Method for controlling vibration disturbance at joint of new bridge and old bridge

Abstract

The invention relates to the technical field of bridge structure engineering, in particular to a new and old bridge joint vibration disturbance control method, which comprises the following steps: S1, arranging a sensor at a new and old bridge joint, and collecting vibration data of the part; according to the method, a multi-source information identification model is constructed by fusing a physical mechanism and deep learning, robust identification of a vibration source and dynamic reconstruction of a transmission path are realized in combination with vibration transmission entropy and Granger causal test, an accurate basis is provided for subsequent vibration reduction scheme design, and an adaptive scheme is screened based on vibration characteristics, so that the accuracy of vibration reduction is improved. A four-dimensional coupling model is constructed through finite element simulation, an improved genetic algorithm is adopted to solve a Pareto optimal solution under the constraint of cooperative work performance of new and old bridges, parameters are selected according to engineering priorities, and cooperative optimization of the vibration reduction effect, structural safety, cost control and durability is achieved. Therefore, the accuracy and engineering applicability of vibration disturbance control at the joint of the new bridge and the old bridge are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of bridge structure engineering, in particular to a vibration disturbance control method for new and old bridge connection. BACKGROUND

[0002] The new and old bridge connection refers to the key part of the structural connection between the newly built bridge and the built (existing) bridge, usually including the connection joint, the transition section and the peripheral associated structure (such as the support, the expansion device, the connecting steel plate, etc.), which is an important interface for the new and old bridge to realize load transfer, deformation coordination and overall cooperative work. Due to the significant differences in construction years, structural forms, material properties (such as concrete strength, steel mechanical properties) and service conditions (new bridge without degradation, old bridge may have damage or stiffness attenuation), this part is prone to stiffness mutation and stress concentration, becoming a sensitive area of vibration disturbance - external excitations such as vehicle load and environmental wind vibration are easy to produce complex vibration transmission and energy accumulation at this point, which may cause fatigue damage of the connecting component, failure of the expansion joint and other problems, threatening the safety and durability of the bridge structure. In the prior art, the traditional vibration disturbance control method for the new and old bridge connection has the problems of low recognition accuracy, lack of multi-objective collaborative optimization of the scheme, and difficulty in considering the vibration reduction effect and the actual engineering demand.

[0003] Based on this, the present application provides a vibration disturbance control method for new and old bridge connection to solve the above technical problems. SUMMARY

[0004] The purpose of the present application is to provide a vibration disturbance control method for new and old bridge connection, which integrates physical mechanism and deep learning to construct a multi-source information recognition model, and realizes robust identification of vibration source and dynamic reconstruction of transmission path by combining vibration transmission entropy and Granger causality test, providing accurate basis for subsequent vibration reduction scheme design, and selecting adaptive scheme based on vibration characteristics, constructing four-dimensional coupled model through finite element simulation, and solving Pareto optimal solution under the constraint of new and old bridge cooperative working performance by using improved genetic algorithm and selecting parameters according to engineering priority, realizing the collaborative optimization of vibration reduction effect, structural safety, cost control and durability, thereby significantly improving the accuracy and engineering applicability of the vibration disturbance control for new and old bridge connection.

[0005] To achieve the above purpose, the present application provides the following technical scheme:

[0006] The present application provides a vibration disturbance control method for new and old bridge connection, comprising the following steps:

[0007] S1: arranging sensors at the new and old bridge connection to collect vibration data of the part;

[0008] S2: Based on a multi-source information identification model that integrates physical mechanisms and deep belief networks, combined with vibration transmission entropy analysis, robust identification of vibration sources at the connection between old and new bridges in complex environments and dynamic reconstruction of transmission paths are achieved.

[0009] S3: Select a tuned mass damper or a local stiffness enhancement scheme based on vibration characteristics. Construct a four-dimensional coupled model of "vibration reduction rate - structural stress - construction cost - durability" through finite element simulation. Based on the performance constraints of the collaborative work of the new and old bridges, use an improved genetic algorithm to solve the multi-objective Pareto optimal solution. Finally, select the scheme parameters according to the project priority.

[0010] S4: After construction according to the optimized plan, the vibration data is dynamically monitored and compared with the standard limits. If the standard is not met, secondary optimization is carried out by adjusting the damper parameters or reinforcing the structure.

[0011] The sensors in S1 include a triaxial accelerometer, a displacement sensor, and a strain gauge, which are symmetrically arranged on both sides of the connection between the old and new bridges, with a spacing of 0.5 to 2m.

[0012] The multi-source information identification model in S2, which integrates physical mechanisms and deep belief networks, combined with vibration transmission entropy analysis, robustly identifies vibration sources at the connection between old and new bridges in complex environments and dynamically reconstructs their transmission paths. The specific steps are as follows:

[0013] S2.1: Integrate vibration sensor data, environmental monitoring data, and bridge design parameters to construct a multi-dimensional input matrix;

[0014] S2.2: Based on Timoshenko beam theory, a transfer matrix model is established to quantify the propagation characteristics of vibration waves between old and new bridges;

[0015] S2.3: Automatically learns the time-frequency characteristics of vibration signals through a 5-layer deep belief network;

[0016] S2.4: Calculate the vibration transfer entropy between different measuring points, and reconstruct the vibration energy transfer path by combining Granger causality test;

[0017] S2.5: The DS theory is used to fuse the physical model output with the deep learning results to generate a vibration source identification report with confidence.

[0018] In step S2.4, the vibration transfer entropy value between different measuring points is calculated, and the vibration energy transfer path is reconstructed using the Granger causality test. The specific steps are as follows:

[0019] S2.4.1: Preprocess the vibration signals at different measuring points to remove noise interference and standardize them to obtain a stationary time series;

[0020] S2.4.2: Based on the pre-processed time series, calculate the vibration transfer entropy value between any two measuring points, the formula is:

[0021] TE(X→Y) = H(Y|{Y(t-1),..., Y(t-k)})

[0022] -H(Y|{Y(t-1),..., Y(t-k), X(t-1),..., X(t-m)})

[0023] Where H is the information entropy, k and m are the delay orders of Y and X respectively;

[0024] S2.4.3: Perform Granger causality test on the vibration signals of the two measuring points, calculate the F statistic by constructing an autoregressive model, and determine whether X is the Granger cause of Y;

[0025] S2.4.4: Fuse the vibration transfer entropy value and the Granger causality test result, mark the paths that meet the conditions at the same time as the dominant vibration energy transfer paths, and complete the path reconstruction.

[0026] In the construction of autoregressive model to calculate F statistic in S2.4.3, let the time series of vibration signals of two measuring points be X t , Y t , the specific formula for calculating F statistic is as follows:

[0027] Unconstrained autoregressive model:

[0028]

[0029] Constrained autoregressive model:

[0030]

[0031] Where α i , β i are model coefficients, ∈ 1t , ∈ 2t are residual terms, and p is the lag order;

[0032] The formula for calculating F statistic is:

[0033]

[0034] Where SSR r is the residual sum of squares of the constrained model, SSR ur is the residual sum of squares of the unconstrained model, and n is the sample size of the vibration signal.

[0035] In S2.5, the D-S theory is used to fuse the physical model output and the deep learning result to generate a vibration source identification report with confidence, the specific steps are as follows:

[0036] S2.5.1: Basic probability assignment:

[0037] ①Physical model output result m phy (A) Generated by residual sum of squares normalization:

[0038]

[0039] where y obs is the measured vibration data of the sensor, and y phy is the predicted value of the physical model;

[0040] ②Deep learning result m dnn (A) Generated by Softmax function:

[0041]

[0042] where, is the original score of the output layer of the deep belief network for hypothesis A, is the value of the i th output node of the deep belief network, and K is the total number of vibration source categories;

[0043] S2.5.2: Evidence fusion: Calculate the fused confidence using the D-S combination rule:

[0044]

[0045] S2.5.3: Conflict processing: When the conflict factor is greater than 0.5, start the manual review process;

[0046] S2.5.4: Report generation: Output the confidence interval [Bel(A), Pl(A)] of each vibration source hypothesis, where:

[0047] In S3, according to the vibration characteristics, the tuned mass damper or local stiffness enhancement scheme is selected, a four-dimensional coupling model of "vibration reduction rate-structure stress-construction cost-durability" is constructed by finite element simulation, based on the constraint of the cooperative working performance of new and old bridges, the improved genetic algorithm is used to solve the multi-objective Pareto optimal solution, and finally the scheme parameters are selected according to the engineering priority, the specific steps are as follows:

[0048] S3.1: According to the vibration characteristics identified in S2, preliminarily screen the applicable vibration reduction measures, and judge whether to use the tuned mass damper or the local stiffness enhancement scheme;

[0049] S3.2: Quantify the vibration reduction rate, structure stress level, construction cost and durability indicators under different scheme parameters through finite element simulation, and unify and normalize the processing to construct a multi-objective comprehensive evaluation system;

[0050] S3.3: Under the constraint of the new and old bridge collaborative performance, the improved genetic algorithm is used for global search to solve the Pareto optimal solution;

[0051] S3.4: Combined with the actual demand of the project, the final parameter combination most suitable for the current engineering priority is selected from the Pareto solution set.

[0052] In S3.3, under the constraint of the new and old bridge collaborative performance, the improved genetic algorithm is used for global search to solve the Pareto optimal solution, and the specific steps are as follows:

[0053] S3.3.1: Constraint processing: dynamic penalty function is used to process the new and old bridge collaborative constraint:

[0054]

[0055] Wherein, g j (x) is the constraint function, "j=1,2,3,4", g1(x) is the material modulus difference constraint ≤15%, g2(x) is the expansion joint displacement constraint ≤50mm, g3(x) is the support reaction force balance constraint ≤10%, g4(x) is the inherent frequency matching constraint ≤5%;

[0056] S3.3.2: Algorithm improvement:

[0057] ① Cross probability self-adaptive adjustment:

[0058]

[0059] Wherein, t is the current generation number, T is the total generation number;

[0060] ② The elite reservation ratio is fixed at 15% of the population size;

[0061] S3.3.3: Termination condition: when the Pareto frontier improvement rate is <1% for 20 consecutive generations, the search is terminated.

[0062] In S3.4, combined with the actual demand of the project, the final parameter combination most suitable for the current engineering priority is selected from the Pareto solution set, and the specific steps are as follows:

[0063] S3.4.1: Priority weight allocation: according to the type of engineering, the target weight is set according to the following rules:

[0064] Key engineering: w1(vibration reduction rate) = 0.5, w2(cost) = 0.2, w3(durability) = 0.3

[0065] General engineering: w1=0.35, w2=0.4, w3=0.25

[0066] S3.4.2: Solution set screening:

[0067] Calculate the comprehensive score of each Pareto solution:

[0068] Score = w1·S 减振 + w2·(1-S 成本 ) + w3·S 耐久

[0069] Where S 减振 , S 成本 , S 耐久 are normalized index values;

[0070] S3.4.3: Final decision: select the solution with the highest comprehensive score as the implementation scheme, and if there is a tie, give priority to the solution with higher vibration reduction rate.

[0071] After the construction according to the optimization scheme in S4, the vibration data is dynamically monitored and compared with the specification limit, and when it does not meet the standard, the damper parameters or reinforcement structure are adjusted for secondary optimization, and the specific steps are as follows:

[0072] S4.1: Use the vibration fingerprint comparison method to ensure that the actual construction parameters deviate by less than 5% from the design scheme;

[0073] S4.2: Real-time collection of acceleration, displacement and strain data through the laid wireless sensor network;

[0074] S4.3: Time history comparison analysis of the monitoring data and the vibration acceleration limit value in the "Urban Bridge Maintenance Specification";

[0075] S4.4: Based on the Bayesian network model, automatically recommend damper adjustment or structural reinforcement strategies according to the exceeding degree;

[0076] S4.5: Adjust the mass-stiffness ratio of TMD using the variable step search method, and perform topology optimization on the reinforcement structure.

[0077] Compared with the prior art, the beneficial effects of the present application are:

[0078] The present application fuses physical mechanism and deep learning to construct a multi-source information recognition model, and realizes robust identification of vibration sources and dynamic reconstruction of transmission paths by combining vibration transmission entropy and Granger causality test, providing accurate basis for subsequent vibration reduction scheme design, and selecting adaptive schemes based on vibration characteristics, constructing a four-dimensional coupled model through finite element simulation, and solving the Pareto optimal solution under the constraint of the cooperative working performance of the new and old bridges using an improved genetic algorithm and selecting parameters according to engineering priority, realizing the collaborative optimization of vibration reduction effect, structural safety, cost control and durability, thereby significantly improving the accuracy and engineering applicability of vibration disturbance control at the connection between the new and old bridges. BRIEF DESCRIPTION OF DRAWINGS

[0079] Fig. 1 This is a flowchart of a vibration disturbance control method at the connection between old and new bridges according to the present invention.

[0080] Fig. 2 This is a flowchart of vibration source identification in a vibration disturbance control method at the connection between old and new bridges according to the present invention.

[0081] Fig. 3 This is a flowchart of the optimization scheme in the vibration disturbance control method at the connection between new and old bridges of the present invention. Detailed Implementation

[0082] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0083] Example:

[0084] like Figs. 1-3 As shown, this embodiment provides a vibration disturbance control method at the connection between new and old bridges, including the following steps: S1: Sensors are deployed at the connection between the new and old bridges to collect vibration data at this location; S2: Based on a multi-source information identification model that integrates physical mechanisms and deep belief networks, combined with vibration transfer entropy analysis, the vibration source at the connection between the new and old bridges under complex environments is robustly identified and the transmission path is dynamically reconstructed; S3: A tuned mass damper or a local stiffness enhancement scheme is selected according to the vibration characteristics. A four-dimensional coupled model of "vibration reduction rate - structural stress - construction cost - durability" is constructed through finite element simulation. Based on the performance constraints of the collaborative work of the new and old bridges, an improved genetic algorithm is used to solve the multi-objective Pareto optimal solution. Finally, the scheme parameters are selected according to the project priority; S4: After construction according to the optimized scheme, the vibration data is dynamically monitored and compared with the standard limits. If the standards are not met, secondary optimization is performed by adjusting the damper parameters or reinforcing the structure.

[0085] In this embodiment, it should also be noted that the sensors in S1 include a triaxial acceleration sensor, a displacement sensor, and a strain gauge, which are symmetrically arranged on both sides of the connection between the old and new bridges, with a spacing of 0.5-2m.

[0086] Further, it needs to be explained that the three-way acceleration sensor: range recommendation ±5g (covers the bridge vibration amplitude range), sampling frequency ≥100Hz (capture 0-50Hz vibration main frequency); displacement sensor: resolution ≤0.01mm (identify the tiny vibration displacement), measurement range 0-100mm (cover the maximum opening and closing amount of expansion joint); strain gauge: sensitivity coefficient 2.0±1%, gauge length 5-10mm (adapt to the characteristics of large stress gradient near the joint). Layout position: 3-5 rows of sensors are arranged on both sides of the joint, the first row is 0.5m away from the joint edge (to capture direct vibration), and the subsequent rows are increased by 0.5-2m (to reflect the vibration attenuation law); preferentially arranged at the sensitive parts of the beam end, support near and the like.

[0087] In this embodiment, it also needs to be explained that the multi-source information recognition model based on physical mechanism and deep belief network fusion in S2 combines vibration transmission entropy analysis to robustly identify the vibration source at the joint of the new and old bridges under complex environment and dynamically reconstruct the transmission path, and the specific steps are as follows: S2.1: integrate vibration sensor data, environmental monitoring data and bridge design parameters to construct a multi-dimensional input matrix; S2.2: establish a transmission matrix model based on Timoshenko beam theory to quantify the propagation characteristics of vibration waves between the new and old bridges; S2.3: automatically learn the time-frequency characteristics of vibration signals through a 5-layer deep belief network; S2.4: calculate the vibration transmission entropy values between different measuring points and reconstruct the vibration energy transmission path combined with Granger causality test; the specific steps are as follows: S2.4.1: pre-process the vibration signals of different measuring points to remove noise interference and standardize to obtain a stationary time series; S2.4.2: based on the pre-processed time series, calculate the vibration transmission entropy values between any two measuring points, and the formula is:

[0088] TE(X→Y)=H(Y|{Y(t-1),...,Y(t-k)})

[0089] -H(Y|{Y(t-1),...,Y(t-k),X(t-1),...,X(t-m)})

[0090] Wherein, H is the information entropy, k and m are the delay orders of Y and X respectively; S2.4.3: perform Granger causality test on the vibration signals of the two measuring points, calculate the F statistic through constructing an autoregressive model, and determine whether X is the Granger cause of Y; when constructing the autoregressive model to calculate the F statistic, the time series of the vibration signals of the two measuring points are set as X t 、Y t , and the specific formula for calculating the F statistic is as follows: unconstrained autoregressive model:

[0091]

[0092] Constrained autoregressive model:

[0093]

[0094] where, a i , b i are model coefficients, e 1t , e 2t are residual terms, and p is the lag order; the F-statistic formula is:

[0095]

[0096] where, SSR r is the residual sum of squares of the constrained model, SSR ur is the residual sum of squares of the unconstrained model, and n is the sample size of the vibration signal. S2.4.4: Fuse the vibration transfer entropy value and the Granger causality test result, and mark the path that meets the condition at the same time as the dominant vibration energy transfer path to complete the path reconstruction. S2.5: Fuse the physical model output and the deep learning result using the D-S theory to generate a vibration source identification report with confidence. The specific steps are as follows: S2.5.1: Basic probability assignment: ①The physical model output result m phy (A) is generated by residual sum of squares normalization:

[0097]

[0098] where, y obs is the measured vibration data of the sensor, and y phy is the predicted value of the physical model; ②The deep learning result m dnn (A) is generated by the Softmax function:

[0099]

[0100] where, is the original score of the output layer of the deep belief network for hypothesis A, is the value of the i-th output node of the deep belief network, and K is the total number of vibration source categories; S2.5.2: Evidence fusion: calculate the fused confidence using the D-S combination rule:

[0101]

[0102] S2.5.3: Conflict processing: when the conflict factor , start the manual review process; S2.5.4: Report generation: output the confidence interval [Bel(A), Pl(A)] of each vibration source hypothesis, where:

[0103] Further, it needs to be explained that the environmental monitoring data includes temperature, humidity, wind speed; the bridge design parameters include beam height, reinforcement ratio, material elastic modulus. The vibration wave type is distinguished: the 5-layer deep belief network structure is as follows: ① input layer: short-time Fourier transform spectrum of vibration signal, size 64x64; ② 1-4 layers: restricted Boltzmann machine, neuron numbers are 512, 256, 128, 64 respectively; ③ output layer: fully connected layer + Softmax, output vibration source category probability distribution. The training process is divided into two stages: ① layer-by-layer pre-training: each RBM is unsupervised trained using the contrast divergence algorithm; ② global fine-tuning: supervised fine-tuning is performed on the labeled data set using the back propagation algorithm, and the loss function is cross entropy. The input signal preprocessing includes detrending, band-pass filtering (0.5-50Hz), normalization. Conflict processing threshold: conflict factor K>0.3 starts manual review, which is derived from engineering statistics: when K≤0.3, the difference between the physical model and the deep learning result is "weak conflict" (the difference can be fused by evidence theory); when K>0.3, the difference is beyond the complementary range of the model (such as sensor failure causing input error of the physical model), which needs manual intervention to troubleshoot.

[0104] In the present embodiment, it also needs to be explained that in S3, the tuned mass damper or local stiffness enhancement scheme is selected according to the vibration characteristics, a four-dimensional coupling model of "vibration reduction rate-structure stress-construction cost-durability" is constructed through finite element simulation, based on the constraint of new and old bridge cooperative working performance, the improved genetic algorithm is used to solve the multi-objective Pareto optimal solution, and finally the scheme parameters are selected according to the engineering priority, the specific steps are as follows: S3.1: according to the vibration characteristics identified in S2, the applicable vibration reduction measures are preliminarily screened, and it is judged whether to use tuned mass damper or local stiffness enhancement scheme; S3.2: the vibration reduction rate, structure stress level, construction cost and durability index under different scheme parameters are quantified through finite element simulation, and they are normalized, and a multi-objective comprehensive evaluation system is constructed; S3.3: under the constraint of new and old bridge cooperative working performance, the improved genetic algorithm is used for global search to solve the Pareto optimal solution; the specific steps are as follows: S3.3.1: constraint processing: the dynamic penalty function is used to process the new and old bridge cooperative constraint:

[0105]

[0106] wherein, g j (x) is the constraint function, "j=1,2,3,4", g1(x) is the material modulus difference constraint ≤15%, g2(x) is the expansion joint displacement constraint ≤50mm, g3(x) is the support reaction force balance constraint ≤10%, g4(x) is the inherent frequency matching constraint ≤5%; S3.3.2: algorithm improvement: ① cross probability self-adaptive adjustment:

[0107]

[0108] Where, t is the current generation number, T is the total generation number; ② The elite reserve ratio is fixed at 15% of the population size; S3.3.3: Termination condition: terminate the search when the improvement rate of the Pareto frontier is less than 1% for 20 consecutive generations. S3.4: Select the final parameter combination that best meets the current engineering priority from the Pareto solution set according to the actual needs of the project. The specific steps are as follows: S3.4.1: Priority weight allocation: set the target weight according to the following rules according to the type of project:

[0109] Key project: w1 (vibration reduction rate) = 0.5, w2 (cost) = 0.2, w3 (durability) = 0.3

[0110] General project: w1 = 0.35, w2 = 0.4, w3 = 0.25

[0111] S3.4.2: Solution set screening: calculate the comprehensive score of each Pareto solution:

[0112] Score = w1 · S 减振 + w2 · (1 - S 成本 ) + w3 · S 耐久

[0113] Where, S 减振 , S 成本 , S 耐久 are the normalized index values; S3.4.3: Final decision: select the solution with the highest comprehensive score as the implementation scheme, and if there is a tie, prefer the solution with higher vibration reduction rate.

[0114] Further, it needs to be explained that the tuning mass damper is suitable for the scene of single vibration frequency (such as 1-3 Hz vertical vibration dominated by vehicle load), which can quickly reduce vibration through resonance energy absorption; Local stiffness enhancement scheme (such as pasting carbon fiber cloth, adding steel longitudinal beam) is suitable for the scene where vibration is caused by stiffness mutation (such as the difference in bending stiffness at the joint between new and old bridges), which can weaken vibration transmission by improving stiffness distribution. S3.2 Index normalization formula: ① Vibration reduction rate (positive direction): ② Structural stress (negative direction): ③ Construction cost (negative direction), durability (positive direction) Similarly, ensure that the index value domain is unified to [0, 1].

[0115] In the present embodiment, it is also necessary to note that after construction in S4 according to the optimization scheme, dynamic monitoring of vibration data is performed and compared with the specification limit, and when the standard is not met, secondary optimization is performed by adjusting the damper parameters or reinforcing the structure, and the specific steps are as follows: S4.1: Adopting the vibration fingerprint comparison method, ensure that the actual construction parameters and the design scheme deviation is less than or equal to 5%; S4.2: Through the laid wireless sensor network, real-time collection of acceleration, displacement and strain data; S4.3: Time comparison analysis of monitoring data and vibration acceleration limit in the "Urban Bridge Maintenance Specification"; S4.4: Based on the Bayesian network model, according to the exceeding standard degree, automatically recommend damper adjustment or structural reinforcement strategy; S4.5: Adjust the mass-stiffness ratio of TMD using the variable step search method, and perform topology optimization on the reinforced structure.

[0116] Further, it is necessary to note that TMD is Tuned Mass Damper; the specific steps of the vibration fingerprint comparison method: collect the environmental vibration data of the first 3 days after construction; calculate the MAC (modal assurance criterion) value of the design model; when MAC<0.9, it is determined that the deviation exceeds the standard.

[0117] In the description of the present specification, the description of the terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are contained in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0118] The preferred embodiments of the application disclosed above are only used to help explain the application. The preferred embodiments do not describe all the details, nor limit the application to the specific embodiments described. Obviously, many modifications and changes can be made according to the content of the present specification. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the application, so that those skilled in the art can well understand and utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A method for controlling vibration disturbance at the connection between old and new bridges, characterized in that, Includes the following steps: S1: Vibration data of the area is collected by installing sensors at the connection between the old and new bridges; S2: Based on a multi-source information identification model that integrates physical mechanisms and deep belief networks, combined with vibration transmission entropy analysis, robust identification of vibration sources at the connection between old and new bridges in complex environments and dynamic reconstruction of transmission paths are achieved. S3: Select a tuned mass damper or a local stiffness enhancement scheme based on vibration characteristics. Construct a four-dimensional coupled model of "vibration reduction rate - structural stress - construction cost - durability" through finite element simulation. Based on the performance constraints of the collaborative work of the new and old bridges, use an improved genetic algorithm to solve the multi-objective Pareto optimal solution. Finally, select the scheme parameters according to the project priority. S4: After construction according to the optimized plan, the vibration data is dynamically monitored and compared with the standard limits. If the standard is not met, secondary optimization is carried out by adjusting the damper parameters or reinforcing the structure.

2. The vibration disturbance control method at the connection between old and new bridges according to claim 1, characterized in that, The sensors in S1 include a triaxial accelerometer, a displacement sensor, and a strain gauge, which are symmetrically arranged on both sides of the connection between the old and new bridges, with a spacing of 0.5 to 2m.

3. The vibration disturbance control method at the connection between old and new bridges according to claim 1, characterized in that, The multi-source information identification model in S2, which integrates physical mechanisms and deep belief networks, combined with vibration transmission entropy analysis, robustly identifies vibration sources at the connection between old and new bridges in complex environments and dynamically reconstructs their transmission paths. The specific steps are as follows: S2.1: Integrate vibration sensor data, environmental monitoring data, and bridge design parameters to construct a multi-dimensional input matrix; S2.2: Based on Timoshenko beam theory, a transfer matrix model is established to quantify the propagation characteristics of vibration waves between old and new bridges; S2.3: Automatically learns the time-frequency characteristics of vibration signals through a 5-layer deep belief network; S2.4: Calculate the vibration transfer entropy between different measuring points, and reconstruct the vibration energy transfer path by combining Granger causality test; S2.5: The DS theory is used to fuse the physical model output with the deep learning results to generate a vibration source identification report with confidence.

4. The vibration disturbance control method at the connection between old and new bridges according to claim 3, characterized in that, In step S2.4, the vibration transfer entropy value between different measuring points is calculated, and the vibration energy transfer path is reconstructed using the Granger causality test. The specific steps are as follows: S2.4.1: Preprocess the vibration signals at different measuring points to remove noise interference and standardize them to obtain a stationary time series; S2.4.2: Based on the preprocessed time series, calculate the vibration transfer entropy between any two measuring points using the following formula: TE(X→Y)=H(Y|{Y(t-1),...,Y(tk)}) -H(Y|{Y(t-1),...,Y(tk),X(t-1),...,X(tm)}) Where H is the information entropy, and k and m are the delay orders of Y and X, respectively; S2.4.3: Perform Granger causality test on the vibration signals of the two measuring points, calculate the F statistic by constructing an autoregressive model, and determine whether X is a Granger cause of Y; S2.4.4: Integrate the vibration transmission entropy value with the Granger causality test results, mark the paths that simultaneously meet the conditions as the dominant vibration energy transmission paths, and complete the path reconstruction.

5. The vibration disturbance control method at the connection between old and new bridges according to claim 4, characterized in that, In S2.4.3, when constructing the autoregressive model to calculate the F statistic, let the time series of the vibration signals at the two measuring points be X. t Y t The specific formula for calculating the F-statistic is as follows: Unconstrained autoregressive model: Constrained autoregressive model: Where, α i β i For model coefficients, ∈ 1t ,∈ 2t The term represents the residual, and p represents the lag order. The formula for calculating the F-statistic is: Among them, SSR r To constrain the sum of squared residuals of the model, SSR ur Let n be the sum of squared residuals of the unconstrained model, and n be the number of vibration signal samples.

6. The vibration disturbance control method at the connection between old and new bridges according to claim 3, characterized in that, In step S2.5, the physical model output and deep learning results are fused using DS theory to generate a vibration source identification report with confidence. The specific steps are as follows: S2.5.1: Basic probability allocation: ① Physical model output results m phy (A) Generated by normalization of the sum of squared residuals: Among them, y obs For the actual vibration data measured by the sensor, y phy These are the predicted values ​​from the physical model; ②Deep learning result m dnn (A) Generated using the Softmax function: Among them, z A z represents the original score of the output layer of the deep belief network for hypothesis A. i Let K be the value of the i-th output node of the deep belief network, and K be the total number of vibration source categories. S2.5.2: Evidence Fusion: The confidence level after fusion is calculated using the DS synthesis rule. S2.5.3: Conflict Handling: When conflict factors In such cases, a manual review process will be initiated. S2.5.4: Report Generation: Outputs the confidence intervals [Bel(A),Pl(A)] for each vibration source hypothesis, where:

7. The vibration disturbance control method at the connection between old and new bridges according to claim 1, characterized in that, In step S3, a tuned mass damper or a local stiffness enhancement scheme is selected based on vibration characteristics. A four-dimensional coupled model of "vibration reduction rate - structural stress - construction cost - durability" is constructed through finite element simulation. Based on the performance constraints of the collaborative work between the old and new bridges, an improved genetic algorithm is used to solve for the multi-objective Pareto optimal solution. Finally, the scheme parameters are selected according to the project priority. The specific steps are as follows: S3.1: Based on the vibration characteristics identified in S2, preliminary screening of suitable vibration reduction measures is conducted to determine whether to use a tuned mass damper or a local stiffness enhancement scheme. S3.2: The vibration reduction rate, structural stress level, construction cost and durability index under different scheme parameters are quantified by finite element simulation, and they are uniformly normalized to construct a multi-objective comprehensive evaluation system. S3.3: Under the performance constraints of the collaborative operation of the old and new bridges, an improved genetic algorithm is used to perform a global search to find the Pareto optimal solution; S3.4: Based on the actual needs of the project, select the final parameter combination that best suits the current project priority from the Pareto solution set.

8. The vibration disturbance control method at the connection between old and new bridges according to claim 7, characterized in that, In S3.3, under the performance constraints of the collaborative operation of the old and new bridges, an improved genetic algorithm is used for global search to find the Pareto optimal solution. The specific steps are as follows: S3.3.1: Constraint Handling: A dynamic penalty function is used to handle the collaborative constraints between the old and new bridges. Among them, g j (x) is a constraint function, "j=1,2,3,4", g1(x) is a material modulus difference constraint ≤15%, g2(x) is an expansion joint displacement constraint ≤50mm, g3(x) is a support reaction force balance constraint ≤10%, and g4(x) is a natural frequency matching constraint ≤5%. S3.3.2: Algorithm Improvement: ① Adaptive adjustment of crossover probability: Where t is the current algebra and T is the total algebra; ②The proportion of elites retained is fixed at 15% of the population size; S3.3.3: Termination condition: The search is terminated when the Pareto front improvement rate is less than 1% for 20 consecutive generations.

9. The vibration disturbance control method at the connection between old and new bridges according to claim 7, characterized in that, In step S3.4, the final parameter combination that best meets the current project priority is selected from the Pareto solution set based on the actual needs of the project. The specific steps are as follows: S3.4.1: Priority Weight Allocation: Set the target weight according to the following rules based on the project type: Key project parameters: w1 (vibration reduction rate) = 0.5, w2 (cost) = 0.2, w3 (durability) = 0.3 For general engineering projects: w1 = 0.35, w2 = 0.4, w3 = 0.25 S3.4.2: Solution set filtering: Calculate the overall score for each Pareto solution: Score=w1·S 减振 +w2·(1-S 成本 )+w3·S 耐久 Among them, S 减振 S 成本 S 耐久 These are the normalized index values; S3.4.3: Final decision: Select the solution with the highest comprehensive score as the implementation plan. If a tie occurs, prioritize the solution with the higher vibration reduction rate.

10. The vibration disturbance control method at the connection between old and new bridges according to claim 1, characterized in that, In step S4, after construction according to the optimized scheme, vibration data is dynamically monitored and compared with the standard limits. If the standards are not met, secondary optimization is performed by adjusting the damper parameters or reinforcing the structure. The specific steps are as follows: S4.1: The vibration fingerprint comparison method is adopted to ensure that the deviation between the actual construction parameters and the design scheme is ≤5%; S4.2: Acceleration, displacement and strain data are collected in real time through a deployed wireless sensor network; S4.3: Compare and analyze the monitoring data with the vibration acceleration limits in the "Urban Bridge Maintenance Code" over time. S4.4: Based on a Bayesian network model, automatically recommend damper adjustment or reinforcement strategies according to the degree of exceedance; S4.5: The mass-stiffness ratio is adjusted by a variable step size search method for the TMD, and topology optimization is performed on the reinforcement structure.

Images