is the actual output of the s th node of the neural network, b s is the predicted output of the s th node.

[0114] A residual carrying capacity prediction system of an RC hollow slab bridge, as shown in Figure 2 , comprises:

[0115] The dynamic load evaluation module obtains a bridge prediction section according to the length of the RC hollow slab beam, collects dynamic load data generated by vehicle driving at a certain moment of the bridge prediction section through a sensor, and evaluates the dynamic load data in a preset bearing evaluation model through a fuzzy comprehensive evaluation method to obtain a dynamic load influence value of the bridge based on the dynamic load data.

[0116] The weight analysis module synchronously collects static load data of the bridge caused by vehicle driving, and obtains the weight of the dynamic load influence value and the static load data on the bridge prediction section according to a grey correlation degree analysis.

[0117] The training set construction module obtains historical bridge section data, compares the historical bridge section data with the bridge prediction section based on the weight, selects historical bridge prediction section data exceeding a preset similarity threshold as samples, and constructs a training set.

[0118] The bearing capacity prediction module trains a preset bearing capacity prediction model according to the training set to obtain residual bearing capacity data of the bridge prediction section.

[0119] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the device disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0120] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the remaining bearing capacity of an RC hollow slab girder bridge, characterized in that, include: Bridge segments are obtained by dividing the bridge into predicted segments based on the length of the RC hollow slab beam. Dynamic load data generated by vehicle movement at a certain moment in the predicted bridge segment are collected by sensors. The dynamic load data is evaluated in a preset load-bearing assessment model using a fuzzy comprehensive evaluation method to obtain the dynamic load influence value of the bridge based on the dynamic load data. The dynamic load data includes natural frequency, damping ratio, and dynamic strain. The first layer of the load-bearing assessment model represents the types of dynamic load data generated during vehicle movement, and the second layer represents the different vehicle positions that generate dynamic load data during vehicle movement. Static load data of bridge changes caused by vehicle movement are collected synchronously. The weights of dynamic load influence value and static load data for the predicted bridge segment are obtained based on grey relational analysis. The static load data includes deflection, strain, and flexural bearing capacity. The weights of dynamic load influence values ​​and static load data for the predicted bridge segment, derived from grey relational analysis, specifically include: choose For a corresponding historical bridge segment with RC hollow slab beams, the maximum deflection of the RC hollow slab beams within that historical bridge segment is... average deflection Minimum deflection mean strain Average flexural bearing capacity Dynamic load influence value Using a linear mapping to the [0,1] interval, six [variables] are constructed respectively. attribute factor vector arrive ; Using the average deflection as the reference vector, and employing a linear mapping to the [0,1] interval, the following is also constructed: Influence vector The grey relational coefficient method was used to calculate the vectors of the six attribute factors and their influence vectors in the 6th... Correlation coefficients of the components: ； in, Indicates the first Influence feature vectors With power vector In the The correlation coefficient of each component The resolution coefficient; Since the weights of dynamic load data are the same for bridges of the same type, and correlation Calculate using the following formula: ； The weights of each factor are calculated using the following formula: ； in, For the first The weights that influence the feature vector; Acquire historical bridge segment data, compare the historical bridge segment data with the predicted bridge segments based on weights, select historical bridge predicted segment data that exceeds a preset similarity threshold as samples, and construct a training set; The similarity comparison specifically includes: ； in, , ,sample , , , All values ​​are normalized impact factor values. Representative sample and samples Similarity between them Represents the sample's first The weights of each factor, , They represent The weighted average of the influencing factors; After classifying the samples, the mean values ​​of the dynamic load influence value and the static load data are calculated separately and used as the cluster centers: ； Among them, the Historical bridge sections include One sample, The number of influencing factors is used to calculate the similarity between the new sample and the cluster center using a formula, and the category with the higher similarity is taken as the category of the new sample. The preset bearing capacity prediction model is trained based on the training set to obtain the remaining bearing capacity data of the predicted bridge segment.

2. The method for predicting the remaining bearing capacity of an RC hollow slab girder bridge according to claim 1, characterized in that, The obtained dynamic load influence value of the bridge based on dynamic load data specifically includes: A load-bearing assessment model is constructed based on the types and locations of dynamic load data generated during vehicle operation to determine the degree of impact of dynamic load data generated during vehicle operation on the predicted bridge section. Data is collected by sensors to collect dynamic load data generated during vehicle operation. Based on the load assessment model, the data collected by the sensors is processed by the fuzzy comprehensive evaluation method to obtain the current influence level vector. The dynamic load influence value of the bridge based on the dynamic load data is determined according to the current influence level vector.

3. The method for predicting the remaining bearing capacity of an RC hollow slab girder bridge according to claim 2, characterized in that, The determination of the impact of dynamic load data generated during vehicle travel on the predicted bridge section specifically includes: establishing a load-bearing assessment model of the types of dynamic load data generated during vehicle travel and the bridge location; comparing the impact of each dynamic load data in the load-bearing assessment model to form a judgment matrix for each dynamic load data; performing a consistency check on the judgment matrix; and normalizing the judgment matrix that passes the consistency check to obtain the weight vector of the dynamic load data.

4. The method for predicting the remaining bearing capacity of an RC hollow slab girder bridge according to claim 3, characterized in that, The obtained current influence level vector specifically includes: collecting dynamic load data generated during vehicle operation through sensors, performing membership vectorization processing on the collected data to obtain the fuzzy relation matrix corresponding to the data at the current moment; and performing fuzzy comprehensive evaluation based on the fuzzy relation matrix and weight vector to determine the current influence level vector.

5. A system for predicting the remaining bearing capacity of an RC hollow slab girder bridge, using the method for predicting the remaining bearing capacity of an RC hollow slab girder bridge as described in any one of claims 1-4, characterized in that, include: The dynamic load assessment module obtains the predicted bridge segment by dividing the bridge into segments based on the length of the RC hollow slab beam. It collects dynamic load data generated by vehicle movement at a certain moment in the predicted bridge segment through sensors. The dynamic load data is evaluated in a preset load assessment model using the fuzzy comprehensive evaluation method to obtain the dynamic load impact value of the bridge based on the dynamic load data. The weight analysis module synchronously collects static load data of bridge changes caused by vehicle movement, and derives the weights of dynamic load influence value and static load data for the predicted bridge segment based on grey relational analysis. The training set construction module acquires historical bridge segment data, compares the historical bridge segment data with the predicted bridge segments based on weights, selects historical bridge predicted segment data that exceeds a preset similarity threshold as samples, and constructs a training set. The bearing capacity prediction module trains a preset bearing capacity prediction model based on the training set to obtain the remaining bearing capacity data of the predicted bridge segment.

Images