Damage monitoring method and device, electronic equipment and storage medium

Abstract

This invention provides a damage monitoring method, device, electronic device, and storage medium. The method includes: acquiring displacement time history data of at least one first test point in a utility tunnel and a damaged area to be detected in the utility tunnel; inputting the displacement time history data into a preset neural network model to obtain stress time history data of the damaged area to be detected; determining damage parameters corresponding to the damaged area to be detected based on the stress time history data; and determining the degree of damage to the utility tunnel based on the damage parameters. This invention addresses the shortcomings of fatigue damage monitoring in the field of integrated utility tunnel operation monitoring, reduces monitoring costs, acquires fatigue damage as a monitoring indicator, improves the safety operation and maintenance system of integrated utility tunnels, and provides more guidance for the operation and maintenance of integrated utility tunnels.

Description

Technical Field

[0001] This invention relates to the field of urban underground integrated pipe gallery monitoring technology, and in particular to a damage monitoring method, device, electronic equipment and storage medium. Background Technology

[0002] With the advancement of urbanization in my country, the construction of integrated utility tunnels, serving as the "lifeline" of cities, is also booming. How to reduce the operation and maintenance risks of these tunnels and ensure their structural safety has become a crucial issue. Integrated utility tunnels are typically buried beneath urban roads with shallow overburden, inevitably subject to the influence of vehicle loads above, causing continuous fluctuations in the internal forces of the underground tunnels, resulting in fatigue damage that affects operational safety.

[0003] Fatigue damage refers to the damage that occurs to a structure under repeated loading below its bearing limit. When the cumulative fatigue damage value exceeds 1, the structure will fail. However, fatigue damage is an abstract concept and is currently difficult to describe with specific physical indicators. It can only be calculated based on stress-time history data. Therefore, fatigue damage is not considered a monitoring indicator in the operation and monitoring of integrated utility tunnels, and there are no corresponding monitoring methods. With the increasing number of integrated utility tunnels in my country, ensuring their operational safety is becoming increasingly important. If fatigue damage monitoring of utility tunnels could be implemented, it would undoubtedly further improve the operational safety monitoring system for utility tunnels.

[0004] In summary, relevant studies indicate that vehicle-induced fatigue damage is a significant factor affecting the operational safety of shallow-buried utility tunnels. However, current monitoring of vehicle-induced fatigue damage to tunnel structures is almost nonexistent. Furthermore, there is currently no effective solution to this problem. Summary of the Invention

[0005] In view of this, the main objective of the present invention is to provide a damage monitoring method, device, electronic device, and storage medium.

[0006] To achieve the above objectives, the technical solution of the present invention is implemented as follows:

[0007] This invention provides a damage monitoring method, the method comprising:

[0008] Obtain displacement time history data of at least one first test point in the pipe gallery under test, as well as the damaged parts to be detected in the pipe gallery under test;

[0009] The displacement time history data is input into a preset neural network model to obtain the stress time history data of the damaged part to be detected;

[0010] The damage parameters corresponding to the damage site to be detected are determined based on the stress time history data.

[0011] The degree of damage to the tested pipe gallery is determined based on the damage parameters.

[0012] In the above scheme, obtaining displacement time history data of at least one first test point in the pipe gallery includes:

[0013] At preset time intervals, displacement time history data of at least one first test point on each cross section of the pipe gallery under test are acquired.

[0014] In the above scheme, determining the damaged location to be detected in the pipe gallery based on the damage value corresponding to each second test point includes:

[0015] The damage values ​​corresponding to each second test point are sorted to obtain a first sorting result of the damage values ​​corresponding to each second test point;

[0016] Identify the second test point corresponding to the damage value greater than or equal to the first preset threshold in the first sorting result, and use the second test point corresponding to the damage value greater than or equal to the first preset threshold in the sorting result as the damage location to be detected in the pipe gallery to be tested.

[0017] In the above scheme, the method further includes:

[0018] Obtain displacement time history data of at least one first measuring point and stress time history data of at least one fatigue-vulnerable part in the sample pipe gallery;

[0019] The displacement time history data and the stress time history data are used as the first training samples to train the first neural network model, thereby obtaining the second neural network model corresponding to each fatigue-vulnerable part in the at least one fatigue-vulnerable part.

[0020] The first training sample is processed based on the second neural network model to obtain at least one second training sample;

[0021] The second neural network model is trained using each of the at least one second training sample to obtain the preset neural network model.

[0022] In the above scheme, the step of processing the first training sample based on the second neural network model to obtain at least one second training sample includes:

[0023] The displacement time history data in the training samples are processed according to the second neural network model to obtain the average influence value (MIV) corresponding to the displacement data in the training samples;

[0024] The at least one second training sample is determined based on the MIV value.

[0025] In the above scheme, determining the at least one second training sample based on the MIV value includes:

[0026] Based on the MIV value, the displacement time history data of each first measuring point in at least one first measuring point in the sample tube gallery are sorted to obtain a second sorting result of the displacement time history data of each first measuring point.

[0027] Determine the displacement time history data that are greater than or equal to a second preset threshold in the second sorting result, and obtain the at least one second training sample based on the displacement time history data that are greater than or equal to the second preset threshold in the second sorting result and the stress time history data.

[0028] In the above scheme, training the second neural network model using each of the at least one second training sample to obtain the preset neural network model includes:

[0029] The second neural network model is trained based on each second training sample to obtain the third neural network model corresponding to each second training sample;

[0030] Determine the training error of each second training sample in the corresponding third neural network model;

[0031] Based on the training error of each second training sample in the corresponding third neural network model, determine the third neural network model corresponding to the minimum value of the training error;

[0032] The third neural network model corresponding to the minimum value of the training error is used as the preset neural network model.

[0033] In the above scheme, obtaining the damaged area to be detected in the pipe gallery includes:

[0034] Obtain stress time history data for at least one second test point in the pipe gallery to be tested;

[0035] A preset algorithm is used to process the stress time history data of each of the at least one second test points to obtain the damage value corresponding to each second test point.

[0036] The location of damage to be detected in the pipe gallery is determined based on the damage value corresponding to each second test point.

[0037] In the above scheme, the step of acquiring displacement time history data of at least one first test point on each cross section of the pipe gallery under test based on a preset time interval includes:

[0038] In the first time period, the first displacement time history data of at least one first test point on each cross section of at least one cross section of the pipe gallery to be tested are acquired;

[0039] After a preset time interval, in the second time period, the second displacement time history data of at least one first test point on each cross section of the pipe gallery to be tested are acquired.

[0040] In the above scheme, the step of inputting the displacement time history data into a preset neural network model to obtain the stress time history data of the damaged part to be detected includes:

[0041] The first displacement time history data is input into a preset neural network model to obtain the first stress time history data of the damaged part to be detected;

[0042] The second displacement time history data is input into a preset neural network model to obtain the second stress time history data of the damaged part to be detected.

[0043] In the above scheme, determining the damage parameters corresponding to the damage site to be detected based on the stress time history data includes:

[0044] The first stress time history data and the second stress time history data are processed by a preset algorithm to obtain the first damage parameter and the second damage parameter corresponding to the damage site to be detected.

[0045] The first damage parameter and the second damage parameter are accumulated to obtain the damage parameter corresponding to the damage site to be detected.

[0046] In the above scheme, determining the degree of damage to the pipe gallery under test based on the damage parameters includes:

[0047] Based on the damage parameters, the safety level of the damaged part to be detected in the pipe gallery to be tested is determined, and the judgment result is obtained; wherein, the safety level represents the safety degree of the damaged part to be detected.

[0048] The degree of damage to the pipe gallery under test is determined based on the judgment result; wherein, the higher the safety level, the lower the degree of damage; the lower the safety level, the higher the degree of damage.

[0049] This invention provides a damage monitoring device, comprising:

[0050] The first acquisition module is used to acquire displacement time history data of at least one first test point in the pipe gallery under test and the damaged parts to be detected in the pipe gallery under test.

[0051] The first prediction module is used to input the displacement time history data into a preset neural network model to obtain the stress time history data of the damage site to be detected.

[0052] The first determining module is used to determine the damage parameters corresponding to the damage site to be detected based on the stress time history data.

[0053] The second determining module is used to determine the degree of damage to the pipe gallery under test based on the damage parameters.

[0054] This invention provides a damage monitoring device, including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the program, it implements any step of the method described above.

[0055] This invention provides a storage medium storing executable instructions, which, when executed by a processor, implement any step of the damage monitoring method described above.

[0056] This invention provides a damage monitoring method, device, electronic device, and storage medium. The method includes: acquiring displacement time history data of at least one first test point in a pipe gallery under test and a damage site to be detected in the pipe gallery; inputting the displacement time history data into a preset neural network model to obtain stress time history data of the damage site to be detected; determining damage parameters corresponding to the damage site to be detected based on the stress time history data; and determining the degree of damage to the pipe gallery under test based on the damage parameters. The technical solution of this invention involves acquiring displacement time history data of a first test point in a utility tunnel and identifying the damaged parts within the tunnel. The displacement time history data is then input into a preset neural network model to obtain stress time history data for the damaged parts. Damage parameters corresponding to the damaged parts are determined based on the stress time history data. Furthermore, the degree of damage to the utility tunnel is determined based on these damage parameters. In essence, by monitoring the displacement of a small number of first test points in the utility tunnel and converting the displacement time history data into stress time history data using a preset neural network model, fatigue damage to key parts of the utility tunnel is calculated based on the stress time history data. This approach addresses the shortcomings of fatigue damage monitoring in the field of integrated utility tunnel operation monitoring, reduces monitoring costs, acquires fatigue damage as a monitoring indicator, improves the safety operation and maintenance system of integrated utility tunnels, and provides more guidance for the operation and maintenance of integrated utility tunnels. Attached Figure Description

[0057] Figure 1 This is a schematic diagram illustrating the implementation process of the damage monitoring method according to an embodiment of the present invention;

[0058] Figure 2This is a schematic diagram of the layout of stress monitoring points in the pipe gallery to be tested in the damage monitoring method of this invention.

[0059] Figure 3 This is a flowchart illustrating the fatigue damage calculation algorithm designed based on three concrete fatigue theories: rainflow counting, Miner's theory, and Cornlissen's formula, in the damage monitoring method of this invention.

[0060] Figure 4 This is a schematic diagram of fatigue-vulnerable parts screened in a certain integrated utility tunnel using the damage monitoring method of this invention.

[0061] Figure 5 This is a schematic diagram of the layout of displacement monitoring nodes at the experimental section of the damage monitoring method according to an embodiment of the present invention;

[0062] Figure 6 This is a schematic diagram of the BP neural network model of the damage monitoring method according to an embodiment of the present invention;

[0063] Figure 7 This is a schematic diagram of the displacement monitoring nodes required after screening a certain utility tunnel according to an embodiment of the present invention;

[0064] Figure 8 This is a flowchart illustrating the fatigue-vulnerable part selection and displacement-stress BP neural network training process of the damage monitoring method according to an embodiment of the present invention.

[0065] Figure 9 This is a schematic diagram of the fatigue damage calculation process during the application stage of the damage monitoring method in an embodiment of the present invention;

[0066] Figure 10 This is a schematic diagram of the composition and structure of the damage monitoring device according to an embodiment of the present invention;

[0067] Figure 11 This is a schematic diagram of the hardware structure of a damage monitoring device in an embodiment of the present invention. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the specific technical solutions of the invention will be further described in detail below with reference to the accompanying drawings of the embodiments of the present invention. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0069] This embodiment proposes a damage monitoring method, which is applied to a damage monitoring device. The function implemented by this method can be achieved by the processor in the damage monitoring device calling program code. Of course, the program code can be stored in a computer storage medium. It can be seen that the computing device includes at least a processor and a storage medium.

[0070] Figure 1This is a schematic diagram illustrating the implementation process of the damage monitoring method according to an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes:

[0071] Step S101: Obtain displacement time history data of at least one first test point in the pipe gallery to be tested and the damaged parts to be detected in the pipe gallery to be tested.

[0072] It should be noted that the damage monitoring method can be determined according to the actual situation and is not limited here. As an example, the damage monitoring method can be a fatigue damage monitoring method for integrated utility tunnels based on back propagation (BP) neural networks. In practical applications, the damage monitoring method can be applied to monitor fatigue damage from vehicles in shallow-buried integrated utility tunnels.

[0073] The utility tunnel to be tested can be selected according to the actual situation and is not limited here. As an example, the utility tunnel to be tested can be an urban underground integrated utility tunnel, which can be a shallow-buried integrated utility tunnel.

[0074] The at least one first test point can be understood as one or more first test points arranged in the pipe gallery to be tested; wherein, the specific number of the first test points can be determined according to the actual situation and is not limited here. As an example, the first test point is equipped with a displacement monitoring node for acquiring displacement time history data of at least one first test point in the pipe gallery to be tested. Multiple displacement monitoring nodes can be set in the pipe gallery to be tested according to a preset distance interval, and any one of the displacement monitoring nodes in the pipe gallery to be tested can be used as the first test point. Alternatively, any one of the displacement monitoring nodes in the pipe gallery to be tested that is of high importance to the output result of the neural network model can be selected as the first test point. The preset distance interval can be determined according to the actual situation and is not limited here. As an example, the preset distance interval can be 0.5m.

[0075] The displacement time history data of the at least one first test point can be obtained through a sensing device. The sensing device can be determined based on actual conditions and is not limited here. As an example, the sensing device can be a distributed optical fiber sensor, which acquires the displacement time history data of the at least one first test point. The arrangement position of the distributed optical fiber sensor can be determined based on actual conditions and is not limited here. As an example, the distributed optical fiber sensor can be arranged on the surface of the pipe gallery structure under test, along the longitudinal direction of the pipe gallery. The distributed optical fiber sensor can measure or monitor the displacement time history data of at least one first test point in the pipe gallery along the optical fiber transmission path. Using distributed optical fiber sensor technology reduces the number of sensors required, enabling continuous measurement of displacement time history data at a lower cost. Simultaneously, arranging the distributed optical fiber sensor on the structural surface avoids pre-embedding the sensors inside the concrete, reducing the technical difficulty and cost of pre-embedding sensors, and also reducing the impact of the sensors on the load-bearing capacity of the concrete structure itself.

[0076] In practical applications, the pipe gallery to be tested may include at least one cross-section, and the structural surface of the at least one cross-section is provided with at least one first test point, and the damaged part to be detected in the pipe gallery to be tested is obtained from the at least one cross-section.

[0077] The displacement time history data includes at least displacement data and time data, and there is a corresponding relationship between the displacement data and the time data. For example, a certain time corresponds to the displacement data of the point to be measured at that time. In practical applications, the displacement data can be obtained by monitoring the displacement data of each first point to be measured at preset time intervals. The preset time can be determined according to the actual situation and is not limited here. As an example, the preset time can be 1 second.

[0078] The method for identifying the damaged areas within the pipe gallery to be tested is as follows: the manner in which these damaged areas cause fatigue damage can be determined based on actual conditions and is not limited here. For example, the damaged areas may be affected by fatigue damage caused by overhead vehicle loads. In practical applications, different vehicle operating conditions applied to the damaged areas will result in different types of fatigue damage.

[0079] Step S102: Input the displacement time history data into a preset neural network model to obtain the stress time history data of the damaged part to be detected.

[0080] It should be noted that the preset neural network model can be obtained by selecting a suitable neural network model for training according to the actual situation, and no limitation is made here. As an example, the preset neural network model can be a trained BP neural network model, and the preset neural network model can be obtained by training the BP neural network model.

[0081] The fitting function obtained by training the preset neural network model can be determined according to the actual situation and is not limited here. As an example, according to structural mechanics theory, a deformation state of a structure corresponds to a unique stress state, and its deformation state can be represented by the displacements of multiple nodals of the structure {d1,d2,...,d...}. n To approximate this, we have: σ = g(d1, d2, ..., d n If the function g:d→σ can be obtained, the structural stress can be obtained by monitoring the nodal displacements. A backpropagation (BP) neural network model algorithm is used to approximate this function. It has been theoretically proven that a BP neural network model with 3 or more layers can fit any function. When the number of training samples m approaches infinity, the function fitted by the BP neural network model becomes... satisfy: By training the preset neural network model, the displacement-stress function relationship of the pipe gallery under test is obtained, resulting in the preset neural network model. The displacement time history data is then converted into stress time history data through the preset neural network model. This significantly reduces the cost of real-time monitoring of stress data at the damaged locations of the pipe gallery under test, thereby reducing the difficulty of calculating fatigue damage in the pipe gallery under test.

[0082] The stress time history data includes at least stress data and time data, and there is a corresponding relationship between the stress data and the time data. For example, a certain time corresponds to the stress data of the test point at that time. In practical applications, the stress data can be monitored at preset time intervals for each second test point. The preset time can be determined according to the actual situation and is not limited here. As an example, the preset time can be 1 second.

[0083] Step S103: Determine the damage parameters corresponding to the damage site to be detected based on the stress time history data.

[0084] It should be noted that the damage parameters can be obtained through theoretical calculations. According to current fatigue theory, only the stress time history data of the structure at the point of damage to be detected is needed to calculate the damage parameters of the structure at that point of damage within that time period, i.e., D(t) = f(σ,t). Based on the above formula, the fatigue damage of the pipe gallery can be indirectly monitored simply by real-time monitoring of the stress at the point of damage to be detected in the pipe gallery. This can be combined with the results fitted by the aforementioned BP neural network model. function Therefore, fatigue damage in the utility tunnel can be expressed as: Therefore, by obtaining displacement time history data at at least one first test point, the damage parameters of the damaged part in the pipe gallery under test can be calculated, that is, fatigue damage can be monitored by monitoring displacement. By performing real-time displacement-stress monitoring on a standard cross-section of the pipe gallery, a BP neural network with displacement as input and stress as output is trained, and this BP neural network is used for displacement-stress conversion of all pipe gallery cross-sections.

[0085] Step S104: Determine the degree of damage to the test tube gallery based on the damage parameters.

[0086] It should be noted that the damage parameters include at least the fatigue damage value. According to relevant fatigue theory, the structure fails when the fatigue damage value is 1. Therefore, the magnitude of the fatigue damage value reflects the degree of damage to the pipe gallery under test. Depending on the degree of damage to the pipe gallery during operation, different operation and maintenance schemes can be designed to improve the service life of the integrated pipe gallery and better ensure its operational safety.

[0087] The fatigue damage value further reflects the changes in structural lifespan. Based on the magnitude of this monitoring index, the lifespan stage of the utility tunnel can be determined. Furthermore, different maintenance plans can be designed according to the lifespan stage of the tunnel, truly achieving sustainable operation and maintenance throughout its entire life cycle, thereby maximizing the service life of the utility tunnel. This not only ensures the safety of the utility tunnel during operation but also fully leverages the role of maintenance, improving the efficiency of its use and reducing social costs.

[0088] This invention provides a damage monitoring method. It acquires displacement time-history data of a first test point in a utility tunnel and identifies the damaged areas within the tunnel. The displacement time-history data is input into a preset neural network model to obtain stress time-history data for the damaged areas. Damage parameters corresponding to the damaged areas are determined based on the stress time-history data. The degree of damage to the utility tunnel is then determined based on these damage parameters. Specifically, by monitoring the displacement of a small number of first test points in the utility tunnel and converting the displacement time-history data into stress time-history data using a preset neural network model, fatigue damage to key components of the utility tunnel is calculated based on the stress time-history data. This method addresses the shortcomings of fatigue damage monitoring in the field of integrated utility tunnel operation monitoring, reduces monitoring costs, acquires fatigue damage as a monitoring indicator, improves the safety operation and maintenance system of integrated utility tunnels, and provides more guidance for the operation and maintenance of integrated utility tunnels.

[0089] In an optional embodiment of the present invention, obtaining the damaged location to be detected in the pipe gallery to be tested includes: obtaining stress time history data of at least one second test point in the pipe gallery to be tested; processing the stress time history data of each second test point in the at least one second test point using a preset algorithm to obtain the damage value corresponding to each second test point; and determining the damaged location to be detected in the pipe gallery to be tested based on the damage value corresponding to each second test point.

[0090] In this embodiment, the location of the second test point can be determined according to the actual situation and is not limited here. As an example, the second test point is equipped with a stress monitoring point for acquiring stress time history data of the second test point in the pipe gallery under test. Multiple stress monitoring points can be set in the pipe gallery under test at preset distance intervals, and the location of any stress monitoring point in the pipe gallery under test can be set as the second test point. Figure 2 This is a schematic diagram of the layout of stress monitoring points in the pipe gallery under test in the damage monitoring method of this invention, as shown in the embodiment of the invention. Figure 2 As shown, taking the mid-span of the top plate of the pipe gallery to be tested as the reference point, a stress monitoring point is arranged every 0.5m along the line connecting the centers of the thickness of the pipe gallery to be tested, and the location of any stress monitoring point in the pipe gallery to be tested is set as the second test point.

[0091] The preset algorithm is designed according to the actual situation and is not limited here. As an example, the preset algorithm includes at least a fatigue damage calculation program algorithm. The fatigue damage calculation program algorithm is designed based on at least one of the three fatigue theories of concrete: rainflow counting method, Miner theory and Cornlissen formula. The fatigue damage calculation program is used to process the stress time history data of each of the at least one second test point to calculate the damage value corresponding to each second test point.

[0092] For ease of understanding, an example is provided here. Figure 3 This is a flowchart illustrating the fatigue damage calculation algorithm designed based on three concrete fatigue theories—rainflow counting, Miner's theory, and Cornlissen's formula—in the damage monitoring method of this invention. Figure 3 As shown, the specific process of this algorithm is as follows:

[0093] (1) Extract the stress cycles at this location using the rainflow counting method. The rainflow counting method is as follows:

[0094] ① Stress curve peak splicing;

[0095] ② Starting from the beginning of the stress curve, select the starting point σ of the raindrop fall. i ;

[0096] ③ Find the next peak point σ j(j＝i+2n,n＝1,2,...) , until σ j >σ i A stress cycle s is obtained. i =[σ i ,σ j ];

[0097] ④ with σ i+1 As the starting point for raindrops, process ③ and ④ are repeated;

[0098] ⑤ Remove the enclosed stress cycles (when the interval formed by the maximum and minimum values ​​of a stress cycle is contained within another larger stress cycle interval, the stress cycle is said to be enclosed, such as stress cycle [4,5] being enclosed by stress cycle [3,5], in which case stress cycle [4,5] should be removed), finally obtaining the stress cycle set S={s i}

[0099] (2) Calculate the stress cycle s for each stress cycle according to the Cornelissen formula. i The fatigue life is given by the following formula:

[0100]

[0101] Where N i For stress cycle s i Fatigue life under the following conditions, σ max With σ min These are stress cycles s i Maximum and minimum stress values, f t This refers to the tensile strength of concrete.

[0102] (3) The fatigue damage of the structure under all stress cycles is calculated according to Miner's theory, and the formula is as follows:

[0103] By using a fatigue damage calculation program designed based on relevant fatigue theory, fatigue damage calculations were performed on the obtained stress time history data, thereby obtaining the fatigue damage of the pipe gallery structure, an indicator that is difficult to monitor directly.

[0104] The damaged parts to be detected in the pipe gallery to be tested are determined according to the actual situation and are not limited here. As an example, the stress time history data of each of the at least one second test points are processed to obtain the damage value corresponding to each second test point. The second test points corresponding to all damage values ​​are taken as the damaged parts to be detected in the pipe gallery to be tested.

[0105] In an optional embodiment of the present invention, determining the damaged location to be detected in the pipe gallery according to the damage value corresponding to each second test point includes: sorting the damage values ​​corresponding to each second test point to obtain a first sorting result of the damage values ​​corresponding to each second test point; determining the second test points corresponding to the damage values ​​greater than or equal to a first preset threshold in the first sorting result; and taking the second test points corresponding to the damage values ​​greater than or equal to the first preset threshold in the sorting result as the damaged locations to be detected in the pipe gallery.

[0106] In this embodiment, the sorting method for the damage values ​​corresponding to each second test point can be determined according to the actual situation and is not limited here. As an example, the damage values ​​corresponding to each second test point are arranged in descending order to obtain the first sorting result of the damage values ​​corresponding to each second test point.

[0107] The first preset threshold can be determined according to the actual situation. The first preset threshold can be the fatigue damage value preset by the pipe gallery to be tested, or it can be the standard value calculated based on the damage value in the first sorting result. It is not limited here. As an example, the standard value calculated based on the damage value in the first sorting result is used as the first preset threshold, and the second test point corresponding to the damage value in the sorting result that is greater than or equal to the standard value calculated based on the damage value in the first sorting result is used as the damage part to be detected in the pipe gallery to be tested.

[0108] The calculation method for the standard value obtained from the damage values ​​in the first sorting result can be determined according to the actual situation. It can be that the damage value corresponding to the preset proportion of the damage values ​​in the first sorting result is used as the standard value, the median of the damage values ​​in the first sorting result is used as the standard value, the average value of the damage values ​​in the first sorting result is used as the standard value, etc. There is no limitation here. As an example, the damage value corresponding to the preset proportion of the damage values ​​in the first sorting result is used as the standard value, and the damage value corresponding to the preset proportion of the damage values ​​in the first sorting result is used as the first preset threshold. The second test point corresponding to the damage value corresponding to the preset proportion of the damage value in the sorting result is greater than or equal to the damage value corresponding to the preset proportion of the damage value in the first sorting result is used as the damage part to be detected in the pipe gallery to be tested.

[0109] The preset proportion of damage values ​​corresponding to the preset proportion of damage values ​​in the first sorting result as the standard value can be determined according to the actual situation. The preset proportion of damage values ​​can be the top 5%, 10%, 15%, ... of the damage values ​​in the first sorting result, and is not limited here. As an example, the damage values ​​corresponding to the top 5% of the damage values ​​in the first sorting result are selected as the first preset threshold, and the second test point corresponding to the damage value corresponding to the top 5% of the damage values ​​in the sorting result is taken as the damage part to be detected in the test tube gallery.

[0110] In this embodiment of the invention, since the degree of fatigue damage varies at different parts of the pipe gallery under vehicle load, only the part with the greatest fatigue damage needs to be monitored. Figure 4 This is a schematic diagram of fatigue-vulnerable parts screened in a certain integrated utility tunnel using the damage monitoring method of this invention, as shown in the embodiment of the invention. Figure 4 As shown, the second test point corresponding to the damage value of the top 5% in the first sorting result is selected from the sorting results, and is taken as the fatigue vulnerable part with greater fatigue damage in the test tube gallery. The fatigue vulnerable part is taken as the damage part to be detected.

[0111] In an optional embodiment of the present invention, the method further includes: acquiring displacement time history data of at least one first measuring point and stress time history data of at least one fatigue-vulnerable part in the sample pipe gallery; using the displacement time history data and the stress time history data as first training samples to train a first neural network model to obtain a second neural network model corresponding to each fatigue-vulnerable part in the at least one fatigue-vulnerable part; processing the first training samples based on the second neural network model to obtain at least one second training sample; and using each of the at least one second training sample to train the second neural network model to obtain the preset neural network model.

[0112] In this embodiment, the sample utility tunnel can be determined according to the actual situation and is not limited here. As an example, one experimental utility tunnel in the comprehensive utility tunnel that needs to be monitored is selected as the sample utility tunnel, and displacement time history data of at least one first measuring point and stress time history data of at least one fatigue-vulnerable part in the experimental utility tunnel of the comprehensive utility tunnel that needs to be monitored are obtained.

[0113] A sample surface can be selected within the sample tube gallery. This sample surface can be an experimental cross-section chosen within the sample tube gallery. At least one first measuring point is set within the experimental cross-section. The position of the first measuring point within the experimental cross-section can be designed according to actual conditions and is not limited here. As an example, the first measuring point is equipped with a displacement monitoring node for acquiring displacement time history data of at least one first measuring point in the sample tube gallery. Multiple displacement monitoring nodes can be set within the experimental cross-section at preset intervals. Figure 5 This is a schematic diagram of the layout of displacement monitoring nodes at the experimental section of the damage monitoring method according to an embodiment of the present invention, as shown below. Figure 5 As shown, a displacement monitoring node is arranged at 0.5m intervals along the inner contour line of an experimental cross section of the sample tube gallery, and the location of any displacement monitoring node in the experimental cross section is set as the first measuring point.

[0114] The load applied to the sample tube gallery can be set according to the actual situation and is not limited here. As an example, various vehicle working conditions are designed to be applied to the road surface above the sample tube gallery, and displacement time history data of at least one first measuring point in the sample tube gallery and stress time history data of at least one fatigue-vulnerable part are acquired in real time.

[0115] The displacement time history data and the stress time history data are used as the first training samples to train the first neural network model. The displacement time history data of at least one first measuring point in the sample pipe gallery are used as the input variables of the first training samples, and the stress time history data of at least one fatigue-vulnerable part is used as the output of the first training samples. The first training samples are divided into training samples, test samples, and verification samples according to a preset sample ratio, and the first neural network model of at least one fatigue-vulnerable part is trained accordingly. The preset sample ratio can be limited according to actual conditions and is not limited here. As an example, the preset sample ratio can be 6:2:2. Figure 6 This is a schematic diagram of the BP neural network model of the damage monitoring method according to an embodiment of the present invention, as shown below. Figure 6 As shown, by continuously adjusting the hidden layers and the number of neurons in each layer of the neural network model, a second neural network model corresponding to each fatigue-prone site is obtained; wherein, the preset sample ratio can be that the ratio of training samples, test samples, and verification samples is 6:2:2, and at least one first neural network model for fatigue-prone site is trained respectively.

[0116] In this embodiment of the invention, since the input variables in the first training sample used by the second neural network model are the displacement time history data monitored by all the first measuring points of the inner contour of an experimental cross section of the sample tube gallery, it is actually unrealistic to monitor the displacement time history data of so many first measuring points at the same time. The number of first measuring points is limited by the monitoring cost. Therefore, the number of input variables in the training sample required by the preset neural network model must be optimized and reduced.

[0117] In an optional embodiment of the present invention, the step of processing the first training sample based on the second neural network model to obtain at least one second training sample includes: processing the displacement time history data in the training sample according to the second neural network model to obtain the average influence value (MIV) corresponding to the displacement data in the training sample; and determining the at least one second training sample based on the MIV value.

[0118] In this embodiment, the MIV value of an input variable reflects the importance of that variable to the output of the neural network model. This embodiment optimizes and reduces the number of input variables based on the MIV values ​​of each node's displacement. The MIV of each input variable (displacement time history data of the displacement monitoring node) in each of the second neural network models is calculated, and its average MIV value is also calculated. mean And according to MIV mean The magnitude is sorted according to the input variables (displacement time history data of displacement monitoring nodes).

[0119] The i-th variable The expression is:

[0120]

[0121] The meanings of each symbol are as follows:

[0122] i — the i-th displacement monitoring node;

[0123] k — the kth displacement sample;

[0124] j — the j-th neural network model, that is, the neural network model used to output the stress of the j-th fatigue-prone part;

[0125] —The k-th sample of the displacement time history data of the i-th displacement monitoring node;

[0126] —The displacement-stress BP neural network model corresponding to the j-th fatigue-vulnerable part;

[0127] t — the total number of second neural network models obtained through training;

[0128] m — the total number of displacement samples;

[0129] n — the total number of input variables.

[0130] It should be noted that, to obtain The process involved two averaging operations. First, the average value of all input variables in the first training sample was calculated to obtain the MIV. i Then, the average value is calculated for all the second neural network models to obtain

[0131] In an optional embodiment of the present invention, determining the at least one second training sample based on the MIV value includes: sorting the displacement time history data of each first measuring point in at least one first measuring point in the sample pipe gallery based on the MIV value to obtain a second sorting result of the displacement time history data of each first measuring point; determining the displacement time history data greater than or equal to a second preset threshold in the second sorting result; and obtaining the at least one second training sample based on the displacement time history data greater than or equal to the second preset threshold in the second sorting result and the stress time history data.

[0132] In this embodiment, the MIV value of an input variable reflects the importance of the input variable to the output of the neural network model. Based on the MIV value, the displacement time history data of each first measuring point in at least one first measuring point in the sample pipe gallery are sorted to obtain a second sorting result of the displacement time history data of each first measuring point, thereby optimizing and reducing the number of input variables.

[0133] The sorting method for the displacement time history data of each first measuring point in at least one first measuring point in the sample pipe gallery based on the MIV value can be determined according to the actual situation and is not limited here. As an example, the MIV values ​​corresponding to the displacement time history data of each first measuring point in at least one first measuring point in the sample pipe gallery are arranged in descending order to obtain the second sorting result of the displacement time history data of each first measuring point.

[0134] The second preset threshold can be determined according to the actual situation and is not limited here. As an example, the first q displacement time history data of each first measuring point according to the second sorting result is used as the second preset threshold; where q = 1, 2, 3...n-1, n represents the total number of displacement time history data of the first measuring point, that is, the total number of input variables. As an example, q can be 1. When q = 1, the first displacement time history data of each first measuring point according to the second sorting result is used as the second preset threshold. The second training sample is obtained based on the first displacement time history data of each first measuring point that is greater than or equal to the second sorting result and the stress time history data. The number of input variables of the second training sample is 1.

[0135] In an optional embodiment of the present invention, training the second neural network model using each of the at least one second training samples to obtain the preset neural network model includes: training the second neural network model based on each second training sample to obtain a third neural network model corresponding to each second training sample; determining the training error of each second training sample in the corresponding third neural network model; determining the third neural network model corresponding to the minimum value of the training errors based on the training errors of each second training sample in the corresponding third neural network model; and using the third neural network model corresponding to the minimum value of the training errors as the preset neural network model.

[0136] In this embodiment, the method for determining the training error of each second training sample in the corresponding third neural network model can be determined according to the actual situation and is not limited here. As an example, when each second training sample is trained in the corresponding third neural network model, the error between the actual stress time history data monitored by the second test point and the predicted stress time history data output by the third neural network model is determined, and the number q of input variables in each second training sample corresponding to each third neural network model is compared with the corresponding training error ε. q Plotting this as a curve, the training error typically decreases first and then increases as the number of input variables increases, indicating an optimal number of input variables that minimizes the training error. Based on the project budget, a suitable number of input variables [q] is selected, and the maximum allowable training error [ε] is determined according to project needs. Then, based on the q-ε corresponding to each third neural network model... q The number of input variables selected for the curve should satisfy both the training error requirement and the limitation on the number of monitoring points. Based on the number of input variables To determine the displacement monitoring nodes selected for the experimental section, see [link / reference]. Figure 7 , Figure 7 This is a schematic diagram of the displacement monitoring nodes required after screening a certain pipe gallery according to an embodiment of the present invention.

[0137] Based on the training error of each second training sample in the corresponding third neural network model, the third neural network model corresponding to the minimum training error is determined. As an example, based on the number of input variables... The minimum training error among the training errors of the second training sample in the corresponding third neural network model is determined, and the third neural network model corresponding to the minimum training error is further determined. The maximum number of input variables in each third neural network model is calculated. The maximum number of input variables q o This refers to the number of displacement monitoring nodes that are of high importance to the output of the neural network model, i.e., the number of the first measurement points. It should be noted that, although the number of input variables for each third neural network model is... They are not the same, but both are included in q. o In the displacement monitoring nodes. The first q o The required displacement monitoring nodes are the displacement monitoring nodes, based on the optimized first q. o The position of the first test point is determined by the position of each displacement monitoring node, and the q is maintained. o Each displacement monitoring node has its plane coordinates fixed on the cross section of the pipe gallery under test. Distributed fiber optic sensors are arranged along the longitudinal direction of the pipe gallery to monitor the displacement time history data of all the displacement monitoring nodes on the pipe gallery cross section.

[0138] In some embodiments, Figure 8 This is a flowchart illustrating the fatigue-vulnerable site selection and displacement-stress BP neural network training process of the damage monitoring method according to an embodiment of the present invention. Figure 8As shown, the fatigue-vulnerable parts screening and displacement-stress BP neural network training process includes at least the following: selecting an experimental cross-section in the sample pipe gallery as the experimental cross-section of the integrated pipe gallery; arranging multiple displacement monitoring nodes for monitoring node displacement and multiple stress monitoring points for monitoring stress at fatigue-damaged parts on the experimental cross-section; acquiring displacement time history data of each node and stress time history data of each fatigue-damaged part when applying various vehicle loads to the experimental cross-section; calculating the fatigue damage of each fatigue-damaged part using a preset algorithm, selecting the top 5% of fatigue-damaged parts as fatigue-vulnerable parts of the integrated pipe gallery, and monitoring the stress time history data of the fatigue-vulnerable parts; training a first BP neural network model for each fatigue-vulnerable part based on the displacement time history data of each node and the stress time history data of each fatigue-vulnerable part, obtaining a second BP neural network model corresponding to each fatigue-vulnerable part; calculating the MIV of the displacement time history data of each node based on the second BP neural network model corresponding to each fatigue-vulnerable part. The values ​​are sorted based on the MIV values ​​of the displacement time history data of each node; based on the sorting results of the MIV values ​​of the displacement time history data of each node, the displacement time history data of the first q nodes are selected, and the second BP neural network model is retrained for each fatigue-prone part to obtain the third BP neural network model corresponding to each fatigue-prone part; it is determined whether the third BP neural network model meets the requirement of selecting an appropriate number of nodes according to the project budget [q], and whether it meets the requirement of determining the maximum allowable error [ε] according to the project needs; if not, the displacement time history data of the first q nodes are reselected, and the third BP neural network model corresponding to each fatigue-prone part is trained; if so, it is determined whether the second BP neural network model corresponding to all fatigue-prone parts has been trained; if not, the second BP neural network model is trained for each fatigue-prone part to obtain the third BP neural network model corresponding to each fatigue-prone part; if so, the maximum value of the number of input variables that meets both the error requirement and the monitoring node number limit is taken. The first q o Each node serves as a required displacement monitoring node.

[0139] In an optional embodiment of the present invention, the step of obtaining displacement time history data of at least one first test point in the pipe gallery to be tested includes: obtaining displacement time history data of at least one first test point on each cross section of at least one cross section in the pipe gallery to be tested at preset time intervals.

[0140] In this embodiment, the preset interval duration may include one day, two days, ..., and is not limited here. As an example, the preset interval duration is one day, and the displacement time history data of at least one first test point on each cross section of the pipe gallery to be tested on the same day are obtained.

[0141] In an optional embodiment of the present invention, the step of acquiring displacement time history data of at least one first test point on each cross section of at least one section of the pipe gallery under test based on a preset time interval includes: acquiring first displacement time history data of at least one first test point on each cross section of at least one section of the pipe gallery under test in a first time period; and acquiring second displacement time history data of at least one first test point on each cross section of at least one section of the pipe gallery under test in a second time period after a preset time interval.

[0142] In this embodiment, the first time period includes at least a first start time and a first end time; the second time period includes at least a second start time and a second end time. A preset time interval is established between the first start time and the second start time, and this preset time interval is the same as the duration of both the first and second time periods. The duration can include one day, two days, etc., and is not limited here. As an example, the preset time interval is the same as the duration of both the first and second time periods, and the duration of both the first and second time periods is one day. At the first start time, first displacement time history data of at least one first test point on each cross-section of at least one section of the pipe gallery under test is acquired on the first day, until the first end time. At the second start time, second displacement time history data of at least one first test point on each cross-section of at least one cross-section of the pipe gallery under test is acquired on the second day, until the second end time.

[0143] In an optional embodiment of the present invention, the step of inputting the displacement time history data into a preset neural network model to obtain stress time history data of the damaged part to be detected includes: inputting the first displacement time history data into the preset neural network model to obtain first stress time history data of the damaged part to be detected; and inputting the second displacement time history data into the preset neural network model to obtain second stress time history data of the damaged part to be detected.

[0144] In this embodiment, the first displacement time history data is input into a preset neural network model corresponding to the at least one damage site to be detected to obtain the first stress time history data of the at least one damage site to be detected; the second displacement time history data is input into the preset neural network model corresponding to the at least one damage site to be detected to obtain the second stress time history data of the at least one damage site to be detected.

[0145] In an optional embodiment of the present invention, determining the damage parameters corresponding to the damage site to be detected based on the stress time history data includes: processing the first stress time history data and the second stress time history data respectively using a preset algorithm to obtain the first damage parameter and the second damage parameter corresponding to the damage site to be detected; and accumulating the first damage parameter and the second damage parameter to obtain the damage parameters corresponding to the damage site to be detected.

[0146] In this embodiment, the preset algorithm is designed according to the actual situation and is not limited here. As an example, the preset algorithm includes at least a fatigue damage calculation program. The fatigue damage calculation program is designed based on at least one of the three concrete fatigue theories: rainflow counting method, Miner theory and Cornlissen formula. The fatigue damage calculation program is used to process the first stress time history data and the second stress time history data to calculate the first damage parameter and the second damage parameter corresponding to the damage site to be detected.

[0147] In an optional embodiment of the present invention, determining the degree of damage to the pipe gallery under test based on the damage parameters includes: determining the safety level of the damaged part to be detected in the pipe gallery under test based on the damage parameters, and obtaining a determination result; wherein, the safety level characterizes the safety level of the damaged part to be detected; determining the degree of damage to the pipe gallery under test according to the determination result; wherein, the higher the safety level, the lower the degree of damage; the lower the safety level, the higher the degree of damage.

[0148] In this embodiment, the damage parameters include at least the fatigue damage value. According to Miner's theory, a structure will fail when the accumulated fatigue damage value is greater than 1. Based on this threshold, the integrated utility tunnel can be divided into three safety levels according to the accumulated fatigue damage value: Excellent: [0-0.3], Good: [0.3-0.7], and Poor: [0.7-1]. The safety level of the damaged part to be detected in the utility tunnel is determined based on the fatigue damage value. A higher safety level indicates a lower degree of damage; a lower safety level indicates a higher degree of damage.

[0149] In some embodiments, Figure 9 This is a schematic diagram of the fatigue damage calculation process during the application stage of the damage monitoring method in this embodiment of the invention, as shown below. Figure 9 As shown, the damage fatigue damage calculation process includes: in the first q... o Each node serves as a required displacement monitoring node, and distributed fiber optic sensors are deployed at these nodes to monitor the cross-section q of each pipe gallery under test. oThe displacement time history data of each node on that day is collected; the displacement time history data is input into a preset displacement-stress BP neural network model, which outputs the stress time history data of each fatigue-prone part of the cross-section of the pipe gallery under test; the fatigue damage D corresponding to the stress time history data of each fatigue-prone part on that day is calculated based on the fatigue damage calculation program. i ; the fatigue damage of the day D i Compared to the previous day's fatigue damage D i-1 The cumulative fatigue damage ∑D of each fatigue-prone part is obtained by summing the data; the safety level of the cumulative fatigue damage ∑D is determined, and different operation and maintenance measures are designed according to different safety levels.

[0150] This embodiment proposes a damage monitoring device. Figure 10 This is a schematic diagram of the composition and structure of the damage monitoring device according to an embodiment of the present invention, as shown below. Figure 10 As shown, the damage monitoring device 200 includes: a first acquisition module 201, a first prediction module 202, a first determination module 203, and a second determination module 204, wherein:

[0151] The first acquisition module 201 is used to acquire displacement time history data of at least one first test point in the pipe gallery under test and the damaged parts to be detected in the pipe gallery under test.

[0152] The first prediction module 202 is used to input the displacement time history data into a preset neural network model to obtain the stress time history data of the damage site to be detected;

[0153] The first determining module 203 is used to determine the damage parameters corresponding to the damage site to be detected based on the stress time history data;

[0154] The second determining module 204 is used to determine the degree of damage to the pipe gallery under test based on the damage parameters.

[0155] In other embodiments, the first acquisition module 201 is further configured to acquire displacement time history data of at least one first test point on each cross section of the pipe gallery to be tested at preset time intervals.

[0156] In other embodiments, the damage monitoring device further includes a third determining module, which is used to sort the damage values ​​corresponding to each second test point to obtain a first sorting result of the damage values ​​corresponding to each second test point; determine the second test points corresponding to the damage values ​​greater than or equal to a first preset threshold in the first sorting result, and take the second test points corresponding to the damage values ​​greater than or equal to the first preset threshold in the sorting result as the damage sites to be detected in the test tunnel.

[0157] In other embodiments, the damage monitoring device further includes a first training module, which is used to acquire displacement time history data of at least one first measuring point in the sample pipe gallery and stress time history data of at least one fatigue-vulnerable part; use the displacement time history data and the stress time history data as first training samples to train a first neural network model to obtain a second neural network model corresponding to each fatigue-vulnerable part in the at least one fatigue-vulnerable part; process the first training samples based on the second neural network model to obtain at least one second training sample; and use each of the at least one second training sample to train the second neural network model to obtain the preset neural network model.

[0158] In other embodiments, the damage monitoring device further includes a fourth determining module, which is used to process the displacement time history data in the training samples according to the second neural network model to obtain the average influence value (MIV) corresponding to the displacement data in the training samples; and to determine the at least one second training sample based on the MIV value.

[0159] In other embodiments, the fourth determining module is further configured to sort the displacement time history data of each first measuring point in at least one first measuring point in the sample tube gallery based on the MIV value, to obtain a second sorting result of the displacement time history data of each first measuring point; determine the displacement time history data in the second sorting result that are greater than or equal to a second preset threshold, and obtain the at least one second training sample based on the displacement time history data in the second sorting result that are greater than or equal to the second preset threshold and the stress time history data.

[0160] In other embodiments, the damage monitoring device further includes a second training module, which is used to train the second neural network model based on each second training sample to obtain a third neural network model corresponding to each second training sample; determine the training error of each second training sample in the corresponding third neural network model; determine the third neural network model corresponding to the minimum value of the training errors based on the training errors of each second training sample in the corresponding third neural network model; and use the third neural network model corresponding to the minimum value of the training errors as a preset neural network model.

[0161] In other embodiments, the first acquisition module 201 is further configured to acquire stress time history data of at least one second test point in the pipe gallery to be tested; process the stress time history data of each second test point in the at least one second test point using a preset algorithm to obtain the damage value corresponding to each second test point; and determine the damage location to be detected in the pipe gallery to be tested based on the damage value corresponding to each second test point.

[0162] In other embodiments, the first acquisition module 201 is further configured to acquire first displacement time history data of at least one first test point on each cross section of at least one cross section in the pipe gallery under test during a first time period; and after a preset time interval, acquire second displacement time history data of at least one first test point on each cross section of at least one cross section in the pipe gallery under test during a second time period.

[0163] In other embodiments, the first prediction module 202 is further configured to input the first displacement time history data into a preset neural network model to obtain the first stress time history data of the damaged part to be detected; and input the second displacement time history data into the preset neural network model to obtain the second stress time history data of the damaged part to be detected.

[0164] In other embodiments, the first determining module 203 is further configured to process the first stress time history data and the second stress time history data respectively using a preset algorithm to obtain the first damage parameter and the second damage parameter corresponding to the damage site to be detected; and to accumulate the first damage parameter and the second damage parameter to obtain the damage parameter corresponding to the damage site to be detected.

[0165] In other embodiments, the first determining module 203 is further configured to determine the safety level of the damaged part to be detected in the pipe gallery to be tested based on the damage parameters, and obtain a determination result; wherein, the safety level characterizes the safety degree of the damaged part to be detected; and determine the degree of damage of the pipe gallery to be tested according to the determination result; wherein, the higher the safety level, the lower the degree of damage; and the lower the safety level, the higher the degree of damage.

[0166] The description of the above device embodiments is similar to that of the above method embodiments, and has similar beneficial effects. For technical details not disclosed in the device embodiments of the present invention, please refer to the description of the method embodiments of the present invention for understanding.

[0167] It should be noted that, in the embodiments of the present invention, if the damage monitoring method described above is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a damage monitoring device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, portable hard drive, read-only memory (ROM), magnetic disk, or optical disk. Thus, the embodiments of the present invention are not limited to any specific hardware and software combination.

[0168] Correspondingly, embodiments of the present invention provide a damage monitoring device, including a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the program, it implements any step of the above-described method.

[0169] Correspondingly, embodiments of the present invention provide a storage medium storing executable instructions, which, when executed by a processor, implement any step of the damage monitoring method described above.

[0170] It should be noted that the descriptions of the storage medium and device embodiments above are similar to those of the method embodiments above, and have similar beneficial effects. For technical details not disclosed in the storage medium and device embodiments of the present invention, please refer to the descriptions of the method embodiments of the present invention for understanding.

[0171] It should be noted that, Figure 11 This is a schematic diagram of a hardware entity structure of a damage monitoring device in an embodiment of the present invention, such as... Figure 11 As shown, the hardware entity of the damage monitoring device 300 includes a processor 301 and a memory 302. Optionally, the damage monitoring device 300 may also include a communication interface 302.

[0172] It is understood that memory 303 can be volatile memory or non-volatile memory, or both. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), ferromagnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM); magnetic surface memory can be disk storage or magnetic tape storage. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), SyncLink Dynamic Random Access Memory (SLDRAM), and Direct Rambus Random Access Memory (DRRAM).The memory 303 described in this embodiment of the invention is intended to include, but is not limited to, these and any other suitable types of memory.

[0173] The methods disclosed in the above embodiments of the present invention can be applied to or implemented by processor 301. Processor 301 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit of the hardware in processor 301 or by instructions in software form. The processor 301 may be a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Processor 301 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or any conventional processor, etc. The steps of the methods disclosed in the embodiments of the present invention can be directly manifested as being executed by a hardware decoding processor, or being executed by a combination of hardware and software modules in the decoding processor. The software modules may be located in a storage medium, which is located in memory 303. Processor 301 reads the information in memory 303 and combines its hardware to complete the steps of the aforementioned method.

[0174] In an exemplary embodiment, the damage monitoring device may be implemented by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), general-purpose processors, controllers, microcontrollers (MCUs), microprocessors, or other electronic components to perform the aforementioned method.

[0175] In the several embodiments provided by this invention, it should be understood that the disclosed methods and apparatus can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components may be combined, or integrated into another observation, or some features may be ignored or not executed. In addition, the communication connections between the various components shown or discussed may be through some interfaces, indirect coupling or communication connections between devices or units, and may be electrical, mechanical, or other forms.

[0176] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0177] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0178] Alternatively, if the integrated units described above in the embodiments of the present invention are implemented as software functional units and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a damage monitoring device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0179] The damage monitoring method, device, and computer storage medium described in this invention are only examples of embodiments of this invention, but are not limited thereto. Any method, device, and computer storage medium related to this damage monitoring is within the protection scope of this invention.

[0180] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of the invention, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the invention. The sequence numbers of the above-described embodiments of the invention are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0181] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0182] The above description is merely an embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A damage monitoring method, characterized in that, The method includes: Obtain displacement time history data of at least one first test point in the pipe gallery under test, as well as the damaged parts to be detected in the pipe gallery under test; The displacement time history data is input into a preset neural network model to obtain the stress time history data of the damaged part to be detected; The damage parameters corresponding to the damage site to be detected are determined based on the stress time history data. The degree of damage to the tested pipe gallery is determined based on the damage parameters. The step of acquiring the damaged area to be detected in the pipe gallery includes: Obtain stress time history data for at least one second test point in the pipe gallery to be tested; A preset algorithm is used to process the stress time history data of each of the at least one second test points to obtain the damage value corresponding to each second test point. The location of damage to be detected in the pipe gallery is determined based on the damage value corresponding to each second test point.

2. The method according to claim 1, characterized in that, The step of determining the damaged location to be detected in the pipe gallery based on the damage value corresponding to each second test point includes: The damage values ​​corresponding to each second test point are sorted to obtain a first sorting result of the damage values ​​corresponding to each second test point; Identify the second test point corresponding to the damage value greater than or equal to the first preset threshold in the first sorting result, and use the second test point corresponding to the damage value greater than or equal to the first preset threshold in the sorting result as the damage location to be detected in the pipe gallery to be tested.

3. The method according to claim 1, characterized in that, The method further includes: Obtain displacement time history data of at least one first measuring point and stress time history data of at least one fatigue-vulnerable part in the sample pipe gallery; The displacement time history data and the stress time history data are used as the first training samples to train the first neural network model, thereby obtaining the second neural network model corresponding to each fatigue-vulnerable part in the at least one fatigue-vulnerable part. The first training sample is processed based on the second neural network model to obtain at least one second training sample; The second neural network model is trained using each of the at least one second training sample to obtain the preset neural network model.

4. The method according to claim 3, characterized in that, The step of processing the first training sample based on the second neural network model to obtain at least one second training sample includes: The displacement time history data in the training samples are processed according to the second neural network model to obtain the average influence value (MIV) corresponding to the displacement data in the training samples; The at least one second training sample is determined based on the MIV value.

5. The method according to claim 4, characterized in that, Determining the at least one second training sample based on the MIV value includes: Based on the MIV value, the displacement time history data of each first measuring point in at least one first measuring point in the sample tube gallery are sorted to obtain a second sorting result of the displacement time history data of each first measuring point. Determine the displacement time history data that are greater than or equal to a second preset threshold in the second sorting result, and obtain the at least one second training sample based on the displacement time history data that are greater than or equal to the second preset threshold in the second sorting result and the stress time history data.

6. The method according to claim 3, characterized in that, The step of training the second neural network model using each of the at least one second training sample to obtain the preset neural network model includes: The second neural network model is trained based on each second training sample to obtain the third neural network model corresponding to each second training sample; Determine the training error of each second training sample in the corresponding third neural network model; Based on the training error of each second training sample in the corresponding third neural network model, determine the third neural network model corresponding to the minimum value of the training error; The third neural network model corresponding to the minimum value of the training error is used as the preset neural network model.

7. The method according to claim 1, characterized in that, The acquisition of displacement time history data of at least one first test point in the pipe gallery to be tested includes: At preset time intervals, displacement time history data of at least one first test point on each cross section of the pipe gallery under test are acquired.

8. The method according to claim 7, characterized in that, The step of acquiring displacement time history data of at least one first test point on each cross section of the pipe gallery under test, based on a preset time interval, includes: In the first time period, the first displacement time history data of at least one first test point on each cross section of at least one cross section of the pipe gallery to be tested are acquired; After a preset time interval, in the second time period, the second displacement time history data of at least one first test point on each cross section of the pipe gallery to be tested are acquired.

9. The method according to claim 8, characterized in that, The step of inputting the displacement time history data into a preset neural network model to obtain the stress time history data of the damaged area to be detected includes: The first displacement time history data is input into a preset neural network model to obtain the first stress time history data of the damaged part to be detected; The second displacement time history data is input into a preset neural network model to obtain the second stress time history data of the damaged part to be detected.

10. The method according to claim 9, characterized in that, The step of determining the damage parameters corresponding to the damage site to be detected based on the stress time history data includes: The first stress time history data and the second stress time history data are processed by a preset algorithm to obtain the first damage parameter and the second damage parameter corresponding to the damage site to be detected. The first damage parameter and the second damage parameter are accumulated to obtain the damage parameter corresponding to the damage site to be detected.

11. The method according to claim 1, characterized in that, Determining the degree of damage to the tested pipe gallery based on the damage parameters includes: Based on the damage parameters, the safety level of the damaged part to be detected in the pipe gallery to be tested is determined, and the judgment result is obtained; wherein, the safety level represents the safety degree of the damaged part to be detected. The degree of damage to the pipe gallery under test is determined based on the judgment result; wherein, the higher the safety level, the lower the degree of damage; the lower the safety level, the higher the degree of damage.

12. A damage monitoring device, characterized in that, include: The first acquisition module is used to acquire displacement time history data of at least one first test point in the pipe gallery under test and the damaged parts to be detected in the pipe gallery under test. The first prediction module is used to input the displacement time history data into a preset neural network model to obtain the stress time history data of the damage site to be detected. The first determining module is used to determine the damage parameters corresponding to the damage site to be detected based on the stress time history data. The second determining module is used to determine the degree of damage to the pipe gallery under test based on the damage parameters; The first acquisition module is further configured to acquire stress time history data of at least one second test point in the pipe gallery to be tested; process the stress time history data of each second test point in the at least one second test point using a preset algorithm to obtain the damage value corresponding to each second test point; and determine the damage location to be detected in the pipe gallery to be tested based on the damage value corresponding to each second test point.

13. A damage monitoring device, characterized in that, The method includes a memory and a processor, the memory storing a computer program executable on the processor, characterized in that, when the processor executes the program, it implements the method according to any one of claims 1 to 11.

14. A storage medium, characterized in that, The storage medium stores executable instructions, which, when executed by a processor, implement the damage monitoring method according to any one of claims 1 to 11.

Images