Bridge engineering hidden disease detection method based on intelligent algorithm and magnetic memory

Abstract

The invention discloses a bridge engineering hidden disease detection method based on an intelligent algorithm and magnetic memory. The method comprises the following steps: acquiring initial information of a bridge structure; the bridge structural members are grouped according to the member types; taking the initial information of the bridge structure and the construction grouping condition as modeling basis, and establishing a bridge structure analysis model by means of structure analysis software; the structural analysis software simulates damage of components at different parts of the bridge structure, and then modal information after damage is obtained; performing analysis, induction and arrangement on modal data of different damaged parts to form a training sample; establishing a damage positioning model by adopting an intelligent algorithm; the magnetic sensor collects data for area detection; positioning defect positions; the field control computer processes the data and transmits the data back to the PC terminal computer, thereby realizing the functions of visual display and statistics. According to the invention, accurate damage position positioning and high damage degree detection efficiency are realized in bridge structure detection.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge engineering detection, and more specifically, to a method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory. Background Art

[0002] With the development of science and technology, metal materials are more widely used in bridge engineering. Due to the influence of harsh climate environments and complex external forces, metal materials will have varying degrees of defect damage. Once defects appear in the materials, if not discovered and measures taken in time, it will not only cause economic losses and major accidents, but may also cause casualties. If the damage to the internal steel bars of the bridge can be detected in advance, scientific maintenance methods can be adopted according to the detection results to extend the service life of the bridge.

[0003] In China, the applied bridge engineering detection methods mainly focus on the detection and evaluation of the bridge appearance and bearing capacity, and less attention is paid to the damage caused by hidden diseases inside the bridge, such as steel bar corrosion, corrosion and broken wires of stay cables, steel bar stress, internal prestress, fatigue damage of steel structures, etc. The detection means for hidden diseases of bridges have not yet formed a complete system, and currently it is still a technical difficulty to be overcome.

[0004] Currently, the non-destructive detection techniques for internal defects of bridges mainly include magnetoacoustic emission detection method (MAE), ultrasonic detection method (UT), ray detection method (RT), magnetic particle detection method (MPT), and magnetic flux leakage detection method (MFL), etc. Although these methods can detect the damaged parts of the bridge structure, they generally have disadvantages such as low accuracy, difficult positioning, and complex operation. As a new method in the field of green non-destructive detection, the metal magnetic memory detection technique can magnetize ferromagnetic materials using the natural geomagnetic field in a complex detection environment, and can perform early prevention and diagnosis on the microscopic defects on the surface of the test piece without preprocessing the surface of the test piece.

[0005] In actual engineering, the workload of detecting internal defects of bridge structures is extremely large, and the proportion of undamaged components is relatively high. Using the existing local detection methods to detect each component of the bridge structure one by one generally has disadvantages such as large workload, low efficiency, and uneconomical. Combining with the rapidly developing intelligent algorithms in recent years will effectively solve this problem. Intelligent algorithms are widely used in many fields such as civil engineering because of their powerful large-scale data processing capabilities, automatic learning capabilities, adaptive capabilities, etc. The structural damage location method applying intelligent algorithms does not require prior knowledge of the structural dynamic characteristics and has the advantage of non-parametric damage location; it can obtain the implicit relationship between input and output hidden in the sample data through training and learning, and can filter out noise and extract the inherent characteristics of things themselves in the presence of noise. Therefore, it is more suitable for damage location of structures with a large amount of noise and measurement errors. Summary of the Invention

[0006] In view of the two major problems of low detection efficiency and difficult damage location existing in the current detection technology for hidden diseases of bridges, the present invention proposes a method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory. By combining the initial information of the bridge structure, intelligent algorithms, and magnetic memory detection technology, this method realizes the accurate location and efficient detection of hidden diseases of bridges. Hidden diseases include welds of steel structure bridges, nodes, and connection parts of composite structures.

[0007] The method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory of the present invention includes the following steps:

[0008] S1 Obtain the initial information of the bridge structure.

[0009] S2 Group the bridge structure components according to the component types.

[0010] S3 Take the initial information of the bridge structure and the component grouping situation as the basis for modeling, and establish a bridge structure analysis model with the help of structural analysis software.

[0011] S4 The structural analysis software performs damage simulation on each component in different parts of the bridge structure, and then obtains the modal data after damage.

[0012] S5 Analyze and summarize the modal data after damage in different parts, and form training samples.

[0013] S6 Use intelligent algorithms to combine, learn, and iterate the training samples formed by the damage conditions and corresponding modal parameters of different components of the bridge structure to establish a damage location model.

[0014] S7 According to the location information given by the damage location model, combined with the collected magnetic memory data, verify and optimize the damage location model.

[0015] The magnetic memory data is collected by using a magnetic memory detection instrument to scan the area to be measured through a micro magnetic sensor driven by a stepping motor.

[0016] The structural analysis software is ansys software.

[0017] The analysis and induction of the modal data are manually sorted out.

[0018] Furthermore, in step S1 of the present invention, the initial information of the bridge structure includes the completion drawing information of the bridge, structural material information, bridge detection and monitoring information, and the current state structural modal information.

[0019] Furthermore, in step S2 of the present invention, in order to improve the accuracy and convergence speed of the intelligent algorithm for determining the location of structural damage, the overall bridge structure is discretized into different sub-components according to component types, and is subdivided into main girders, piers and abutments, bridge towers, truss joints, box girder welds, suspenders, arch ribs, arch feet, and shear keys according to the force conditions of the bridge structure.

[0020] Furthermore, in step S3 of the present invention, when using the construction analysis software for modeling, it is necessary to first establish an analysis model of the bridge structure in a non-damaged state according to the initial information of the bridge structure, group the model according to sub-component types, and digitize the damage conditions of each sub-component for later statistical information to form training samples.

[0021] Furthermore, in step S4 of the present invention, in the process of simulating the structural damage location, all sub-components that may be damaged should be damaged one by one, and the damage modal information in the single-component damage state should be obtained;

[0022] The introduced damage method adopts reducing the modulus of the damaged part, setting unit notches, or even removing some units; specifically, damage is introduced into the original non-damaged model, and the damaged parts are determined according to the grouping of sub-components, ensuring that at least one damage is introduced into each sub-component, and the modal parameters of different parts of the bridge structure in the single-damage state are output by using the structural analysis software.

[0023] The single-damage state means that in the structural analysis model, only the damaged condition of a single component needs to be simulated each time. The existing intelligent algorithm randomly combines the damaged conditions and corresponding modal samples of each single-component to obtain the modal information corresponding to the multi-damage condition of multiple components damaged simultaneously; after the single-damage state and the data of randomly combining single-component damage, a bridge structure damage location model is obtained.

[0024] Furthermore, in step S5 of the present invention, the training samples are universal for the current and subsequent detections of the bridge structure to be detected. No matter how the performance of the bridge structure degrades during subsequent service, this data is still applicable and belongs to the permanent bridge structure damage location model.

[0025] Furthermore, in step S6 of the present invention, after the bridge structure damage location model is established, it is necessary to calculate the bridge structure damage location by back-calculating the modal information of the measured bridge structure, and judge the damage location accuracy of the intelligent algorithm. If the preset accuracy requirement is met, the intelligent damage location model is established. Otherwise, it is necessary to re-iterate and optimize the intelligent algorithm until the preset accuracy requirement is met.

[0026] Furthermore, the intelligent algorithm of the present invention uses the normalized damage signal index NDSI as the network input vector for specific damage location and uses the normalized damage signal index NDSI as the damage identification of the network input vector.

[0027] The beneficial effects of the present invention are as follows:

[0028] Improve detection efficiency: By combining intelligent algorithms and magnetic memory detection technology, the present invention can achieve rapid positioning of hidden diseases in bridges, greatly shortening the detection time and improving the detection efficiency. This helps reduce detection costs, while accelerating the progress of bridge repair and maintenance, ensuring the safety and stability of the bridge structure.

[0029] Precisely locate diseases: The present invention uses intelligent algorithms to learn and iterate on the damage conditions of different components of the bridge structure and the corresponding magnetic memory signal characteristics, establishing a high-precision damage location model. This enables inspectors to more accurately locate the positions of hidden diseases in bridges, reducing the possibility of misjudgment and missed judgment.

[0030] Enhance detection versatility: The established structural data model of the present invention is universal. Regardless of how the performance of the bridge structure degrades during subsequent service, this data model is still applicable. This means that this method can be long-term applied to the monitoring and maintenance of bridge structures, providing continuous guarantee for the safe operation of bridges.

[0031] Optimize bridge management: By presenting the current status of the bridge and the prediction results in real time, the present invention provides strong decision-making support for bridge management and maintenance units. The management and maintenance units can formulate repair and maintenance plans in a timely manner based on this information, optimize bridge management strategies, and ensure the long-term safe operation of the bridge structure.

[0032] Promote technological progress: The proposal and implementation of the present invention promote the development and innovation of hidden disease detection technology for bridges. By combining intelligent algorithms and magnetic memory detection technology, the present invention provides new ideas and methods for the monitoring and maintenance of bridge structures, helping to promote the technological progress and development of the entire industry.

[0033] In summary, the present invention shows significant beneficial effects in improving detection efficiency, precisely locating diseases, enhancing detection versatility, optimizing bridge management, and promoting technological progress. Brief Description of the Drawings

[0034] Figure 1a is the flow chart of the present invention.

[0035] Figure 1b is the structural schematic diagram of the detection method of the present invention.

[0036] Figure 2 is the flow chart of the intelligent algorithm of the present invention.

[0037] Figure 3 is the flow chart of the magnetic memory technology of the present invention. Detailed Description of the Invention

[0038] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0039] Figure 1a 、 1b It is a structural diagram of a method for detecting hidden diseases of bridge structures based on intelligent algorithms and magnetic memory technology, which includes the Figure 2 described intelligent algorithm system and the Figure 3 described magnetic memory technology system, and specifically includes the following steps

[0040] I. Preliminary preparation stage

[0041] Obtain the initial information of the bridge structure (S1):

[0042] Collect the completion drawings, structural material information, inspection and monitoring records from the bridge structure operator to comprehensively understand the basic structure and historical conditions of the bridge.

[0043] Conduct dynamic characteristic tests to obtain the modal parameters of the current bridge structure and provide basic data for subsequent analysis.

[0044] Group the bridge structure components (S2):

[0045] According to the stress conditions and component types of the bridge structure, discretize the overall structure into multiple sub-components, such as steel truss joints, steel box girder welds, suspenders, arch ribs, arch feet, shear keys, etc.

[0046] To improve the accuracy and convergence speed of the intelligent algorithm, ensure that at least one damage is introduced to each sub-component in subsequent simulations.

[0047] II. Model construction and damage simulation stage

[0048] Establish an analysis model of the bridge structure (S3):

[0049] When using construction analysis software to build a model, first establish an analysis model of the bridge structure in a non-damaged state according to the initial information of the bridge structure, group the model according to the sub-component types, and digitize the damage conditions of each sub-component for forming training samples for later statistical information.

[0050] Bridge structure damage simulation (S4):

[0051] In the simulation process of the damage parts of the structure, damage simulations should be carried out one by one for all the sub-components where damage may occur, and the damage modal information in the single-component damage state should be obtained;

[0052] The introduced damage method adopts reducing the modulus of the damaged part, setting unit notches, or even removing some units; specifically, damage is introduced into the original undamaged model, and the damaged parts are determined according to the grouping of sub-components to ensure that at least one damage is introduced into each sub-component. The modal parameters of different parts of the bridge structure in the single-damage state are output using structural analysis software.

[0053] The single-damage state means that in the structural analysis model, only the damaged conditions of a single component need to be simulated each time. The damaged conditions of each single component and the corresponding modal samples are randomly combined by the existing intelligent algorithm to obtain the modal information corresponding to the multi-damage conditions where multiple components are damaged simultaneously. After the data of the single-damage state and the randomly combined single-component damage, a bridge structure damage location model is obtained.

[0054] Form training samples (S5):

[0055] Organize the modal data after damage at different parts to form training samples containing the damage conditions of different components of the bridge structure and the corresponding modal parameters.

[0056] The training samples are universal for the current and subsequent inspections of the bridge structure to be detected. No matter how the performance of the bridge structure degrades during subsequent service, this data is still applicable and belongs to the permanent bridge structure damage location model.

[0057] III. Intelligent algorithm and damage location model construction stage

[0058] Establish a damage location model (S6):

[0059] Adopt intelligent algorithms such as machine learning and deep learning to combine, learn, and iterate the training samples to establish a bridge structure damage location model.

[0060] After the bridge structure damage location model is established, it is necessary to calculate the damage location of the bridge structure by back-calculating according to the measured modal information of the bridge structure, and judge the damage location accuracy of the intelligent algorithm. If the preset accuracy requirement is met, the intelligent damage location model is established; otherwise, the intelligent algorithm needs to be iteratively optimized until the preset accuracy requirement is met.

[0061] Calculate the damage location by back-calculating according to the measured modal information of the bridge structure, and verify the location accuracy of the intelligent algorithm. If the location error is less than the maximum scanning range of the magnetic memory technology, the model is established; otherwise, the algorithm is iteratively optimized until the accuracy requirement is met.

[0062] IV. On-site inspection and data processing stage

[0063] On-site inspection and location (S7):

[0064] Based on the positioning information given by the bridge structure damage location model, use the magnetic memory system to conduct regional detection.

[0065] By changing the settings of the multi-channel instrument, the movement of the magnetic memory focus is realized to cover the area to be detected.

[0066] The detector holds the array probe to emit signals to the area to be detected, the receiving end receives the echo signals, and locates the defect position through time series inversion and the maximum amplitude time point.

[0067] Data processing and display (S8 & S9):

[0068] The on-site control computer processes the received data, including signal filtering, feature extraction, etc., to improve the data quality.

[0069] Through long-distance transmission communication mode, the processed data is input into the PC terminal computer in the control room.

[0070] The PC terminal computer stores, analyzes and visually displays the data, such as generating three-dimensional images, drawing damage distribution maps, etc., to realize statistical functions and assist in decision-making.

[0071] During the data transmission process, bus conversion technology is adopted to ensure the stability and reliability of long-distance transmission.

[0072] Main applications of the method of the present invention: on-site quantitative detection and pre-judgment:

[0073] Based on the positioning information given by the bridge structure damage location model, guide the detector to conduct on-site quantitative detection accurately and efficiently. During the detection process, use the magnetic memory detection instrument to realize the scanning of the area to be measured through the micro magnetic sensor driven by the stepping motor, and collect magnetic memory data to further verify and optimize the damage location model.

[0074] Combined with factors such as vehicle load, pre-judge the damage of bridge components, and display the pre-judgment results and the current situation of the bridge in real time in the bridge three-dimensional model.

[0075] Guide maintenance and repair:

[0076] Utilize the real-time data and pre-judgment results in the bridge three-dimensional model to guide the maintenance unit to conduct maintenance and repair work in a timely manner, ensuring the safety and stability of the bridge structure.

[0077] Through the above implementation steps, the present invention successfully combines the intelligent algorithm with the magnetic memory detection technology, realizing the accurate positioning and efficient detection of hidden diseases of bridges. This not only improves the detection efficiency, but also enhances the accuracy and reliability of the detection results, providing strong support for the safety management and timely repair of bridge structures.

[0078] The content described in the embodiments of this specification is only an enumeration of the implementation forms of the inventive concept. The protection scope of the present invention should not be regarded as limited to the specific forms stated in the embodiments. The protection scope of the present invention also includes equivalent technical means that those skilled in the art can think of based on the inventive concept.

Claims

1. A method for detecting hidden diseases in bridge engineering based on intelligent algorithms and magnetic memory, characterized in that, It includes the following steps: S1 Obtain the initial information of the bridge structure; S2 Group the bridge structure components according to the component types; S3 Use the initial information of the bridge structure and the component grouping situation as the basis for modeling, and establish a bridge structure analysis model with the help of structural analysis software; S4 The structural analysis software conducts damage simulation on the components of different parts of the bridge structure one by one, and then obtains the modal data after damage; S5 Analyze, summarize and organize the modal data after damage to different parts to form training samples; S6 Use an intelligent algorithm to combine, learn and iterate the training samples formed by the damage conditions and corresponding modal parameters of different components of the bridge structure to establish a damage location model; S7 According to the location information given by the damage location model, combined with the collected magnetic memory data, verify and optimize the damage location model; The magnetic memory data is collected by using a magnetic memory detection instrument to scan the area to be measured driven by a micro magnetic sensor through a stepping motor.

2. The method for detecting hidden diseases in bridge engineering based on intelligent algorithms and magnetic memory according to claim 1, wherein In step S1, the initial information of the bridge structure includes the completion drawing information of the bridge, the structural material information, the bridge detection and monitoring information, and the current state structural modal information.

3. The method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory according to claim 1, characterized in that, In step S2, in order to improve the accuracy and convergence speed of the intelligent algorithm for determining the structural damage location, the overall bridge structure is discretized into different sub-components according to the component types, and is subdivided into main girders, piers, bridge towers, truss joints, box girder welds, suspenders, arch ribs, arch feet, and shear keys according to the force conditions of the bridge structure.

4. The method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory according to claim 1, characterized in that, In step S3, when using the construction analysis software for modeling, it is necessary to first establish a bridge structure analysis model in a non-damaged state according to the initial information of the bridge structure, group the model according to the sub-component types, and digitize the damage conditions of each sub-component for forming training samples for later statistical information.

5. The method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory according to claim 4, wherein In step S4, in the process of simulating the damaged parts of the structure, all possible damaged sub-components should be damaged one by one, and the damaged modal information in the single-component damaged state should be obtained; The introduced damage method adopts reducing the modulus of the damaged part, setting unit notches, or even removing some units; specifically, damage is introduced into the original non-damaged model, and the damaged parts are determined according to the sub-component grouping situation to ensure that at least one damage is introduced to each sub-component, and the structural analysis software is used to output the modal parameters of different parts of the bridge structure in the single-damaged state; The single-damaged state means that each time only the damaged condition of a single component needs to be simulated in the structural analysis model. The existing intelligent algorithm randomly combines the damaged conditions of each single component and the corresponding modal samples to obtain the modal information corresponding to the multi-damaged condition of multiple components damaged at the same time; after the single-damaged state and the data of randomly combined single-component damages, a bridge structure damage location model is obtained.

6. The method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory according to claim 5, characterized in that In step S5, the training samples are universal for the current detection and subsequent detections of the bridge structure to be detected. No matter how the performance of the bridge structure degrades during subsequent service, this data is still applicable and belongs to the permanent bridge structure damage location model.

7. The method for detecting hidden diseases of bridge engineering based on intelligent algorithm and magnetic memory according to claim 5, characterized in that, In step S6, after the establishment of the bridge structure damage location model, it is necessary to calculate the bridge structure damage location by back-calculating according to the measured bridge structure modal information, and judge the damage location accuracy of the intelligent algorithm. If the preset accuracy requirement is met, the intelligent damage location model is established; otherwise, the intelligent algorithm needs to be iteratively optimized until the preset accuracy requirement is met.

8. The method for detecting hidden diseases of bridge engineering based on intelligent algorithms and magnetic memory according to claim 5, characterized in that, The intelligent algorithm uses the normalized damage signal index NDSI as the network input vector for specific damage location, and uses the normalized damage signal index NDSI as the damage identification of the network input vector.

Images