Tunnel deformation maximum value prediction method, device, storage medium and program product

Abstract

The application discloses a tunnel deformation maximum value prediction method and device, a storage medium and a program product, and relates to the technical field of tunnel engineering. The method comprises the following steps: obtaining a historical cross section deformation monitoring time sequence set based on original deformation monitoring data of each cross section monitoring point which has been completed monitoring; obtaining deformation monitoring data of a to-be-tested cross section which is being monitored and has deformation, and extracting a short-term deformation monitoring sequence of the to-be-tested cross section before a current first preset time; obtaining an ending index of a best matching subsequence of a most similar reference cross section corresponding to a minimum DTW distance; and predicting a deformation maximum value of the to-be-tested cross section based on the ending index of the best matching subsequence of the most similar reference cross section and a subsequent data trend thereof. The application realizes accurate prediction of a future deformation maximum value of a to-be-tested cross section which is being monitored.

Description

Technical Field

[0001] This invention relates to the field of tunnel engineering technology, specifically to a method, device, storage medium, and program product for predicting the maximum value of tunnel deformation. Background Technology

[0002] After tunnel excavation, surrounding rock deformation is a key indicator for assessing tunnel structural stability and construction safety. The amount of deformation evolves over time, exhibiting complex nonlinear and multi-stage characteristics, influenced by multiple factors including geological conditions, construction loads, and environmental factors. Real-time monitoring of tunnel deformation and accurate prediction of its potential maximum value are crucial for setting timely warning thresholds, identifying potential risks in advance, and ensuring construction and operational safety.

[0003] Traditional methods for predicting tunnel deformation, such as numerical simulation and empirical regression models, often require detailed geological and mechanical parameters. Furthermore, their generalization ability and prediction accuracy may be limited when processing large-scale, high-frequency monitoring data. In addition, actual monitoring data often suffers from various problems, such as irregular sampling (multiple or missing data per day), incomplete data (insufficient monitoring period), and outliers, all of which pose challenges to prediction. Current research largely focuses on time-series prediction of deformation at a specific future moment or analysis of deformation patterns under different working conditions, while direct prediction methods for the key indicator of the "maximum value that tunnel deformation may reach" remain insufficient.

[0004] Those skilled in the art urgently need a method that can fully utilize historical monitoring data, effectively address data quality issues, and accurately predict the future maximum deformation of the currently monitored cross-section by analyzing the similarity of deformation patterns of different tunnel sections, in order to provide a more reliable basis for risk assessment and decision-making in tunnel engineering. Summary of the Invention

[0005] This invention addresses the shortcomings of current methods for directly predicting the key indicator of tunnel deformation's "maximum potential value." It provides a method, device, storage medium, and program product for predicting the maximum value of tunnel deformation, enabling accurate prediction of the future maximum deformation value of the currently monitored cross-section. This provides a more reliable basis for risk assessment and decision-making in tunnel engineering.

[0006] The present invention is achieved through the following technical solution.

[0007] In a first aspect, the present invention provides a method for predicting the maximum deformation of a tunnel. The method includes: acquiring a set of historical cross-sectional deformation monitoring time series based on the original deformation monitoring data of each cross-sectional measuring point that has been monitored, wherein the monitoring data includes monitoring time, monitored cross-section, and cross-sectional deformation amount; acquiring deformation monitoring data of a cross-section under test that is currently being monitored and has deformation, and extracting a short-term deformation monitoring sequence of the cross-section under test before the current first preset time; acquiring all historical cross-sections based on the set of historical cross-sectional deformation monitoring time series and using them as reference cross-sections, and calculating the dynamic time warping (DTW) distance between multiple historical deformation monitoring sub-sequences of a first length in the historical cross-sectional deformation monitoring time series of each reference cross-section and the short-term deformation monitoring sequence of the cross-section under test, to obtain the minimum DTW distance and the end index of the best matching sub-sequence of the most similar reference cross-section corresponding to the minimum DTW distance; and predicting the maximum deformation of the cross-section under test based on the end index of the best matching sub-sequence of the most similar reference cross-section and its subsequent data trend.

[0008] In some embodiments, before obtaining a set of historical cross-sectional deformation monitoring time series based on the original deformation monitoring data of each cross-sectional measuring point that has been monitored, the method further includes preprocessing the original deformation monitoring data, wherein the preprocessing includes: aggregating the original deformation monitoring data and interpolating the data series with missing dates, wherein the aggregation is to take the average value of multiple data; and performing outlier detection and interpolation repair on the data series after difference processing to obtain a cleaned data series.

[0009] In some embodiments, outlier detection and interpolation repair are performed on the data sequence after difference processing to obtain a cleaned data sequence, including: performing moving average smoothing on the deformation monitoring time series in the historical cross-section deformation monitoring time series set; calculating the standard score Z-score and interquartile range (IQR) based on the smoothed data sequence; identifying data exceeding a preset Z-score threshold or preset IQR upper and lower limits as outliers; and interpolating and replacing the identified outliers using the average of the data before and after them.

[0010] In some embodiments, before calculating the Dynamic Time Warped (DTW) distance between a plurality of historical deformation monitoring subsequences of a first length in the historical deformation monitoring time series of each reference section and the short-term deformation monitoring sequence of the section to be tested, the method further includes: normalizing each historical deformation monitoring subsequence of the reference section based on the maximum and minimum values ​​of each historical deformation monitoring subsequence of the first length in the historical deformation monitoring time series of the reference section; and normalizing the deformation monitoring time series of the section to be tested based on the maximum and minimum values ​​of the short-term deformation monitoring time series of the section to be tested.

[0011] In some embodiments, predicting the maximum deformation of the section under test based on the end index of the best matching subsequence of the most similar reference section and its subsequent data trend includes: obtaining the data after the end index of the best matching subsequence of the most similar reference section; determining the maximum value in the data from the end index of the best matching subsequence of the most similar reference section to its subsequent values; obtaining the true value of the last data in the short-term deformation monitoring time series of the section under test; and predicting the maximum deformation of the section under test using a proportional increment method based on the reference section value of the most similar reference section at the end index of the best matching subsequence, the subsequent maximum value of the most similar reference section, and the true value of the last data in the short-term deformation monitoring time series of the section under test.

[0012] In some embodiments, the proportional increment method includes: when the reference section value of the most similar reference section at the end index of the best matching subsequence is not zero, the predicted maximum deformation value = the true value of the section to be tested * (the subsequent maximum value of the most similar reference section / the reference section value of the most similar reference section at the end index of the best matching subsequence); when the reference section value of the most similar reference section at the end index of the best matching subsequence is zero and the subsequent maximum value of the most similar reference section is greater than zero, the predicted maximum deformation value = the true value of the section to be tested + the subsequent maximum value of the most similar reference section; or when the reference section value of the most similar reference section at the end index of the best matching subsequence is zero and the subsequent maximum value of the most similar reference section is zero, the predicted maximum deformation value = the true value of the section to be tested.

[0013] In some embodiments, a preset window size is used when calculating the DTW distance to allow for local flexibility in the alignment of time series.

[0014] In a second aspect, the present invention provides a tunnel deformation maximum prediction device, the device comprising: at least one processor; and at least one memory coupled to the at least one processor and storing instructions for execution by the at least one processor, the instructions being executed by the at least one processor in accordance with the method described above.

[0015] Thirdly, the present invention provides a computer-readable storage medium storing instructions that, when executed by a computer, cause the computer to perform the methods described above.

[0016] Thirdly, the present invention provides a computer program product, the computer program product including instructions, which, when executed by a computer, cause the computer to perform the methods described above.

[0017] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0018] 1. High prediction accuracy and practicality: By capturing the inherent similarity of time series through the DTW algorithm, historical patterns can be identified more accurately, thereby effectively predicting the maximum value of tunnel deformation and providing key information for risk warning in tunnel engineering;

[0019] 2. Effectively addressing data quality issues: The introduced preliminary data processing steps can effectively handle problems such as multiple data points within a day and missing data in the raw monitoring data, thereby improving the reliability of data input;

[0020] 3. No complex physical parameters required: It mainly relies on pattern recognition of monitoring data, eliminating the need to acquire a large number of complex geomechanical parameters in advance, thus simplifying the model building process;

[0021] 4. Strong robustness: Data preprocessing (including outlier detection and repair) effectively addresses outliers and noise in the monitoring data, improving the model's stability and resistance to uncertain data;

[0022] 5. Wide adaptability: For tunnels with similar geological conditions and construction methods, it can effectively utilize existing monitoring data to achieve knowledge transfer between different cross sections. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as limiting the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a method for predicting the maximum deformation of a tunnel according to an embodiment of the present invention.

[0025] Figure 2 This is a schematic diagram comparing the predicted and actual values ​​of the maximum deformation of 50 cross sections according to an embodiment of the present invention.

[0026] Figure 3 This is a structural block diagram of a tunnel deformation maximum prediction device according to an embodiment of the present invention.

[0027] Figure 4 This is a schematic diagram of a tunnel deformation maximum prediction device according to an embodiment of the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments and accompanying drawings. The illustrative embodiments and descriptions of this invention are for illustrative purposes only and are not intended to limit the invention.

[0029] On the one hand, the present invention provides a method for predicting the maximum value of tunnel deformation. Figure 1 This is a flowchart of a method for predicting the maximum deformation of a tunnel according to an embodiment of the present invention. (Reference) Figure 1 The method for predicting the maximum tunnel deformation includes S10 to S40.

[0030] In S10, based on the original deformation monitoring data of each cross-section measuring point that has been monitored, a set of historical cross-section deformation monitoring time (data) sequences is obtained. The monitoring data includes monitoring time, monitored cross-section, and cross-section deformation.

[0031] In S20, deformation monitoring data of the test section that is being monitored and is deformed is acquired, and the short-term deformation monitoring sequence of the test section before the current first preset time is extracted.

[0032] In S30, all historical cross sections are obtained based on the set of historical cross section deformation monitoring time series and used as reference cross sections. The dynamic time warping (DTW) distance between multiple historical deformation monitoring subsequences with a first length in the historical cross section deformation monitoring time series of each reference cross section and the short-term deformation monitoring sequence of the cross section to be tested is calculated to obtain the minimum DTW distance and the end index of the best matching subsequence of the most similar reference cross section corresponding to the minimum DTW distance.

[0033] In S40, the maximum deformation value of the cross section to be tested is predicted based on the end index of the best matching subsequence of the most similar reference cross section and its subsequent data trend.

[0034] The following describes S10 to S40 in detail with reference to the accompanying drawings.

[0035] The application scenario of this invention is the prediction of the maximum value of tunnel deformation monitoring data.

[0036] In S10, before obtaining the historical cross-sectional deformation monitoring time series based on the original deformation monitoring data of each cross-sectional measuring point that has been monitored, the method also includes preprocessing the original deformation monitoring data. This preprocessing includes: S101 and S102.

[0037] In step S101, the raw deformation monitoring data is aggregated, and data sequences with missing dates are interpolated. For each historical cross-section, the average value of the raw monitoring data collected daily (if more than one record) is first calculated to obtain the representative deformation value for each day. If there is no monitoring data for a certain day, linear interpolation of the adjacent monitoring data is used to fill in the gaps, ensuring that data is recorded every day.

[0038] In step S102, outlier detection and interpolation repair are performed on the differenced data sequence to obtain a cleaned data sequence. Outlier detection and repair specifically include steps A1 to A4.

[0039] A1: Perform a moving average smoothing process on the deformation monitoring time series in the historical cross-section deformation monitoring time series set (the window size can be set to 7 days to ensure the smoothness of the series).

[0040] A2: Calculate the Z-score (Standard Score) and IQR (Interquartile Range) statistics based on the smoothed data sequence (the Z-score threshold can be set to 3.0, and the IQR factor can be set to 1.5).

[0041] A3: Identify data points that exceed the preset Z-score threshold or the preset IQR upper and lower limits as outliers.

[0042] A4: For identified outliers, use the average of the data points before and after them for interpolation to replace them, in order to ensure the smoothness and continuity of the sequence.

[0043] In S20, the test section currently under monitoring, with only early deformation monitoring data available, is acquired, and its short-term post-cleaning deformation monitoring sequence for the previous n days (first preset time) is extracted. The test section has its previous n=30 days of cleaned deformation monitoring sequence as the current mode input. This short-term sequence represents the recent deformation behavior pattern of the test section.

[0044] In S30, all reference cross-sections are obtained from the historical cross-section deformation monitoring time series set. For each reference cross-section, local maximum-minimum normalization is performed on all subsequences of length n (first length) in both the short-term cleaned deformation monitoring time series of the cross-section to be tested and the complete cleaned deformation monitoring time series of the reference cross-section. Then, the Dynamic Time Warping (DTW) algorithm is used to calculate the DTW distance, and the minimum DTW distance and the ending index of the corresponding best-matching subsequence of the reference cross-section are determined. Dynamic Time Warping (DTW) is an algorithm used to measure the similarity between two time series. Its core idea is to achieve nonlinear alignment through dynamic programming to solve the sequence matching problem caused by differences in time scales (such as speech rate changes, signal compression / stretching).

[0045] For example, for each reference section in the historical cross-section deformation monitoring time series set, the following operation is performed: from the complete (cleaned and unnormalized) sequence of the reference section, all subsequences of length 30 are extracted one by one (e.g., from day 1 to day 30, from day 2 to day 31, and so on).

[0046] Local maximum-minimum normalization was performed on the 30-day sequence of the test section, that is, normalization was performed based on its own maximum and minimum values ​​in the previous 30 days. Local maximum-minimum normalization was also performed on the reference subsequence of each cut reference section (which has a historical deformation monitoring subsequence of length 30), that is, normalization was performed based on its own minimum and maximum values.

[0047] For the normalized 30-day sequence before the test fracture and the normalized reference subsequence, the DTW distance between them is calculated using the Dynamic Time Warping (DTW) algorithm. In this embodiment, the window size is set to 10 during DTW calculation to allow for some local stretching or compression of the time series on the time axis. That is, a preset window size is used when calculating the DTW distance to allow for some local flexibility in the alignment of the time series. In other words, there is a step of setting the window size in the DTW calculation process. Too large or too small a window size will lead to inaccurate matching data. The set window size ensures that even if the events (such as deformation acceleration and deceleration points) of the two time series occur at slightly different times, the algorithm can still find the best correspondence (i.e., the best path).

[0048] Record the minimum DTW distance among all calculation results, and the end index of the reference subsequence corresponding to the minimum distance in the complete sequence of the original reference section.

[0049] Based on the minimum DTW distance, the reference section that is most similar to the short-term pattern of the corresponding test section is determined, and its corresponding best matching subsequence end index is determined.

[0050] In this embodiment, by comparing the minimum DTW distance between all reference sections and the short-term patterns of the section under test, a reference section with the smallest DTW distance is finally determined as the optimal similarity reference section. Simultaneously, the end index of the optimal subsequence in this optimal similarity reference section that matches the short-term pattern of the section under test is obtained.

[0051] In S40, the prediction process includes: B1 to B4.

[0052] B1: Obtain the cleaned, unstandardized data of the most similar reference section after (including) the index of the best matching subsequence.

[0053] B2: Determine the maximum value in the data from the end index of the best-matching subsequence of the most similar reference section to its subsequent index, denoted as... .

[0054] B3: Obtain the true value of the last data point in the short-term deformation monitoring sequence (first 30 days sequence) of the cross-section under test, denoted as... .

[0055] B4: The reference section value at the end index of the best-matching subsequence based on the most similar reference section, the subsequent maximum value of the reference section, and the true value of the last data in the short-term deformation monitoring time series of the section under test (denoted as...). The maximum deformation of the test section is predicted using the proportional increment method. :

[0056] when When the value is not zero, predict the maximum value. ;

[0057] when It is zero, and When greater than 0, predict the maximum value. ;

[0058] like and If all are zero, then .

[0059] To improve the robustness of the prediction, the final predicted deformation maximum value is... It must be ensured that it is not lower than the current maximum value of the short-term sequence of the section under test, i.e. .in, This represents the current maximum value of the short-term sequence of the section to be tested.

[0060] To evaluate the predictive performance of this invention, leave-one-out cross-validation can be used for testing. In this evaluation scenario, using the 50 historical cross-sectional data points, each cross-section in the historical dataset is sequentially extracted as a "test cross-section," and its historical data from the previous n days is used as input into the above method for prediction. The complete true maximum value of this cross-section is used as the true value to evaluate the prediction error. All remaining historical cross-sections are used as a reference cross-section set. After completing this iterative evaluation for all historical cross-sections, the mean absolute error and standard deviation between the prediction result and the true maximum value of each "test cross-section" are calculated to objectively reflect the generalization ability and prediction accuracy of this method.

[0061] The prediction result of this embodiment is as follows: Figure 2 As shown. After testing and evaluation on 50 cross-sectional data points, the method of this invention can generate effective predictions for the vast majority of cross-sections. The evaluated R-squared (coefficient of determination) of the prediction is 0.992. The evaluated average absolute error of prediction per effective experiment is 6.01 mm. The evaluated standard deviation of the absolute error of prediction is 2.92 mm.

[0062] This method demonstrates good performance in handling complex preprocessing of time series data, similarity matching, and maximum value trend inference, verifying its effectiveness and accuracy in predicting the maximum value of tunnel deformation.

[0063] The present invention has the following beneficial technical effects:

[0064] 1. High prediction accuracy and practicality: By capturing the inherent similarity of time series through the DTW algorithm, historical patterns can be identified more accurately, thereby effectively predicting the maximum value of tunnel deformation and providing key information for risk warning in tunnel engineering.

[0065] 2. Effectively address data quality issues: The introduced preliminary data processing steps can effectively handle problems such as multiple data points within a day and missing data in the original monitoring data, thereby improving the reliability of data input.

[0066] 3. No complex physical parameters required: It mainly relies on pattern recognition of monitoring data, eliminating the need to acquire a large number of complex geomechanical parameters in advance, thus simplifying the model building process.

[0067] 4. Strong robustness: Data preprocessing (including outlier detection and repair) effectively addresses outliers and noise in the monitoring data, improving the model's stability and resistance to uncertain data.

[0068] 5. Wide adaptability: For tunnels with similar geological conditions and construction methods, it can effectively utilize existing monitoring data to achieve knowledge transfer between different cross sections.

[0069] The foregoing primarily describes the solutions of the embodiments of this disclosure from a methodological perspective. It is understood that, in order to achieve the aforementioned functions, the tunnel deformation maximum prediction device includes at least one of the hardware structures and software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this disclosure can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of this disclosure.

[0070] This disclosure embodiment can divide the tunnel deformation maximum value prediction device into functional modules according to the above method embodiment. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one functional module. The integrated module can be implemented in hardware or software. It should be noted that the module division in this disclosure embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods. The following description uses the example of dividing each functional module according to each function.

[0071] On the other hand, the present invention provides a device for predicting the maximum value of tunnel deformation. Figure 3 This is a structural block diagram of a tunnel deformation maximum prediction device according to an embodiment of the present invention. (Reference) Figure 3 The tunnel deformation maximum prediction device includes: a historical data acquisition module, a short-term data acquisition module, a minimum DTW distance calculation module, and a deformation maximum calculation module.

[0072] The historical data acquisition module is used to: acquire a set of historical cross-sectional deformation monitoring time series based on the original deformation monitoring data of each cross-sectional measuring point that has been monitored. The monitoring data includes the monitoring time, the monitored cross-section, and the cross-sectional deformation.

[0073] The short-term data acquisition module is used to: acquire deformation monitoring data of the test section that is being monitored and is deformed, and extract the short-term deformation monitoring sequence of the test section before the current first preset time.

[0074] The minimum DTW distance calculation module is used to: obtain all historical cross sections based on the historical cross section deformation monitoring time series set and use them as reference cross sections, and calculate the dynamic time warping (DTW) distance between multiple historical deformation monitoring subsequences with a first length in the historical cross section deformation monitoring time series of each reference cross section and the short-term deformation monitoring sequence of the cross section to be tested, so as to obtain the minimum DTW distance and the end index of the best matching subsequence of the most similar reference cross section corresponding to the minimum DTW distance.

[0075] The maximum deformation calculation module is used to predict the maximum deformation of the section under test based on the end index of the best matching subsequence of the most similar reference section and its subsequent data trend.

[0076] In some embodiments, before obtaining the historical cross-sectional deformation monitoring time series set based on the original deformation monitoring data of each cross-sectional measuring point that has been monitored, the historical data acquisition module is further configured to: preprocess the original deformation monitoring data. The preprocessing includes: aggregating the original deformation monitoring data and interpolating the data series with missing dates, wherein the aggregation is an average of multiple data points; and performing outlier detection and interpolation repair on the data series after interpolation processing to obtain a cleaned data series.

[0077] In some embodiments, outlier detection and interpolation repair are performed on the data sequence after difference processing to obtain a cleaned data sequence, including: performing moving average smoothing on the deformation monitoring time series in the historical cross-section deformation monitoring time series set; calculating the standard score Z-score and interquartile range (IQR) based on the smoothed data sequence; identifying data exceeding the preset Z-score threshold or preset IQR upper and lower limits as outliers; and interpolating and replacing the identified outliers with the average of the data before and after them.

[0078] In some embodiments, before calculating the dynamic time warping (DTW) distance between multiple historical deformation monitoring subsequences of a first length in the historical deformation monitoring time series of each reference section and the short-term deformation monitoring sequence of the section to be tested, the minimum DTW distance calculation module is further configured to: normalize each historical deformation monitoring subsequence of the reference section based on the maximum and minimum values ​​of each historical deformation monitoring subsequence of the first length in the historical deformation monitoring time series of the reference section; and normalize the deformation monitoring time series of the section to be tested based on the maximum and minimum values ​​of the short-term deformation monitoring time series of the section to be tested.

[0079] In some embodiments, when predicting the maximum deformation of the section under test based on the end index of the best matching subsequence of the most similar reference section and its subsequent data trend, the maximum deformation calculation module is further configured to: obtain the data after the end index of the best matching subsequence of the most similar reference section; determine the maximum value in the data from the end index of the best matching subsequence of the most similar reference section to its subsequent values; obtain the true value of the last data in the short-term deformation monitoring time series of the section under test; and predict the maximum deformation of the section under test using the proportional increment method based on the reference section value at the end index of the best matching subsequence of the most similar reference section, the subsequent maximum value of the reference section, and the true value of the last data in the short-term deformation monitoring time series of the section under test.

[0080] The proportional increment method includes: when the reference section value at the end index of the best matching subsequence of the most similar reference section is not zero, the predicted maximum deformation value = the true value of the section to be tested * (the subsequent maximum value of the most similar reference section / the reference section value at the end index of the best matching subsequence of the most similar reference section); when the reference section value at the end index of the best matching subsequence of the most similar reference section is zero and the subsequent maximum value of the most similar reference section is greater than zero, the predicted maximum deformation value = the true value of the section to be tested + the subsequent maximum value of the most similar reference section; or when the reference section value at the end index of the best matching subsequence of the most similar reference section is zero and the subsequent maximum value of the most similar reference section is zero, the predicted maximum deformation value = the true value of the section to be tested.

[0081] In some embodiments, a preset window size is used when calculating the DTW distance to allow for local flexibility in the alignment of time series.

[0082] In implementing the functions of the integrated modules described above in hardware, this disclosure provides a structure for the tunnel deformation maximum value prediction device involved in the above embodiments. Figure 4 This is a schematic diagram of a tunnel deformation maximum prediction device according to an embodiment of the present invention. (Reference) Figure 4 The tunnel deformation maximum prediction device includes: at least one processor; and at least one memory. The at least one memory is coupled to the at least one processor and stores instructions for execution by the at least one processor, which, when executed by the at least one processor, implement the method described above.

[0083] A processor can be a set of various exemplary logic blocks, modules, and circuits that implement or execute the embodiments described in connection with this disclosure. The processor can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in connection with the embodiments of this disclosure. A processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, etc.

[0084] The memory may be read-only memory (ROM) or other types of static storage devices that can store static information and instructions, random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, electrically erasable programmable read-only memory (EEPROM), disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0085] In one implementation, the memory can exist independently of the processor. The memory can be connected to the processor via a bus and used to store instructions or program code. When the processor calls and executes the instructions or program code stored in the memory, it can implement the methods provided in the embodiments of this disclosure. In another implementation, the memory can also be integrated with the processor.

[0086] On the other hand, the present invention also provides a computer-readable storage medium (e.g., a non-transitory computer-readable storage medium) storing computer program instructions that, when executed on a computer, cause the computer to perform the method as described in any of the above embodiments.

[0087] Exemplary examples show that the aforementioned computer-readable storage media may include, but are not limited to: magnetic storage devices (e.g., hard disks, floppy disks, or magnetic tapes), optical discs (e.g., compact disks (CDs), digital versatile disks (DVDs), etc.), smart cards, and flash memory devices (e.g., erasable programmable read-only memory (EPROMs), cards, sticks, or key drives, etc.). The various computer-readable storage media described in this disclosure may represent one or more devices for storing information and / or other machine-readable storage media. The term "machine-readable storage media" may include, but is not limited to, wireless channels and various other media capable of storing, containing, and / or carrying instructions and / or data.

[0088] This disclosure provides a computer program that, when run on a computer, causes the computer to perform the method of any of the above embodiments.

[0089] This disclosure provides a computer program product containing instructions that, when run on a computer, cause the computer to perform the method of any of the embodiments described above.

[0090] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A tunnel deformation maximum prediction method characterized by, The method comprises: obtaining a historical cross-section deformation monitoring time sequence set based on original deformation monitoring data of each cross-section monitoring point of completed monitoring, wherein the monitoring data comprises monitoring time, monitoring cross-section and cross-section deformation amount; obtaining deformation monitoring data of a to-be-tested cross-section being monitored and deformed, and extracting a short-term deformation monitoring sequence of the to-be-tested cross-section before a current first preset time; obtaining all historical cross-sections as reference cross-sections based on the historical cross-section deformation monitoring time sequence set, and calculating a dynamic time warping (DTW) distance between a plurality of historical deformation monitoring subsequences with a first length in a historical cross-section deformation monitoring time sequence of each reference cross-section and the short-term deformation monitoring sequence of the to-be-tested cross-section, to obtain a minimum DTW distance and an ending index of a best matching subsequence of a most similar reference cross-section corresponding to the minimum DTW distance; and predicting a deformation maximum value of the to-be-tested cross-section based on the ending index of the best matching subsequence of the most similar reference cross-section and a subsequent data trend thereof, wherein predicting the deformation maximum value of the to-be-tested cross-section based on the ending index of the best matching subsequence of the most similar reference cross-section and the subsequent data trend thereof comprises: obtaining data after the ending index of the best matching subsequence of the most similar reference cross-section; determining a maximum value in data from the ending index of the best matching subsequence of the most similar reference cross-section to data after the ending index; obtaining a true value of a last data of the short-term deformation monitoring time sequence of the to-be-tested cross-section; and predicting the deformation maximum value of the to-be-tested cross-section by a proportional increment method based on a reference cross-section value of the most similar reference cross-section at the ending index of the best matching subsequence, a subsequent maximum value of the most similar reference cross-section and the true value of the last data of the short-term deformation monitoring time sequence of the to-be-tested cross-section, the proportional increment method comprising: when the reference cross-section value of the most similar reference cross-section at the ending index of the best matching subsequence is not zero, predicting the deformation maximum value = true value of the to-be-tested cross-section * (subsequent maximum value of the most similar reference cross-section / reference cross-section value of the most similar reference cross-section at the ending index of the best matching subsequence); when the reference cross-section value of the most similar reference cross-section at the ending index of the best matching subsequence is zero and the subsequent maximum value of the most similar reference cross-section is greater than zero, predicting the deformation maximum value = true value of the to-be-tested cross-section + subsequent maximum value of the most similar reference cross-section; or when the reference cross-section value of the most similar reference cross-section at the ending index of the best matching subsequence is zero and the subsequent maximum value of the most similar reference cross-section is zero, predicting the deformation maximum value = true value of the to-be-tested cross-section.

2. The method of claim 1, wherein, Before obtaining the historical cross-section deformation monitoring time sequence set based on original deformation monitoring data of each cross-section monitoring point of completed monitoring, the method further comprises pre-processing the original deformation monitoring data, wherein the pre-processing comprises: aggregating the original deformation monitoring data and performing interpolation processing on a data sequence with a missing date, wherein the aggregation is to take an average value for a plurality of data; and The data sequence after interpolation processing is subjected to outlier detection and interpolation repair to obtain a cleaned data sequence.

3. The method of claim 2, wherein, The data sequence after difference processing is subjected to outlier detection and interpolation repair to obtain a cleaned data sequence, comprising: Performing a moving average smoothing process on the deformation monitoring time series in the set of historical cross-section deformation monitoring time series; Calculating the standard score Z-score and the interquartile range IQR based on the smoothed data sequence; Identifying data exceeding the preset Z-score threshold or the preset IQR upper and lower limits as outliers; and Using the average value of the data before and after the identified outliers for interpolation replacement.

4. The method of claim 1, wherein, Before calculating the dynamic time warping DTW distance between the plurality of historical deformation monitoring subsequences with the first length in the historical cross-section deformation monitoring time series of each reference cross-section and the short-term deformation monitoring sequence of the cross-section to be tested, the method further comprises: Based on the maximum and minimum values of each historical deformation monitoring subsequence with the first length in the historical cross-section deformation monitoring time series of the reference cross-section, respectively normalizing each historical deformation monitoring subsequence of the reference cross-section; and Based on the maximum and minimum values of the short-term deformation monitoring time series of the cross-section to be tested, normalizing the deformation monitoring time series of the cross-section to be tested.

5. The method of claim 1, wherein, When calculating the DTW distance, a preset window size is adopted to allow local flexibility when aligning the time series.

6. A tunnel deformation maximum prediction device characterized by comprising: The device comprises: at least one processor; and at least one memory coupled to the at least one processor and storing instructions for execution by the at least one processor, which, when executed by the at least one processor, implement the method of any one of claims 1 to 5.

7. A computer readable storage medium characterized by The computer-readable storage medium stores instructions which, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 5.

8. A computer program product, characterised in that, The computer program product comprises instructions which, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 5. The computer program product comprises instructions which, when executed by a computer, cause the computer to perform the method according to any one of claims 1 to 5.

Images