Method and apparatus for intelligent judgment of shield tail seal failure

Abstract

Disclosed in the present application are a method and apparatus for intelligent judgment of a shield tail seal failure. The method comprising: acquiring gap data between segments and a shield tail, external soil pressure data and grease state data; inputting the gap data into a shield attitude anomaly warning model, and outputting a first prediction result; inputting the external soil pressure data into an external soil anomaly warning model, and outputting a second prediction result, the external soil anomaly warning model being used for determining the turning point of a sudden change of an external soil pressure; inputting the grease state data into a grease anomaly warning model, and outputting a third prediction result, the grease anomaly warning model being used for determining, on the basis of the grease state data, whether grease contains water, whether the grease temperature is abnormal, and whether the grease pressure is abnormal; and on the basis of a Bayesian algorithm, fusing the first prediction result, the second prediction result and the third prediction result to output an overall prediction result.

Description

Intelligent Judgment Method and Device for Tail Shield Seal Failure

[0001] Related applications

[0002] This application claims Chinese Patent Application No. 202411684012.0, filed on November 22, 2024, and incorporates the disclosure of the aforementioned patent application as part of this application. Technical Field

[0003] This application relates to the field of tunnel boring machine technology, and in particular to an intelligent method and device for judging the failure of the shield tail seal. Background Technology

[0004] This section is intended to provide background or context for the embodiments of this application set forth in the claims. The description herein is not an admission that it is prior art simply because it is included in this section.

[0005] The tail shield sealing system generally includes tail shield seals, a grease injection system, and a grout stop plate. It is a critical component of the tunnel boring machine (TBM), and seal failure can lead to the influx of external mud and groundwater into the tunnel, causing ground subsidence and casualties. Factors such as the TBM's burial depth, tail shield attitude, external soil pressure, grease injection conditions in the tail shield cavity, and tail brush wear all affect the sealing effect. Failure to diagnose and provide timely warnings will result in seal failure and leakage.

[0006] Existing systems for diagnosing shield tail seal failure primarily analyze raw data from the tunnel boring machine (TBM) or parameters collected by sensing terminals to propose early warning rules. For example, analyzing the TBM's grease injection system, aggregating pressures in various cavities and branches, and combining this with grease levels for real-time warnings only addresses the grease injection system. This results in warning schemes that fail to effectively reflect the shield tail seal's performance and offer no practical guidance for on-site operations. Similarly, techniques using time series segmentation and unsupervised learning to process and detect anomalies in the shield tail seal system pressure data focus solely on pressure data, ignoring the influence of various practical factors in field operations, leading to a high error rate in diagnosing shield tail seal failure. Summary of the Invention

[0007] This application provides an intelligent method for judging shield tail seal failure, which comprehensively considers multi-dimensional information such as tail shield attitude, external soil pressure, and grease injection quality to monitor the shield tail seal status from all angles and perspectives, thereby achieving intelligent diagnosis and accurate early warning of shield tail seal failure and ensuring safe tunneling under complex working conditions. The method includes:

[0008] Acquire data on the gap between the tunnel lining segments and the tail shield, external soil pressure, and grease condition.

[0009] The gap data between the tunnel segment and the tail shield is input into the shield attitude anomaly early warning model, and the first prediction result is output. The shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segment and the tail shield; determine the first prediction result based on the uniformity of the tail shield gap distribution; the first prediction result reflects the degree of compression of the tail shield brush.

[0010] External soil pressure data is input into the external soil anomaly early warning model, and a second prediction result is output. The external soil anomaly early warning model is used to: determine the first external soil pressure data of the sudden change in external soil pressure based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data. The second prediction result reflects the influence of external soil pressure on the shield tail sealing performance.

[0011] The grease state data is input into the grease anomaly early warning model, and the third prediction result is output. The grease anomaly early warning model is used to determine the third prediction result by using the results of whether the grease contains water, whether the grease temperature is abnormal, and whether the grease pressure is abnormal, as determined by the grease state data. The third prediction result reflects the impact of the grease state on the shield tail sealing performance.

[0012] Based on the Bayesian algorithm, the first, second, and third prediction results are fused together to output a comprehensive prediction result. The comprehensive prediction result includes no risk, oil spill, leakage risk level, and leakage location.

[0013] This application embodiment also provides an intelligent detection device for shield tail seal failure, which comprehensively considers multi-dimensional information such as tail shield attitude, external soil pressure, and grease injection quality to monitor the shield tail seal status from all angles and perspectives, thereby achieving intelligent diagnosis and accurate early warning of shield tail seal failure and ensuring safe tunneling under complex working conditions. The device includes:

[0014] The data acquisition module is used to acquire gap data between the tunnel lining segments and the tail shield, external soil pressure data, and grease status data.

[0015] The shield attitude anomaly early warning module is used to input the gap data between the tunnel segments and the tail shield into the shield attitude anomaly early warning model and output the first prediction result. The shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segments and the tail shield; determine the first prediction result based on the uniformity of the tail shield gap distribution; and reflect the degree of compression of the tail shield brush.

[0016] The external soil anomaly early warning module is used to input external soil pressure data into the external soil anomaly early warning model and output a second prediction result. The external soil anomaly early warning model is used to: determine the first external soil pressure data of the sudden change in external soil pressure based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data. The second prediction result reflects the influence of external soil pressure on the shield tail sealing performance.

[0017] The grease condition anomaly early warning module is used to input grease condition data into the grease anomaly early warning model and output a third prediction result. The grease anomaly early warning model is used to determine the third prediction result by using the results of whether the grease contains water, whether the grease temperature is abnormal, and whether the grease pressure is abnormal, as determined from the grease condition data. The third prediction result reflects the impact of the grease condition on the shield tail sealing performance.

[0018] The fusion processing module is used to fuse the first, second, and third prediction results based on the Bayesian algorithm and output a comprehensive prediction result. The comprehensive prediction result includes no risk, oil spill, leakage risk level, and leakage location.

[0019] This application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned intelligent judgment method for shield tail seal failure.

[0020] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described intelligent judgment method for shield tail seal failure.

[0021] This application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-mentioned intelligent judgment method for shield tail seal failure.

[0022] In this embodiment, gap data between the tunnel lining segment and the tail shield, external soil pressure data, and grease status data are acquired. The corresponding data are then processed using independently operating shield attitude anomaly early warning models, external soil anomaly early warning models, and grease anomaly early warning models. The influence of tail shield attitude, external soil pressure, tail shield cavity status, and tail shield sealing grease injection on the sealing effect is comprehensively considered. This allows for all-round, multi-angle monitoring of the tail shield sealing status, achieving intelligent diagnosis and assessment of tail shield sealing failure through multi-information fusion. This significantly improves the accuracy of tail shield sealing failure early warnings and provides timely and accurate early warning information for construction personnel. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0024] Figure 1 is a flowchart illustrating the intelligent judgment method for shield tail seal failure in an embodiment of this application;

[0025] Figure 2 is a schematic diagram of the uniform distribution of the shield tail gap in an embodiment of this application;

[0026] Figure 3 is a schematic diagram of excessive shield tail gap in an embodiment of this application;

[0027] Figure 4 is a flowchart of the change point monitoring algorithm in the embodiments of this application;

[0028] Figure 5 is a flowchart illustrating the abnormal oil pressure warning process in an embodiment of this application;

[0029] Figure 6 is an example of the intelligent judgment method for shield tail seal failure in the embodiments of this application;

[0030] Figure 7 is another example of the intelligent judgment method for shield tail seal failure in the embodiments of this application;

[0031] Figure 8 is a schematic diagram of the intelligent judgment device for shield tail seal failure in the embodiments of this application. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the embodiments of this application will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments and their descriptions are used to explain this application, but are not intended to limit this application.

[0033] To address the shortcomings of existing technologies, this application proposes a method and device for intelligent judgment of shield tail seal failure involving multiple coupled systems. This solves the problems of unclear diagnosis and inaccurate early warning of shield tail seal failure in existing technologies, aiming to identify seal failure risks, optimize construction decisions, and ensure the safe tunneling of tunnel boring machines.

[0034] Figure 1 is a flowchart illustrating the intelligent judgment method for shield tail seal failure in an embodiment of this application. As shown in Figure 1, the method includes:

[0035] Step 101: Obtain the gap data between the tunnel lining segment and the tail shield, the external soil pressure data, and the grease status data;

[0036] Step 102: Input the gap data between the tunnel segment and the tail shield into the shield attitude anomaly early warning model and output the first prediction result; the shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segment and the tail shield; and determine the first prediction result based on the uniformity of the tail shield gap distribution.

[0037] Step 103: Input the external soil pressure data into the external soil anomaly early warning model and output the second prediction result; the external soil anomaly early warning model is used to: determine the first external soil pressure data of the external soil pressure change based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data.

[0038] Step 104: Input the oil state data into the oil anomaly early warning model and output the third prediction result; The oil anomaly early warning model is used to: determine the third prediction result by using the results of whether the oil contains water, whether the oil temperature is abnormal, and whether the oil pressure is abnormal, as determined from the oil state data.

[0039] Step 105: Based on the Bayesian algorithm, the first prediction result, the second prediction result, and the third prediction result are fused together to output a comprehensive prediction result. The comprehensive prediction result includes no risk, oil spill, leakage risk level, and leakage location.

[0040] As shown in Figure 1, this embodiment acquires gap data between the tunnel segment and the tail shield, external soil pressure data, and grease status data. It then processes the corresponding data using independently operating shield attitude anomaly early warning models, external soil anomaly early warning models, and grease anomaly early warning models. By comprehensively considering the influence of tail shield attitude, external soil pressure, tail shield cavity status, and tail shield sealing grease injection on the sealing effect, the tail shield sealing status is monitored in a comprehensive and multi-angle manner. This achieves intelligent diagnosis and assessment of tail shield sealing failure through multi-information fusion, significantly improving the accuracy of tail shield sealing failure early warnings and providing timely and accurate early warning information for construction personnel.

[0041] During tunnel boring machine (TBM) excavation, the stability and reliability of the tail shield sealing system are crucial factors in ensuring tunnel construction safety. However, due to the complexity and unique environment of the tail shield sealing system, leaks are difficult to directly observe and accurately assess, posing a potential risk to construction safety. This application's embodiment installs sensors at multiple locations. For example, a laser sensor is installed at the junction of the tail shield and the tunnel lining segments to measure the tail shield gap (i.e., the gap data between the tunnel lining segments and the tail shield). Sensing and monitoring sensors are installed within the tail shield sealing cavity to acquire parameters such as grease pressure, grease temperature, and moisture content. This application's embodiment employs a multi-system coupled intelligent tail shield seal failure judgment method that comprehensively considers the influence of tail shield attitude, external soil pressure, tail shield cavity state parameters, and tail shield seal grease injection parameters on the sealing effect. Based on the coupling between each sub-model, multiple independent models with parallel computation are formed, achieving intelligent judgment and assessment of tail shield seal failure through multi-information fusion. This provides timely and accurate early warning information for construction personnel and has significant guiding significance on-site. When the comprehensive prediction results indicate that the leakage risk level exceeds the preset level, the shield tail seal will be deemed to have failed. An early warning message will be immediately pushed to the terminal where the main driver is located, and the leakage location will be transmitted in real time to the user terminal connected to the server so that the user can carry out timely maintenance.

[0042] In step 101, the gap data between the tunnel segment and the tail shield, the external soil pressure data, and the grease status data are obtained.

[0043] Preferably, the system acquires gap data between the tunnel segment and the tail shield detected by the first sensor, external soil pressure data detected by the second sensor, and grease status data detected by the third sensor. The first sensor is installed at the junction of the tail shield and the tunnel segment; the second sensor is installed at the part of the shield tail that is in direct contact with the surrounding soil; and the third sensor is installed inside the shield tail sealing cavity. The grease status data reflects the moisture content, temperature, and pressure of the grease.

[0044] For example, acquiring gap data collected by laser sensors, external soil pressure data collected by mud pressure sensors, and oil condition data such as oil temperature and pressure data collected by temperature and pressure sensors.

[0045] These sensors communicate directly or indirectly with the server, which ultimately performs calculations based on the real-time sensor data.

[0046] In one embodiment, after acquiring the gap data between the tunnel segment and the tail shield detected by the first sensor, the external soil pressure data detected by the second sensor, and the grease condition data detected by the third sensor, the process may further include:

[0047] The gap data between the tunnel lining segment and the tail shield detected by the first sensor, the external soil pressure data detected by the second sensor, and the grease condition data detected by the third sensor are preprocessed as follows:

[0048] Moving average filtering and outlier removal.

[0049] In this example, moving average filtering and outlier removal are used to process the acquired data, thereby improving data quality.

[0050] In step 102, the gap data between the tunnel segments and the tail shield is input into the shield attitude anomaly early warning model, and a first prediction result is output. The shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segments and the tail shield; and determine the first prediction result based on the uniformity of the tail shield gap distribution. In this embodiment, the first prediction result reflects the degree of compression of the tail shield brush.

[0051] In this embodiment, a shield attitude anomaly early warning model is pre-constructed. This model is used to predict whether the shield attitude is abnormal, and thus predict the risk of leakage.

[0052] The uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield is determined by the gap data between the tunnel segments and the tail shield. This mainly involves determining the size of the tail shield gap at different positions measured along the entire circumference of the tail shield of the tunnel boring machine. By comparing the gap values ​​at different positions, it can be determined whether the tunnel boring machine has shifted or whether the tunnel segments have deformed.

[0053] The gap data between the tunnel lining segments and the tail shield can include the gap value and monitoring location. In one embodiment, the shield attitude anomaly early warning model is specifically used for:

[0054] Determine the tail shield gap value at each point on the circumference of the tail shield based on the gap value and monitoring location;

[0055] Based on the average value and standard deviation of the shield tail gap at each point, determine the uniformity of the shield tail gap distribution at multiple points on the circumference of the tail shield; the uniformity of distribution includes uniform distribution and non-uniform distribution.

[0056] When the distribution is uneven, the relative height difference between the segment and the tail shield seal is calculated based on the deviation of the tail shield and the segment on the axis.

[0057] A first prediction result is determined based on the relative height difference. This first prediction result indicates whether there is a risk of leakage in the current shield tunneling posture dimension. For example, if there is a risk of leakage, a first value is used as the first prediction result; if there is no risk of leakage, a second value is used as the first prediction result. The first and second values ​​are different values; for example, the first value is 1 and the second value is 0.

[0058] Changes in the shield's attitude directly affect the tail gap. The gap between the tail seal and the tunnel lining segments is calculated based on this tail gap to indirectly reflect the degree of compression by the tail brush. The specific modeling approach in this application is as follows:

[0059] 1) Traverse the maximum and minimum points of the shield tail gap to determine whether the tail shield attitude has a large deviation;

[0060] 2) Given the tail shield diameter, segment diameter, and relevant design parameters of the tail shield sealing brush, calculate the height difference at the junction of the segment and the tail shield seal using the deviation angle α between the tail shield and the center of the segment.

[0061] 3) When the height difference h is less than or equal to the set threshold, a sealing leakage warning is issued (i.e., there is a risk of leakage). It can be understood that, as shown in Figure 2, when the height difference on the first side (e.g., the left side) of the junction is less than or equal to the set threshold, the height difference on the second side (e.g., the right side) opposite to the first side must be greater than or equal to the set threshold.

[0062] During implementation, based on the gap between the tunnel segment and the tail shield detected by the first sensor, the tail shield gap value at each point on the circumference of the tail shield is calculated by back-calculating the extreme points of the current gap, thereby accurately depicting the relative positional relationship between the tail shield and the tunnel segment. When the tail shield gap is uniformly distributed, it is considered that there is no risk of leakage to the tail shield seal at this position. When there is a large deviation in the tail shield gap in the lateral or longitudinal direction, the relative height difference between the tunnel segment and the tail shield seal is calculated based on the deviation of the tail shield and the tunnel segment on the axis, thereby reflecting the degree of compression of the tail shield brush and realizing early warning assessment of leakage risk.

[0063] Figure 2 is a schematic diagram of the uniform distribution of the shield tail gap in the embodiment of this application. As shown in Figure 2, when the center of the tail shield is coaxial with the center of the tube segment, the shield tail gaps in the upper, lower, left, and right parts are all in a uniform distribution state. The shield tail sealing brush will be in close contact with the tube segment, and the shield tail sealing cavity is filled with grease, resulting in a good sealing effect.

[0064] Figure 3 is a schematic diagram of excessive shield tail gap in an embodiment of this application. As shown in Figure 3, when the shield attitude deviates, the center of the tail shield and the center of the segment are no longer coaxial. If the shield tail gap is too large or too small, the shield tail sealing brush will detach from the segment, the shield tail seal will fail, and it is very easy to cause sealing failure and mud leakage.

[0065] In step 103, external soil pressure data is input into the external soil anomaly early warning model, and a second prediction result is output. The external soil anomaly early warning model is used to: determine the first external soil pressure data with a sudden change in external soil pressure based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data with a sudden change in external soil pressure. In this embodiment, the second prediction result reflects the influence of external soil pressure on the shield tail sealing performance. For example, when it is determined that there is a first external soil pressure data with a sudden change in external soil pressure, the first value is used as the second prediction result; if there is no first external soil pressure data with a sudden change in external soil pressure, the second value is used as the first prediction result. The first value and the second value are different values, for example, the first value is 1 and the second value is 0. Further, the second prediction result may also include, for example, the leakage risk level and the leakage location.

[0066] External soil pressure has a decisive impact on the shield tail sealing effect. A sudden increase in external soil pressure greatly increases the risk of shield tail seal leakage. To address this issue, a Pruned Exact Linear Time (PELT) algorithm is employed to capture the moments of sudden increases or decreases in external soil pressure. Based on precise time division and an efficient load attenuation mechanism, it sensitively captures subtle changes in the pressure data of the shield tail slurry chamber, identifying key turning points of sudden pressure increases or decreases. This application's embodiment closely integrates the dynamic monitoring capability of the PELT algorithm with the real-time impact of external soil pressure on the shield tail sealing effect, achieving accurate assessment of the risk of shield tail seal leakage.

[0067] The change point detection algorithm in this embodiment of the application finds the optimal change segmentation point, i.e., the anomaly detection point, by minimizing the cost function through dynamic programming. Figure 4 is a flowchart of the change point monitoring algorithm in this embodiment of the application. Referring to Figure 4, the external soil anomaly early warning model in this embodiment of the application is specifically used for:

[0068] The external soil pressure data is divided into multiple segments according to the detection time.

[0069] Calculate the difference, volatility, and mean of the external soil pressure data within each segment;

[0070] Then, the mean squared error is used as the cost function to measure the deviation between the mean and the true value within the current interval of the input test time series. Specifically, using the difference, volatility, and mean of the external soil pressure data within each segment, the minimum cost function for the external soil pressure data within each segment is calculated according to the following formula:

[0071] In the formula, L(t',t) is the cost of the time interval [t',t], i represents the time interval [t',t], and y iμ is the difference or volatility of the test time series (external soil pressure data) over the time interval [t', t]. [t',t] It is the mean of external soil pressure data within the interval [t', t]; where the cost function is minimized by dynamic programming according to the following formula:

[0072] In the formula, C(t) represents the minimum cost from time point 0 to time point t, C(t′) represents the minimum cost from time point 0 to time point t′, and L(t′,t) is the cost of [t′,t].

[0073] The second prediction result is determined based on the minimum cost function of the external soil pressure data within each segment.

[0074] In this example, based on the PELT algorithm, pruning techniques are used to reduce computation. This effectively eliminates parts that are unlikely to be the optimal solution during dynamic programming, thereby significantly reducing computation and improving the efficiency and accuracy of change point detection.

[0075] In step 104, the grease state data is input into the grease anomaly early warning model, and a third prediction result is output. The grease anomaly early warning model is used to determine the third prediction result based on the results determined from the grease state data regarding whether the grease contains water, whether the grease temperature is abnormal, and whether the grease pressure is abnormal. In this embodiment, the third prediction result reflects the influence of the grease state on the shield tail sealing performance. Specifically, for example, the data corresponding to whether the grease contains water, whether the grease temperature is abnormal, and whether the grease pressure is abnormal are used as the third prediction result. For example, when the grease contains water (there is a risk), the first data is used as the third prediction result; otherwise (no water, i.e., no risk), the second data is used as the third prediction result. The first and second values ​​are different values; for example, the first value is 1, and the second value is 0. Similarly, the third prediction result corresponding to whether the grease temperature is abnormal and whether the grease pressure is abnormal is determined. Further, this embodiment may also have only one third prediction result. Specifically, if there is a corresponding risk of leakage among whether the grease contains water, whether the grease temperature is abnormal, and whether the grease pressure is abnormal, the first value is used as the third prediction result; otherwise, the second value is used as the third prediction result. Furthermore, the third prediction result may also include, for example, the leakage risk level, the location of the leakage, and information indicating the presence of grease spills.

[0076] The grease anomaly early warning model is mainly a comprehensive early warning judgment of grease moisture content, grease temperature and grease pressure in the shield tail sealing cavity. It can include grease temperature anomaly early warning algorithm, grease moisture content anomaly early warning algorithm and grease pressure anomaly early warning algorithm.

[0077] In one embodiment, the oil state data includes oil moisture content data, oil temperature data, and oil pressure data;

[0078] The oil abnormality early warning model is specifically used for:

[0079] Based on the moisture content data of oils, determine whether the oil contains water; when water is detected in the oil, issue an immediate warning.

[0080] Based on the oil temperature data, determine whether the oil temperature exceeds the first preset value; if the oil temperature exceeds the first preset value, immediately issue an early warning message;

[0081] Outlier values ​​were removed from the grease pressure data of a single monitoring point in the tail seal cavity at different tail brush cavities, and the grease pressure data at the moment of grease injection in the tail seal cavity were also removed, resulting in grease pressure data after removing outliers.

[0082] A change point detection algorithm is used to identify abnormal data in the grease pressure data after removing anomalies. The abnormal data is then filtered out, and the current abnormal grease pressure monitoring location in the sealing cavity is output.

[0083] The third prediction result is formed by combining whether the grease contains water, whether the grease temperature exceeds the first preset value, and the current abnormal grease pressure monitoring location in the sealing cavity.

[0084] During implementation, an abnormal moisture content early warning algorithm is used for oil moisture content data. If water content is detected in the oil, an early warning signal is issued immediately. The specific judgment formula is as follows:

[0085] In the formula, f water This indicates the result of the water content warning for oils.

[0086] Moisture in grease is a significant sign of shield tail seal failure, and immediate measures must be taken to prevent further damage and ensure the safety of tunnel boring machine construction.

[0087] An abnormal oil temperature warning algorithm is used to directly monitor the oil temperature data collected by the sensor integration module. If the oil temperature exceeds a preset threshold range, an early warning is triggered immediately. The specific judgment formula is as follows:

[0088] In the formula, f temp This indicates the result of the oil temperature warning; T represents the temperature.

[0089] Excessive temperature indicates that the grease is overheated, affecting the sealing performance; excessively low temperature may indicate sensor malfunction or poor grease condition, so it is important to ensure timely response to prevent sensor failure.

[0090] The grease pressure anomaly early warning algorithm is used to detect anomalies in the grease pressure at a single monitoring point within the shield tail sealing cavity in different tail brush cavities. It eliminates the grease injection time within the shield tail sealing cavity and, combined with the simulated leakage mechanism of the shield tail seal and the anomaly detection results of each cavity, comprehensively assesses the leakage risk at the current monitoring point. The specific algorithm is illustrated in Figures 4 and 5, with Figure 5 showing a flowchart of the grease pressure anomaly early warning algorithm in this embodiment. The grease pressure anomaly early warning algorithm first calculates the differential, fluctuation rate, and mean pressure within the shield tail sealing cavity. Then, it combines the pressure to form a 4-dimensional feature, which serves as the input to the PELT algorithm for anomaly change points (the PELT algorithm principle is consistent with external soil pressure). Finally, it outputs the anomaly points for the current sealing cavity and the current monitoring point.

[0091] Specifically, the difference method is used to measure the change between adjacent observations in time series data (grease pressure data sorted by acquisition time). It can represent the instantaneous change in shield tail seal grease pressure data over time. The specific calculation formula is as follows: ΔP t =P t -P t-1

[0092] In the formula, ΔP t P is the difference at time t. t P is the shield tail seal pressure value at time t. t-1 It is the shield tail seal pressure value at time t-1.

[0093] Volatility reflects the stability and magnitude of changes in data over a period of time. It measures the stability of the trend change in the pressure of the shield tail seal grease over a certain period of time. The specific calculation formula is as follows:

[0094] In the formula, μ is the average pressure of the shield tail sealing grease over time, and P i is the tail seal pressure value at time point i, and N is the total number of samples within the time period. var is the variance of the tail seal pressure, and vol is the fluctuation rate of the tail seal pressure.

[0095] In Figure 5, the abnormal grease pressure data of the front and rear chambers are statistically analyzed by the abnormal change point detection unit, and the grease pressure data at the time of grease injection is removed to obtain a series of abnormal monitoring points of the front and rear chambers. Finally, the results of the leakage risk assessment of the current monitoring point are comprehensively judged and output.

[0096] Finally, in step 105, based on the Bayesian algorithm, the first, second, and third prediction results are fused to output a comprehensive prediction result. The comprehensive prediction result includes no risk, grease spillage, leakage risk level, and leakage location. Specifically, if the fused result is less than a preset threshold, the comprehensive prediction result is determined to be no risk; if it is greater than or equal to the preset threshold, the grease spillage and / or leakage risk level and leakage location from the second and / or third prediction results are used as the comprehensive prediction result.

[0097] In one embodiment, based on the Bayesian algorithm, the first prediction result, the second prediction result, and the third prediction result are fused to output a comprehensive prediction result, which may include:

[0098] Based on the Bayesian algorithm, the posterior probabilities of the shield attitude anomaly early warning model, external soil anomaly early warning model, and grease anomaly early warning model are calculated using the pre-set prior probabilities of these models.

[0099] Based on the posterior probabilities and the first, second, and third prediction results of the shield attitude anomaly early warning model, the external soil anomaly early warning model, and the grease anomaly early warning model, a comprehensive prediction result is output.

[0100] Figure 6 is an example of the intelligent judgment method for shield tail seal failure in this application embodiment. Referring to Figure 6, the acquired data undergoes preprocessing such as outlier removal, filling, and moving average filtering. It is then processed by the shield attitude anomaly early warning model M1, the external soil anomaly early warning model M2, and the shield tail seal cavity grease anomaly early warning model M3, respectively. Finally, a fusion decision is output based on the Bayesian algorithm to send the leakage risk. When outputting the fusion decision based on the Bayesian algorithm, the prior probabilities P(M1, M2, and M3 of the shield attitude anomaly early warning model M1, external soil anomaly early warning model M2, and grease anomaly early warning model M3 can be pre-set using expert experience. i Then, Bayes' theorem is used to calculate the posterior probability P(M) of each model. i / F):

[0101] Wherein, P(F / M i ) is model M i The likelihood is P(F), which is the marginal likelihood. The final fusion decision's early warning result is a weighted average of the early warning models:

[0102] Among them, W i This is the model's early warning result, while Warn is the final result. For example, if the final result is less than a preset threshold, it is determined that there is no risk; otherwise, it is determined that there is a risk.

[0103] By organically integrating the early warning results of each sub-model through the Bayesian algorithm, and fully considering the correlation between the sub-models, a comprehensive decision-making system for shield tail seal failure diagnosis is realized through multi-information fusion. This significantly improves the accuracy of early warning of shield tail seal leakage risk and can be applied to various complex working conditions and multiple devices, providing important guidance for the field.

[0104] Figure 7 is another example of the intelligent judgment method for shield tail seal failure in this application embodiment. As shown in Figure 7, based on the shield attitude anomaly early warning model, the shield tail gap is used to determine the state of the shield tail and segments; based on the external soil pressure monitoring device, the external soil anomaly early warning model, mud pressure, and formulas are used to determine the state of the mud-water cavity; based on the shield tail seal state monitoring system, the grease anomaly early warning model is used to comprehensively process grease pressure, grease moisture content, and grease temperature, and the corresponding formulas are used to determine the state of the shield tail seal cavity. Finally, the structure is comprehensively processed to output the shield tail seal cavity leakage risk assessment result, including no risk, grease overflow, and leakage risk. It should be noted that the input parameters in this application embodiment include, but are not limited to, grease injection parameters, shield tail attitude parameters, shield tail cavity state parameters, and external soil pressure parameters. In this application embodiment, the constructed change point detection algorithm can be replaced by a more suitable method, using a machine learning model for processing, and can be arbitrarily changed.

[0105] In summary, compared with the prior art, the embodiments of this application provide a multi-system coupled intelligent judgment method for shield tail seal failure. It comprehensively considers the influence of multiple factors on the shield tail seal effect, can accurately predict the risk of leakage due to shield tail seal failure, provide suggestions for the main driver for grease injection and grouting, reduce potential safety hazards, and ensure the safe tunneling of the tunnel boring machine.

[0106] This application also provides an intelligent detection device for shield tail seal failure, as described in the following embodiments. Since the principle by which this device solves the problem is similar to the intelligent detection method for shield tail seal failure, the implementation of this device can refer to the implementation of the intelligent detection method for shield tail seal failure; repeated details will not be elaborated further.

[0107] Figure 8 is a schematic diagram of the intelligent detection device for shield tail seal failure in an embodiment of this application. As shown in Figure 8, the intelligent detection device 800 for shield tail seal failure includes:

[0108] The data acquisition module 801 is used to acquire gap data between the tunnel segment and the tail shield, external soil pressure data, and grease status data.

[0109] The shield attitude anomaly early warning module 802 is used to input the gap data between the tunnel segment and the tail shield into the shield attitude anomaly early warning model and output the first prediction result; the shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segment and the tail shield; and determine the first prediction result based on the uniformity of the tail shield gap distribution.

[0110] The external soil anomaly early warning module 803 is used to input external soil pressure data into the external soil anomaly early warning model and output a second prediction result; the external soil anomaly early warning model is used to: determine the first external soil pressure data of the external soil pressure change based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data.

[0111] The oil and fat state anomaly early warning module 804 is used to input oil and fat state data into the oil and fat anomaly early warning model and output a third prediction result; the oil and fat anomaly early warning model is used to: determine the third prediction result by using the results of whether the oil and fat contain water, whether the oil and fat temperature is abnormal, and whether the oil and fat pressure is abnormal as determined by the oil and fat state data.

[0112] The fusion processing module 805 is used to fuse the first prediction result, the second prediction result and the third prediction result based on the Bayesian algorithm, and output a comprehensive prediction result. The comprehensive prediction result includes no risk, oil spill, leakage risk level and leakage location.

[0113] In one embodiment, the data acquisition module 801 is specifically used to: acquire gap data between the tunnel segment and the tail shield detected by the first sensor, external soil pressure data detected by the second sensor, and grease state data detected by the third sensor; the first sensor is installed at the junction of the tail shield and the tunnel segment; the second sensor is installed at the part where the tail shield is in direct contact with the surrounding soil; the third sensor is installed in the sealing cavity of the tail shield, and the grease state data reflects the moisture content, temperature, and pressure of the grease.

[0114] In one embodiment, the gap data between the segment and the tail shield includes the gap value and the monitoring position;

[0115] The shield tunneling attitude anomaly early warning module 802 is specifically used for:

[0116] The tail gap value at each point on the tail shield circumference is determined based on the gap value and monitoring location; the uniformity of the tail gap distribution at multiple points on the tail shield circumference is determined based on the average value and standard deviation of the tail gap value at each point; the uniformity of distribution includes uniform distribution and non-uniform distribution; when the distribution is non-uniform, the relative height difference between the segment and the tail shield seal is calculated based on the deviation of the tail shield and the segment on the axis; the first prediction result is determined based on the relative height difference.

[0117] In one embodiment, the external soil anomaly early warning module 803 is specifically used for:

[0118] The external soil pressure data is divided into multiple segments according to the detection time.

[0119] Calculate the difference, volatility, and mean of the external soil pressure data within each segment;

[0120] Using the difference, volatility, and mean of the external earth pressure data within each segment, the minimum cost function for the external earth pressure data within each segment is calculated according to the following formula:

[0121] In the formula, L(t',t) is the cost of the time interval [t',t], i represents the time interval [t',t], and y i Let μ be the difference or volatility of the external soil pressure data over the time interval [t', t]. [t',t] It is the mean of external soil pressure data within the interval [t', t]; where the cost function is minimized by dynamic programming according to the following formula:

[0122] In the formula, C(t) represents the minimum cost from time point 0 to time point t, C(t′) represents the minimum cost from time point 0 to time point t′, and L(t′,t) is the cost of [t′,t].

[0123] The second prediction result is determined based on the minimum cost function of the external soil pressure data within each segment.

[0124] In one embodiment, the oil state data includes oil moisture content data, oil temperature data, and oil pressure data;

[0125] The abnormal oil condition early warning module 804 is specifically used for:

[0126] Based on the moisture content data of oils, determine whether the oil contains water; when water is detected in the oil, issue an immediate warning.

[0127] Based on the oil temperature data, determine whether the oil temperature exceeds the first preset value; if the oil temperature exceeds the first preset value, immediately issue an early warning message;

[0128] Outlier values ​​were removed from the grease pressure data of a single monitoring point in the tail seal cavity at different tail brush cavities, and the grease pressure data at the moment of grease injection in the tail seal cavity were also removed, resulting in grease pressure data after removing outliers.

[0129] A change point detection algorithm is used to identify abnormal data in the grease pressure data after removing anomalies. The abnormal data is then filtered out, and the current abnormal grease pressure monitoring location in the sealing cavity is output.

[0130] In one embodiment, the fusion processing module 805 is specifically used for:

[0131] Based on the Bayesian algorithm, the posterior probabilities of the shield attitude anomaly early warning model, external soil anomaly early warning model, and grease anomaly early warning model are calculated using the pre-set prior probabilities of these models.

[0132] Based on the posterior probabilities and the first, second, and third prediction results of the shield attitude anomaly early warning model, the external soil anomaly early warning model, and the grease anomaly early warning model, a comprehensive prediction result is output.

[0133] In one embodiment, the device 800 may further include:

[0134] The data preprocessing module is used to preprocess the gap data between the tunnel segment and the tail shield detected by the first sensor, the external soil pressure data detected by the second sensor, and the grease status data detected by the third sensor after the data acquisition module 801 acquires the grease status data.

[0135] Moving average filtering and outlier removal.

[0136] This application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned intelligent judgment method for shield tail seal failure.

[0137] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described intelligent judgment method for shield tail seal failure.

[0138] This application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-mentioned intelligent judgment method for shield tail seal failure.

[0139] In this embodiment, sensors are arranged at the junction of the tail shield and the tunnel segment, at the location where the tail shield directly contacts the surrounding soil, and inside the tail shield sealing cavity to acquire data on the gap between the tunnel segment and the tail shield, external soil pressure, and grease status. The corresponding data are processed using independently operating shield attitude anomaly early warning models, external soil anomaly early warning models, and grease anomaly early warning models. The system comprehensively considers the influence of tail shield attitude, external soil pressure, tail shield cavity status, and tail shield sealing grease injection on the sealing effect, enabling all-round, multi-angle monitoring of the tail shield sealing status. This achieves intelligent diagnosis and assessment of tail shield sealing failure through multi-information fusion, significantly improving the accuracy of tail shield sealing failure early warnings and providing timely and accurate early warning information for construction personnel.

[0140] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0141] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more flowchart illustrations and / or one or more block diagrams.

[0142] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0143] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0144] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for intelligently determining shield tail seal failure, characterized in that, include: Acquire data on the gap between the tunnel lining segments and the tail shield, external soil pressure, and grease condition. The gap data between the tunnel segments and the tail shield is input into the shield attitude anomaly early warning model, and the first prediction result is output. The shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segments and the tail shield; and determine the first prediction result based on the uniformity of the tail shield gap distribution. The external soil pressure data is input into the external soil anomaly early warning model, and the second prediction result is output. The external soil anomaly early warning model is used to: determine the first external soil pressure data of the external soil pressure change based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data. The oil state data is input into the oil anomaly early warning model, and the third prediction result is output. The oil anomaly early warning model is used to determine the third prediction result by using the results of whether the oil contains water, whether the oil temperature is abnormal, and whether the oil pressure is abnormal, as determined by the oil state data. Based on the Bayesian algorithm, the first, second, and third prediction results are fused together to output a comprehensive prediction result. The comprehensive prediction result includes no risk, oil spill, leakage risk level, and leakage location.

2. The intelligent judgment method for shield tail seal failure as described in claim 1, characterized in that, Acquire data on the gap between the tunnel lining segments and the tail shield, external soil pressure, and grease condition, including: The system acquires data on the gap between the tunnel segment and the tail shield detected by the first sensor, data on the external soil pressure detected by the second sensor, and data on the grease condition detected by the third sensor. The first sensor is installed at the junction of the tail shield and the tunnel segment. The second sensor is installed at the part of the shield tail that is in direct contact with the surrounding soil. The third sensor is installed inside the sealing cavity of the shield tail. The grease condition data reflects the moisture content, temperature, and pressure of the grease.

3. The intelligent judgment method for shield tail seal failure as described in claim 1, characterized in that, The gap data between the tunnel segment and the tail shield includes the gap value and the monitoring location; The shield tunneling attitude anomaly early warning model is specifically used for: Determine the tail shield gap value at each point on the circumference of the tail shield based on the gap value and monitoring location; Based on the average value and standard deviation of the shield tail gap at each point, determine the uniformity of the shield tail gap distribution at multiple points on the circumference of the tail shield; the uniformity of distribution includes uniform distribution and non-uniform distribution. When the distribution is uneven, the relative height difference between the segment and the tail shield seal is calculated based on the deviation of the tail shield and the segment on the axis. The first prediction result is determined based on the relative height difference.

4. The intelligent judgment method for shield tail seal failure as described in claim 1, characterized in that, The external soil anomaly early warning model is specifically used for: The external soil pressure data is divided into multiple segments according to the detection time. Calculate the difference, volatility, and mean of the external soil pressure data within each segment; Using the difference, volatility, and mean of the external earth pressure data within each segment, the minimum cost function for the external earth pressure data within each segment is calculated according to the following formula: In the formula, L(t',t) is the cost of the time interval [t',t], i represents the time interval [t',t], and y i Let μ be the difference or volatility of the external soil pressure data over the time interval [t', t]. [t',t] It is the mean of external soil pressure data within the interval [t', t]; where the cost function is minimized by dynamic programming according to the following formula: In the formula, C(t) represents the minimum cost from time point 0 to time point t, C(t′) represents the minimum cost from time point 0 to time point t′, and L(t′,t) is the cost of [t′,t]. The second prediction result is determined based on the minimum cost function of the external soil pressure data within each segment.

5. The intelligent judgment method for shield tail seal failure as described in claim 1, characterized in that, Oil condition data includes oil moisture content data, oil temperature data, and oil pressure data; The oil abnormality early warning model is specifically used for: Determine whether the oil contains water based on the oil's moisture content data; An alert will be issued immediately if water is detected in the oil. Based on the oil temperature data, determine whether the oil temperature exceeds the first preset value; If the temperature of the oil exceeds the first preset value, an early warning message will be issued immediately. Outlier values ​​were removed from the grease pressure data of a single monitoring point in the tail seal cavity at different tail brush cavities, and the grease pressure data at the moment of grease injection in the tail seal cavity were also removed, resulting in grease pressure data after removing outliers. A change point detection algorithm is used to identify abnormal data in the grease pressure data after removing anomalies. The abnormal data is then filtered out, and the current abnormal grease pressure monitoring location in the sealing cavity is output.

6. The intelligent judgment method for shield tail seal failure as described in claim 1, characterized in that, Based on the Bayesian algorithm, the first, second, and third prediction results are fused to output a comprehensive prediction result, including: Based on the Bayesian algorithm, the posterior probabilities of the shield attitude anomaly warning model, external soil anomaly warning model, and grease anomaly warning model are calculated by using the pre-set prior probabilities of these models. Based on the posterior probabilities and the first, second, and third prediction results of the shield attitude anomaly early warning model, the external soil anomaly early warning model, and the grease anomaly early warning model, a comprehensive prediction result is output.

7. The intelligent judgment method for shield tail seal failure as described in claim 1, characterized in that, After acquiring the gap data between the tunnel segment and the tail shield detected by the first sensor, the external soil pressure data detected by the second sensor, and the grease condition data detected by the third sensor, the system also includes: The gap data between the tunnel lining segment and the tail shield detected by the first sensor, the external soil pressure data detected by the second sensor, and the grease condition data detected by the third sensor are preprocessed as follows: Moving average filtering and outlier removal.

8. A smart device for determining shield tail seal failure, characterized in that, include: The data acquisition module is used to acquire gap data between the tunnel lining segments and the tail shield, external soil pressure data, and grease status data. The shield attitude anomaly early warning module is used to input the gap data between the tunnel segments and the tail shield into the shield attitude anomaly early warning model and output the first prediction result; the shield attitude anomaly early warning model is used to: determine the uniformity of the tail shield gap distribution at multiple points on the circumference of the tail shield based on the gap data between the tunnel segments and the tail shield; and determine the first prediction result based on the uniformity of the tail shield gap distribution. The external soil anomaly early warning module is used to input external soil pressure data into the external soil anomaly early warning model and output a second prediction result; the external soil anomaly early warning model is used to: determine the first external soil pressure data of the external soil pressure change based on the external soil pressure data, and determine the second prediction result based on the first external soil pressure data. The oil and fat state anomaly early warning module is used to input oil and fat state data into the oil and fat anomaly early warning model and output a third prediction result; the oil and fat anomaly early warning model is used to determine the third prediction result by using the results of whether the oil and fat contain water, whether the oil and fat temperature is abnormal, and whether the oil and fat pressure is abnormal, as determined from the oil and fat state data. The fusion processing module is used to fuse the first, second, and third prediction results based on the Bayesian algorithm and output a comprehensive prediction result. The comprehensive prediction result includes no risk, oil spill, leakage risk level, and leakage location.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the intelligent judgment method for shield tail seal failure as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the intelligent judgment method for shield tail seal failure as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the intelligent judgment method for shield tail seal failure as described in any one of claims 1 to 7.

Images