Soil and water pressure prediction method, device, equipment, system and medium

Abstract

The application discloses a kind of water and soil pressure prediction method, device, equipment, system and medium, it is related to tunnel construction technical field, by the historical sample data obtained according to acquisition, the current data set of water and soil pressure is constructed, current state vector is constructed, the current distance between current state vector and each feature vector X in current data set is calculated, select k from current data set The sample with minimum current distance as the most similar historical working condition sequence, based on the most similar historical working condition sequence, predict future water and soil pressure, so that the application can be used, according to the most similar historical working condition sequence in the historical sample data obtained by acquisition future water and soil pressure prediction, and then also makes the application can be in the receiving process of pipe jacking machine, carry out the advance and accurate prediction of water and soil pressure, improve the safety of receiving operation.

Description

Technical Field

[0001] This invention relates to the field of tunnel construction technology, and in particular to a method, apparatus, equipment, system and medium for predicting soil and water pressure. Background Technology

[0002] Three-circle interlocking pipe jacking is a new type of pre-support method for large-section tunnels, and its receiving stage is one of the most risky stages of construction. When the pipe jacking machine head enters the receiving shaft, the breaking of the tunnel portal will change the original water and soil pressure balance. If the sealing measures are not in place, water inrush and sand inrush accidents can easily occur, leading to ground subsidence or even flooding of the receiving shaft.

[0003] Traditional water-stopping methods primarily rely on mechanical seals (such as rubber sheets and water-stopping airbags) for passive sealing. However, when geological conditions are complex and water and soil pressure fluctuate significantly, passive seals struggle to adapt to pressure changes, resulting in poor sealing performance. While some technologies have attempted to utilize the principle of slurry pressure balance, their control strategies often depend on real-time feedback, leading to response lag. When changes occur in the tunneling speed of the pipe jacking machine, geological conditions, or the external hydrological environment, the feedback control system cannot anticipate sudden pressure changes, still posing a risk of instability. Therefore, there is an urgent need for an intelligent control system capable of predicting pressure change trends and achieving proactive feedforward balancing. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method, device, equipment, system and medium for predicting soil and water pressure, so as to achieve advanced and accurate control of soil and water pressure during the receiving process and fundamentally improve the safety of receiving operations.

[0005] To achieve the above objectives, in a first aspect, the present invention proposes a method for predicting soil and water pressure, comprising the following steps: Based on the collected historical sample data, a current dataset of soil and water pressure is constructed; wherein, the current dataset includes multiple sample pairs, each sample pair includes a feature vector X and a corresponding target value Y, wherein the feature vector X is the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, as well as the soil and water pressure value and the current speed of the pipe jacking machine at the past n time steps, and the target value Y is the soil and water pressure value at the future k time steps. Construct the current state vector; wherein the current state vector includes the soil and water pressure values ​​of the current time step and the past n time steps, and the current speed of the pipe jacking machine; Calculate the current distance between the current state vector and each of the feature vectors X in the current dataset; Select k samples with the smallest distance from the current dataset as the most similar historical working condition sequence; Based on the most similar historical working condition sequence, future water and soil pressure is predicted.

[0006] In one embodiment, before the step of constructing the current dataset of soil and water pressure based on the collected historical sample data, the method further includes: Construct the feature vector X; wherein the feature vector X is represented by Formula 1, which is: Xi=[P(t),P(t 1), P(t) 2),...,P(t n),V(t)] P(t) represents the soil and water pressure value at time step t, and V(t) represents the speed of the pipe jacking machine at time step t.

[0007] In one embodiment, before the step of constructing the current dataset of soil and water pressure based on the collected historical sample data, the method further includes: Construct a second equation to describe the target value Y; wherein the second equation is: Yi = P(t+k) Where P(t+k) represents the predicted soil and water pressure for the next k steps.

[0008] In one embodiment, the step of calculating the current distance between the current state vector and each of the feature vectors X in the current dataset includes: Formula 3 is used to calculate the current distance between the current state vector and each feature vector X in the current dataset; wherein, Formula 3 is:

[0009] Let X represent the distance between the i-th sample and the current state vector, X represent the current state vector, Xi represent the feature vector of the i-th sample, and m represent the dimension of the feature vector.

[0010] In one embodiment, the step of predicting future soil and water pressure based on the most similar historical working condition sequence includes: Based on the most similar historical operating condition sequence, the target value Y corresponding to all the historical operating condition sequences is weighted and averaged using Formula 4 to obtain the target value of the i-th sample; wherein, Formula 4 is:

[0011] For the first i The weights of each sample, Let be the distance between the i-th sample and the current state vector. To prevent the denominator from being zero; Based on the target value of the i-th sample, future soil and water pressure is predicted using Formula 5; wherein, Formula 5 is:

[0012] Let i be the target value for the i-th sample. k The number of nearest neighbor samples selected.

[0013] In one embodiment, after the step of predicting future soil and water pressure based on the most similar historical working condition sequence, the method further includes: The output pressure of the external mud pump is obtained based on the predicted future earth pressure.

[0014] Based on the same technical concept, in a second aspect, the present invention also proposes a water and soil pressure prediction device, comprising: The dataset modeling module is used to construct the current dataset of soil and water pressure based on the collected historical sample data. The current dataset includes multiple sample pairs, each of which includes a feature vector X and a corresponding target value Y. The feature vector X is the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, as well as the soil and water pressure values ​​and the current speed of the pipe jacking machine over the past n time steps. The target value Y is the soil and water pressure value over the next k time steps. A vector construction module is used to construct the current state vector; wherein, the current state vector includes the soil and water pressure values ​​of the current time step and the past n time steps, and the current speed of the pipe jacking machine; The calculation module is used to calculate the current distance between the current state vector and each feature vector X in the current dataset; The selection module is used to select k samples with the smallest distance from the current dataset as the most similar historical working condition sequence; The prediction module is used to predict future soil and water pressure based on the most similar historical working condition sequence.

[0015] Based on the same technical concept, in a third aspect, the present invention also proposes a soil and water pressure prediction device, which includes a processor and a memory. The memory stores a soil and water pressure prediction program. When the soil and water pressure prediction program is executed by the processor, it implements the soil and water pressure prediction method described in the first aspect.

[0016] Based on the same technical concept, in a fourth aspect, the present invention also proposes a soil and water pressure prediction system, comprising: A pipe jacking machine, the pipe jacking machine having a driver's cab and a machine head, the driver's cab being installed inside the machine head; The water and soil pressure prediction device described in the third aspect is installed inside the cockpit; and... A mud pump is mounted on the machine head and is communicatively connected to the soil and water pressure prediction device.

[0017] Based on the same technical concept, in a fifth aspect, the present invention also proposes a computer-readable storage medium storing a computer program, which, when executed by one or more processors, implements the water and soil pressure prediction method described in the first aspect.

[0018] The technical solution of this invention constructs a current dataset of soil and water pressure based on collected historical sample data, builds a current state vector, calculates the current distance between the current state vector and each feature vector X in the current dataset, selects k samples with the smallest current distance from the current dataset as the most similar historical working condition sequence, and predicts future soil and water pressure based on the most similar historical working condition sequence. This allows the invention to predict future soil and water pressure based on the most similar historical working condition sequence in the collected historical sample data, thereby enabling the invention to make advanced and accurate predictions of soil and water pressure during the pipe jacking machine receiving process, improving the safety of the receiving operation. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0020] Figure 1 A flowchart of the water and soil pressure prediction method provided by the present invention; Figure 2 A schematic diagram of the structure of the three-circle interlocking jacking pipe pressure balance receiving and sealing chamber provided by the present invention; Figure 3 for Figure 2 A schematic diagram of the cross-sectional structure of the monitoring component layout in the example; Figure 4 for Figure 2 A schematic cross-sectional view of the door connection assembly shown in the example; Figure 5 for Figure 2 The example is a schematic diagram of the cross-sectional structure of the pilot tube sealing portal.

[0021] Figure label: 1. Monitoring components; 2. Precast tunnel portal; 3. Grouting components; 4. Water-stop airbag; 5. Waterproof gasket; 6. Bolts; 7. Receiving and sealing chamber; 8. Mud pump; 9. Circulation pipeline; 10. Shield tail brush The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0023] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a specific posture. If the specific posture changes, the directional indicators will also change accordingly.

[0024] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the use of "and / or" or "and / or" throughout the text includes three parallel solutions. For example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0025] In the traditional three-circle interlocking pipe jacking receiving stage, the portal breaching operation disrupts the water and soil pressure balance. Traditional mechanical sealing measures rely on passive sealing mechanisms, which cannot adapt to water and soil pressure fluctuations under complex geological conditions. Due to the lack of predictive ability for water and soil pressure changes, existing feedback control systems exhibit response lag, failing to adjust sealing parameters in advance when the pipe jacking machine's excavation speed, geological conditions, or external hydrological environment change. This leads to reduced sealing reliability and compromised system stability. The water and soil pressure imbalance directly affects the structural safety of the receiving shaft and the continuity of construction. When sudden pressure changes occur, the risk of seal failure increases, further threatening the overall integrity of the tunnel pre-support.

[0026] For example, during the construction of receiving wells in soft soil layers with high groundwater levels, the soil and water pressure changes drastically the moment the tunnel portal is breached as the pipe jacking machine approaches the well. Simultaneously, the pipe jacking machine's excavation speed fluctuates due to uneven geological conditions, and changes in the external hydrological environment further exacerbate these pressure fluctuations. In this scenario, the feedback control system, due to delays in data acquisition and processing, cannot respond promptly to sudden changes in soil and water pressure. The mechanical seal cannot maintain an effective seal, increasing the risk of soil erosion and leading to ground subsidence. Furthermore, this problem is particularly pronounced in sandy strata, where fluctuations in pipe jacking machine speed and changes in soil and water pressure are coupled, making it difficult to match sealing measures to actual working conditions, significantly increasing the construction risks during the receiving phase.

[0027] This invention proposes a method, apparatus, equipment, system, and medium for predicting soil and water pressure.

[0028] Please see Figures 1 to 5 To facilitate understanding, this method for predicting soil and water pressure includes the following steps: S100. Based on the collected historical sample data, construct the current dataset of soil and water pressure; wherein, the current dataset includes multiple sample pairs, each sample pair includes a feature vector X and a corresponding target value Y, the feature vector X is the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, and the soil and water pressure values ​​and the current speed of the pipe jacking machine at the past n time steps, and the target value Y is the soil and water pressure value at the future k time steps. S200. Construct the current state vector; wherein, the current state vector includes the soil and water pressure values ​​of the current time step and the past n time steps, and the current speed of the pipe jacking machine; S300. Calculate the current distance between the current state vector and each feature vector X in the current dataset; S400. Select k samples with the smallest distance from the current dataset as the most similar historical working condition sequence; S500. Based on the most similar historical working condition sequence, predict future water and soil pressure.

[0029] Specifically, firstly, a current dataset of soil and water pressure is constructed based on the collected historical sample data. The purpose of this step is to provide a data foundation for subsequent predictions. Historical sample data can be collected in various ways, such as by deploying pressure sensors and speed sensors at key locations like the pipe jacking machine head and receiving well to monitor and record soil and water pressure values ​​and the operating speed of the pipe jacking machine in real time. This raw data is stored in a database or data file. When constructing the current dataset, manual screening can be used to extract data relevant to the current prediction task from a large amount of historical data based on a preset time window or working condition type. For example, pipe jacking construction data from the past few months or years under similar geological conditions can be manually selected. This current dataset is organized into multiple sample pairs, each containing a feature vector X and a corresponding target value Y. The feature vector X can be defined as the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, as well as the combination of soil and water pressure values ​​and the speed of the pipe jacking machine over the past n time steps. The target value Y can be defined as the soil and water pressure value over the next k time steps. For example, continuous time series in historical data can be manually divided into fixed-length input sequences (feature vectors X) and corresponding future output sequences (target values ​​Y).

[0030] Construct the current state vector. This step aims to obtain the current working condition information for comparison with historical datasets. The current state vector is constructed similarly to the feature vector X, containing the soil and water pressure values ​​at the current time step and the current speed of the pipe jacking machine over the past n time steps. For example, real-time monitoring data at the current moment and historical data from the previous n time steps can be manually entered into the system to form a vector representing the current working condition.

[0031] Calculate the current distance between the current state vector and each of the corresponding feature vectors X in the current dataset. The purpose of this step is to quantify the similarity between the current state and historical states. Distance calculation can employ various mathematical methods. For example, the Manhattan distance (L1 norm) can be used to measure the difference between two vectors, which is the sum of the absolute differences of their corresponding elements. Alternatively, the Chebyshev distance (L∞ norm) can be used, where the maximum value of the absolute differences of their corresponding elements is taken as the distance. By iterating through each feature vector X in the current dataset and calculating its distance to the current state vector, a series of distance values ​​can be obtained.

[0032] From the current dataset, select k samples with the smallest distance to the current state vector as the most similar historical working condition sequence. This step aims to identify the historical scenario most similar to the current working condition. After calculating the distances between all samples and the current state vector, these distance values ​​can be sorted. For example, basic sorting algorithms such as bubble sort or selection sort can be used to arrange all distance values ​​from smallest to largest. Subsequently, select the k samples with the smallest distances from the sorted results. These samples and their corresponding target values ​​Y constitute the most similar historical working condition sequence.

[0033] Based on the most similar historical working condition sequence, future soil and water pressure is predicted. This step is the ultimate goal of the entire method: to infer future pressure changes based on the identified similar historical working conditions. Prediction can be performed in several ways. For example, one can simply take the arithmetic mean of all target values ​​Y in the most similar historical working condition sequence and use this average as the predicted future soil and water pressure. Alternatively, one can select the target value Y corresponding to the closest sample in the most similar historical working condition sequence and directly use it as the predicted future soil and water pressure.

[0034] In this embodiment, a current dataset of soil and water pressure is constructed based on the collected historical sample data. A current state vector is constructed, and the current distance between the current state vector and each feature vector X in the current dataset is calculated. The k samples with the smallest current distance from the current dataset are selected as the most similar historical working condition sequence. Based on the most similar historical working condition sequence, future soil and water pressure is predicted. This allows the present invention to predict future soil and water pressure based on the most similar historical working condition sequence in the collected historical sample data. Consequently, the present invention can perform advanced and accurate prediction of soil and water pressure during the pipe jacking machine receiving process, thereby improving the safety of the receiving operation.

[0035] In one embodiment, before the step of constructing the current dataset of soil and water pressure based on the collected historical sample data, the method further includes: Construct the feature vector X; wherein the feature vector X is represented by Formula 1, which is: Xi=[P(t),P(t 1), P(t) 2),...,P(t n),V(t)] P(t) represents the soil and water pressure value at time step t, and V(t) represents the speed of the pipe jacking machine at time step t.

[0036] Specifically, the feature vector X is a multi-dimensional data point used to describe a specific historical working condition. Its function is to transform complex time-series data into a calculable fixed-length representation, thereby quantifying the similarity between different working conditions. Formula 1 clearly defines the specific composition of the feature vector X, which combines the soil and water pressure value P(t) at the current time step t, the soil and water pressure values ​​P(t-1) to P(tn) at the past n time steps, and the pipe jacking machine speed V(t) at the current time step t. This combination method aims to comprehensively capture key operating parameters at the current and historical moments, providing sufficient information for subsequent similarity matching. Among them, the soil and water pressure value P(t) is a core parameter in the pipe jacking construction process, which can be collected in real time by sensors or obtained through averaging calculation. The pipe jacking machine speed V(t) is an important dynamic parameter affecting soil and water pressure, reflecting the progress and intensity of construction, and can also be monitored in real time by the equipment. The introduction of soil and water pressure values ​​P(t-1) to P(tn) from the past n time steps is to reflect the time dependence of soil and water pressure, that is, the current soil and water pressure is not only affected by the current working conditions, but also closely related to the previous working conditions. The parameter n is a positive integer, and its value can be determined based on actual engineering experience or through data analysis. For example, it can be set to 3, 5, or 10 time steps to fully reflect the historical trend of soil and water pressure.

[0037] In this embodiment, the feature vector X is explicitly defined to include the soil and water pressure value at the current time step, the current speed of the pipe jacking machine, and the soil and water pressure values ​​over the past n time steps. When constructing the current dataset of soil and water pressure, the feature vector X in each sample pair is precisely defined as including the soil and water pressure value P(t) at the current time step, the soil and water pressure values ​​P(t-1) to P(tn) over the past n time steps, and the speed V(t) of the pipe jacking machine at the current time step. This structured feature vector X can more comprehensively characterize the working condition at a specific moment because it not only considers instantaneous data but also incorporates historical evolution information. When subsequent steps calculate the distance between the current state vector and these feature vectors X, this feature vector containing temporal information ensures that the similarity measurement is more accurate, thereby more accurately identifying historical working condition sequences similar to the current working condition development trend. This explicit construction of the feature vector X enables the system to capture the dynamic correlation between soil and water pressure and the speed of the pipe jacking machine, as well as the temporal dependence of soil and water pressure, providing a solid foundation for subsequent predictions based on similar historical working conditions.

[0038] In one embodiment, before the step of constructing the current dataset of soil and water pressure based on the collected historical sample data, the method further includes: Construct a second equation to describe the target value Y; wherein the second equation is: Yi = P(t+k) Where P(t+k) represents the predicted soil and water pressure for the next k steps.

[0039] Specifically, a data preprocessing module can be used to extract sample pairs from the original historical data, directly extracting P(t+k) as the target value Yi from the original time series based on the current time step t and the preset future step size k. Alternatively, during the data storage or database design phase, the soil and water pressure value P(t+k) corresponding to each feature vector X at the k-th future time step can be stored as an independent field or associated data for direct reading and use later. P(t+k) is the value of soil and water pressure at the k-th future time step. This definition further emphasizes that the focus of prediction is the soil and water pressure at a single time point (t+k), rather than the average or sequence over a period of time. This is crucial for precise control of pipe jacking machine operation, as it provides key pressure information for a specific future moment. P(t+k) can be the soil and water pressure measurement value at time step t+k collected by sensors in actual engineering. Alternatively, P(t+k) can also be the soil and water pressure value estimated at time step t+k based on historical data trends using interpolation or extrapolation methods, used to construct training samples.

[0040] In this embodiment, when constructing the current dataset of soil and water pressure, the system first uses the collected historical sample data according to formula 1 Xi=[P(t),P(t)] 1), P(t) 2),...,P(t The feature vector X for each sample pair is constructed using [n), V(t)]. Based on this, to ensure the clarity and consistency of the target value Y, this application introduces a second equation, Yi = P(t+k), precisely defining the target value Y as the soil and water pressure value P(t+k) at the k-th time step after the current time step t. This means that when extracting each sample pair from historical data, the system accurately locates and extracts the soil and water pressure value P(t+k) at time step t+k from the original time series based on the current time step t and the preset future prediction step size k, as the target value for that sample. This clear definition ensures that each sample pair in the dataset has a clear and consistent correspondence between input (feature vector X) and output (target value Y), thereby providing high-quality training data for subsequent prediction algorithms, enabling them to specifically learn the mapping relationship from current and historical working condition characteristics to soil and water pressure at a specific future time point.

[0041] In one embodiment, the step of calculating the current distance between the current state vector and each of the feature vectors X in the current dataset includes: Formula 3 is used to calculate the current distance between the current state vector and each feature vector X in the current dataset; wherein, Formula 3 is:

[0042] Let X represent the distance between the i-th sample and the current state vector, X represent the current state vector, Xi represent the feature vector of the i-th sample, and m represent the dimension of the feature vector.

[0043] Specifically, Formula 3 uses Euclidean distance. Euclidean distance is a commonly used metric for measuring the "straight-line" distance between two points in a multidimensional space. Its core idea is to quantify the overall difference by calculating the square root of the sum of the squares of the differences in each dimension. In this application, its role is to provide a quantitative and objective means to assess the similarity between the current operating state of the pipe jacking machine and historical sample operating conditions. A smaller Euclidean distance value intuitively indicates that the current operating condition is closer to a certain historical sample operating condition in multiple feature dimensions such as soil and water pressure and pipe jacking machine speed, that is, the two have a higher similarity. Here, X represents the operating state of the pipe jacking machine at the current moment. It includes the soil and water pressure values ​​of the current time step and the current speed of the pipe jacking machine in the past n time steps, and is a multidimensional vector. Xi represents the feature vector of the i-th sample in the historical dataset. Its structure is consistent with X and is used to compare it with the current state one by one. m represents the dimension of the feature vector, that is, the number of independent features that constitute each state or sample. For example, if the feature vector contains the current soil and water pressure, the soil and water pressure over the past n times, and the current velocity, then the value of m can be n+2. This distance calculation method can be implemented in various ways. For instance, it can be implemented through software programming, utilizing mathematical library functions in various programming languages ​​(such as Python, Java, C++, etc.) to operate on each vector element individually, performing square difference, summation, and square root operations. Furthermore, for scenarios requiring the processing of massive amounts of data or with high real-time requirements, hardware acceleration solutions can be considered. For example, leveraging the parallel computing capabilities of field-programmable gate arrays (FPGAs) or graphics processing units (GPUs) can parallelize the Euclidean distance calculation task, thereby significantly improving computational efficiency.

[0044] In this embodiment, Euclidean distance is introduced as a distance metric between the current state vector and the feature vectors of historical samples. This enables a comprehensive and accurate similarity assessment of multi-dimensional working condition features based on the construction of the current dataset, feature vector X, and target value Y. During the method execution, the current dataset containing feature vector X and target value Y is first constructed based on historical sample data. Feature vector X is constructed according to Equation 1, and target value Y is constructed according to Equation 2. Subsequently, the system obtains the soil and water pressure values ​​at the current time step and the current speed of the pipe jacking machine over the past n time steps, forming the current state vector. Then, using Equation 3, the current state vector is compared one by one with the feature vector X of each historical sample in the current dataset. The Euclidean distance calculation mechanism ensures that the differences in soil and water pressure values ​​and pipe jacking machine speed at different time steps are taken into account, and the influence of larger differences is amplified through the sum of squares, allowing the distance value to accurately reflect the overall deviation of the two working conditions in the multi-dimensional feature space. This precise distance calculation provides a solid foundation for selecting the k samples with the smallest distance from the current dataset as the most similar historical working condition sequence, thus ensuring a high degree of matching between the selected historical working condition sequence and the current actual working condition. In this way, this application effectively solves the problem of how to accurately identify the historical experience data most similar to the current working condition in a complex and variable soil and water pressure environment, laying the foundation for reliable future soil and water pressure prediction based on these similar historical working conditions.

[0045] In one embodiment, the step of predicting future soil and water pressure based on the most similar historical working condition sequence includes: Based on the most similar historical operating condition sequence, the target value Y corresponding to all the historical operating condition sequences is weighted and averaged using Formula 4 to obtain the first... The target value for each sample; where formula four is:

[0046] Let i be the weight of the i-th sample. Let be the distance between the i-th sample and the current state vector. To prevent the denominator from being zero; Based on the target value of the i-th sample, future soil and water pressure is predicted using Formula 5; wherein, Formula 5 is:

[0047] Yi is the target value of the i-th sample, and k is the number of nearest neighbor samples selected.

[0048] Specifically, weighted averaging of the target value Y corresponding to all historical working condition sequences is a statistical processing method. This method aims to calculate the average value by assigning different weights to different samples, thereby reflecting their relative importance in the overall picture. In this scheme, its role is to make historical working condition sequences (i.e., those more similar to the current state vector) more similar to the current state vector. Smaller (or less) values ​​have a larger weight in the prediction results, thus effectively improving the accuracy of the prediction. This weighted averaging process can be implemented in various ways; for example, a weighting function can be preset, which is based on distance. Dynamically generate weights based on size This ensures that the smaller the distance, the greater the weight; alternatively, a kernel-based method can be used to map the distance to the weight, such as the Gaussian kernel function, so that samples that are closer to each other have higher weights.

[0049] Obtain the The target value for each sample refers to the original target value extracted from the most similar historical operating condition sequence during the weighted average calculation process. Y These target values ​​are the foundational data for predicting future soil and water pressure. By weighted averaging, the future trends of multiple historical samples can be synthesized. In practice, the original target value for each sample can be directly extracted from the most similar historical working condition sequence selected from the existing data. Y Alternatively, in certain application scenarios, the original target value for each sample can also be used. Y Preprocessing or normalization is performed before weighted averaging.

[0050] Formula 4: The weights are defined. The calculation method is an inverse distance weighting function. The core function of this formula is to ensure the distance between the sample and the current state vector. The smaller the value, the higher the calculated weight. The larger the value, the greater its contribution to the final prediction result in the subsequent weighted average calculation. (Constant) The introduction of this is to avoid when the distance When the denominator is zero, it becomes infinitely large, thus ensuring the numerical stability and validity of the calculation process. In practical applications, this formula can be calculated in the data processing module after obtaining the distance for each sample. Then, directly apply this formula to calculate the corresponding weights. Alternatively, it can be done by looking up a table or pre-calculating, based on... Quickly obtain the corresponding range To improve computational efficiency.

[0051] Formula 5: This is the standard weighted average formula used to calculate the final predicted value. The function of this formula is to select all... k The weighted target value of the most similar historical operating condition sequence The summation is performed and divided by the sum of all weights to obtain a future water and soil pressure prediction that comprehensively considers similarity differences. In this way, the final prediction result can more accurately reflect the future trend indicated by historical data most similar to the current working conditions. In implementation, the prediction module can receive the weights of each sample. and target value Then, calculate according to Formula 5; in order to improve calculation efficiency, especially... When the value is large, the calculation process of Formula 5 can be accelerated by parallel computing or vectorization operations.

[0052] In this embodiment, by determining the vector most similar to the current state vector... Following the historical load condition series, a weighted average mechanism is further introduced to optimize the prediction of soil and water pressure. Specifically, for each selected historical load condition series, its weight is first calculated using Formula 4. The formula is based on the distance between the historical operating condition sequence and the current state vector. This allows sequences with smaller distances to receive greater weights, thus playing a more important role in prediction. Subsequently, these calculated weights are... Target value corresponding to each historical working condition sequence The products are summed and divided by the sum of all weights, i.e., a weighted average is calculated using Formula 5. This method ensures the final predicted future soil and water pressure. This approach not only considers historical similarity but also more precisely reflects the impact of historical data with varying degrees of similarity on current predictions, making the prediction results closer to actual working conditions. In this way, this application effectively addresses the potential inaccuracies of simple predictions and improves the precision of soil and water pressure prediction.

[0053] In one embodiment, after the step of predicting future soil and water pressure based on the most similar historical working condition sequence, the method further includes: The output pressure of the external mud pump is obtained based on the predicted future earth pressure.

[0054] Specifically, this application further proposes that after predicting future soil and water pressure based on the most similar historical working condition sequence, it also includes obtaining the output pressure of the external mud pump based on the predicted future soil pressure. The predicted future soil pressure is an estimate of the soil and water pressure at a specific future time step calculated through the aforementioned steps. It serves as input for subsequent control decisions, reflecting the impending soil pressure environment. The output pressure of the external mud pump is the specific pressure value required to be generated by the mud pump to balance the soil and water pressure at the excavation face of the pipe jacking machine. Its precise control is crucial for maintaining the stability of the excavation face. Obtaining this output pressure means determining the appropriate output pressure of the mud pump based on the predicted future soil and water pressure through a certain mechanism or algorithm. This can be achieved through a preset mapping relationship, empirical formula calculation, or a dynamic adjustment strategy based on real-time working condition parameters. For example, a control strategy based on engineering experience or theoretical calculation can be preset. This strategy can be a function such that the output pressure of the external mud pump equals the predicted future soil and water pressure plus a preset compensation pressure value, used to ensure that the pressure at the excavation face is slightly higher than the soil and water pressure to maintain stability. Alternatively, a lookup table could be included to directly provide the corresponding output pressure value of the external mud pump based on different predicted future water and soil pressure ranges.

[0055] In this embodiment, after predicting future soil and water pressure, the prediction result is further used as input to directly determine the output pressure of the external mud pump. This design transforms soil and water pressure prediction from merely an information output into a direct control command for the construction equipment (external mud pump). By establishing a direct correlation between the predicted value and the output pressure of the mud pump, the system can achieve proactive balancing of soil and water pressure at the excavation face. When a change in future soil and water pressure is predicted, the system can calculate and adjust the output pressure of the mud pump in advance, thus preparing for the eventual change. This proactive control mechanism enables the pipe jacking machine to traverse complex strata more smoothly, effectively avoiding risks such as excavation face instability, surface subsidence, or sudden surges caused by pressure imbalance, significantly improving the safety, stability, and efficiency of construction.

[0056] More specifically, the present invention can also be implemented in the following manner: First, the present invention proposes a three-circle interlocking jacking pipe pressure balance receiving sealing chamber.

[0057] The receiving sealed chamber includes a mechanical seal assembly, a door connection assembly, a pressure balance receiving assembly, and a monitoring assembly.

[0058] The mechanical seal assembly is used to passively prevent water inflow during the exit phase of the pipe jacking machine. The mechanical seal assembly includes a prefabricated portal, a water-stop airbag, and a grouting assembly. The prefabricated portal is pre-installed at the exit point of the pipe jacking machine. The water-stop airbag is located inside the prefabricated portal and close to the starting end of the pipe jacking. Its expansion presses against the pipe jacking head, providing initial water stoppage. The water-stop airbag connects to the subsequent receiving sealing chamber. By adjusting the pressure inside the receiving sealing chamber, a tight fit between the water-stop airbag and the pipe jacking head can be ensured, effectively filling the gap between the machine head and the prefabricated portal, preventing groundwater from flowing out. The grouting assembly fills the gap between the interlocking pipe segments and the ground, preventing water inflow at the portal due to gaps during subsequent disassembly of the mechanical seal assembly, portal connection assembly, and pressure balance receiving assembly, forming an additional safety barrier.

[0059] The door / cabin connecting assembly is used to connect the mechanical seal assembly and the pressure-balanced receiving assembly, ensuring a watertight seal at the connection point. The door / cabin connecting assembly includes bolts and a watertight gasket. The watertight gasket is positioned at the flange connection surface between the precast portal and the receiving sealed cabin, and is tightened by bolts. The compression deformation of the gasket effectively prevents mud and water from gushing out from the connection between the receiving sealed cabin and the precast portal.

[0060] The pressure balancing receiving assembly is used to actively balance the water and soil pressure, achieve dynamic water stoppage during the receiving process, and complete the receiving of the jacking head and transition pipe. The pressure balancing receiving assembly includes a receiving sealed chamber, a mud pump, a circulation pipeline, a pilot pipe sealing portal, and a pressure regulating system. The receiving sealed chamber is a sealed steel chamber filled with mud, with a receiving platform at its bottom for final placement and receiving of the jacking head and transition pipe. The mud pump is connected to the receiving sealed chamber through the circulation pipeline and is controlled by the pressure regulating system. The pressure regulating system controls the output pressure of the mud pump to fill the receiving sealed chamber with mud and maintain a certain pressure inside, ensuring that the mud-water pressure inside the receiving sealed chamber is balanced with the mud-water pressure outside. This balancing pressure also acts on the water-stopping airbag, ensuring its continuous contact with the jacking head and playing a role in active water stoppage. The pilot tube sealing portal is an openable and closable sealing door on the receiving sealing chamber. Its inner side is equipped with a 360-degree shield tail brush to adhere to the pilot tube, ensuring the sealing effect of the receiving sealing chamber during the pilot tube reception process, thus serving as a single-tube water stop. Once the pilot tube arrives and the shield tail brush is fully engaged, the pilot tube sealing portal can be opened for subsequent operations.

[0061] The monitoring components are used to collect feedforward feature datasets and real-time feature datasets, providing data support for AI prediction models. The monitoring components include a soil and water pressure monitoring unit, a pipe jacking machine status monitoring unit, a geological and hydrological monitoring unit, and a data acquisition and communication unit. The soil and water pressure monitoring unit includes soil pressure sensors and pore water pressure gauges arranged on the outer wall of the precast tunnel portal, the outer wall of the receiving sealed chamber, and in front of the pipe jacking machine head, used to collect total soil and water pressure and pore water pressure in real time. The pipe jacking machine status monitoring unit is used to collect real-time tunneling speed and thrust data of the pipe jacking machine. The geological and hydrological monitoring unit is used to collect groundwater level data around the receiving well and can be combined with equipment such as ground-penetrating radar to obtain information about the soil ahead. The data acquisition and communication unit is used to collect all monitoring data, preprocess it, and transmit it to the pressure regulation system.

[0062] like Figure 2 As shown, the receiving sealing chamber system of the present invention is installed in the receiving well. The prefabricated portal 2 in the mechanical seal assembly is prefabricated in the factory and fixed to the opening during the construction of the receiving well. Before the pipe jacking machine exits the well, the water-stopping airbag 4 is activated to initially inflate it. Subsequently, quick-setting grout is injected into the gap between the interlocking pipe segments and the stratum through the grouting assembly 3 to form a ring-shaped water-stopping ring.

[0063] like Figure 3 As shown, the pressure balancing receiving assembly is reliably connected to the mechanical seal assembly via the door chamber connecting assembly. During installation, first, the water-proof gasket 5 is placed on the flange face of the prefabricated door 2, then the flange of the receiving sealing chamber 7 is aligned with it, and high-strength bolts 6 are tightened evenly in a diagonal sequence to compress the water-proof gasket 5 and form a reliable seal.

[0064] The pressure balancing receiving assembly is the core functional unit. The receiving sealed chamber 7 is a sealed chamber welded from thick steel plates, and its internal volume is determined according to the dimensions of the pipe jacking head and the transition pipe. The mud pump 8 (preferably a frequency-controlled mud pump) is connected to the bottom and top of the receiving sealed chamber 7 through circulation pipelines 9, forming a mud circulation loop, which helps to maintain the uniformity of mud and the stability of pressure inside the chamber. The pilot tube sealing portal is opened and closed hydraulically or electrically, and the shield tail brush 10 installed on its inner side adopts a combination of wire brush and rubber plate to ensure a tight fit with the outer wall of the pilot tube.

[0065] The monitoring component 1 is the "sensory nerve" of the system. For example... Figure 3As shown, the soil pressure sensor and pore water pressure gauge are respectively installed on the outer surface of the steel structure of the precast tunnel portal 2, the outer wall of the receiving sealed chamber 7, and in the strata along the direction of the pipe jacking machine's advance via pre-embedding. The pipe jacking machine's status data is acquired through its own PLC system. Groundwater level monitoring is achieved through a water level observation well located near the receiving well. All sensor data is collected and converted from analog to digital by a data acquisition and communication unit (e.g., an industrial IoT gateway), and transmitted to the pressure regulation system in the central control room via industrial Ethernet or a wireless network. The pressure regulation system is essentially an industrial computer (IPC) or a high-performance PLC, running the program for the soil and water pressure prediction and control algorithm.

[0066] Secondly, this invention provides a method for predicting and controlling water and soil pressure in the aforementioned receiving sealed chamber. The core of this method is the use of a spatiotemporal feature similarity matching and adaptive weighted prediction algorithm based on this invention.

[0067] The method includes the following steps: Step 1: Collect historical sample data and construct a training set. The training set includes multiple sample pairs, each sample pair including a spatiotemporal feature vector X and a corresponding target value Y. The feature vector X includes the soil and water pressure values ​​at the current time step and the past n time steps, as well as the current speed of the pipe jacking machine. The target value Y is the soil and water pressure value at the next k time steps.

[0068] Step 2: Construct the current spatiotemporal feature vector. The current state vector includes the soil and water pressure values ​​at the current time step and the past n time steps, as well as the current speed of the pipe jacking machine.

[0069] Step 3: Similarity matching: Calculate the distance between the current state vector and the feature vector X of each sample in the training set.

[0070] Step 4: Select k samples from the training set that are closest to the current spatiotemporal feature vector as the most similar historical working condition sequence.

[0071] Step 5: Adaptive Weighted Prediction: Based on the selected k samples, calculate the predicted value of future soil and water pressure. Specifically, a weighted average is calculated for the target value Y of each sample, with the weight being the reciprocal of the distance between the corresponding sample and the current state vector.

[0072] Step Six: Multiply the predicted future water and soil pressure value by the dynamic safety factor λ to obtain the final predicted value W, and send it to the pressure regulation system in the receiving sealed chamber. The pressure regulation system adjusts the output pressure of the mud pump to maintain a dynamic balance between the water and soil pressure in the receiving sealed chamber and the predicted value W.

[0073] In step one, the feature vector X is represented as: the target value Y is represented as: Yi = P(t+k), where P(t+k) represents the predicted water and soil pressure in the next k steps.

[0074] In step three, the distance is calculated using Euclidean distance:

[0075] Where di represents the distance between the i-th sample and the current state vector, X represents the current state vector, Xi represents the feature vector of the i-th sample, and m represents the dimension of the feature vector.

[0076] In step five, the weight ωi is calculated as follows:

[0077] Where ε is a small constant to prevent the denominator from being zero.

[0078] The predicted value P is calculated as follows:

[0079] In step six, the safety factor λ is calculated as follows: λ = λ0+ α · Cw + β · (1 - Cs) + γ · Rd Wherein, λ0 is the basic safety factor, which is a constant greater than or equal to 1; Cw is the hydrological risk factor calculated based on real-time monitoring data; Cs is the soil stability factor determined based on geological data; Rd is the construction dynamic risk factor calculated based on the tunneling speed of the pipe jacking machine; α, β, γ are the weights corresponding to each coefficient, which are determined through laboratory data or historical data.

[0080] The hydrological risk coefficient Cw is calculated as follows:

[0081] Where Wc is the current groundwater level, Wavg is the annual average groundwater level, and Wmax is the highest groundwater level in historical statistics.

[0082] The value of the soil stability coefficient Cs is determined according to the geological conditions: when the soil ahead is determined to be a homogeneous cohesive soil layer, Cs is 0.8~1.0; when the soil ahead is determined to be a silt or silty sand layer, Cs is 0.5~0.7; when the soil ahead is determined to be a water-rich sandy gravel layer or there is an obvious geological anomaly zone, Cs is 0~0.4.

[0083] The construction dynamic risk coefficient The calculation method is as follows:

[0084] Where Vc is the current tunneling speed of the pipe jacking machine, and Vavg is the average speed under stable tunneling conditions.

[0085] The "Spatiotemporal Feature Similarity Matching and Adaptive Weighted Prediction Algorithm" proposed in this invention is not a simple improvement on existing general algorithms, but an original algorithm specifically designed to address the challenge of time-series prediction of soil and water pressure in underground engineering. It finds the most similar historical working conditions through high-dimensional spatiotemporal feature matching and uses an adaptive weighting mechanism to extrapolate the future, achieving true feedforward intelligent control and overcoming the inherent lag of traditional feedback control.

[0086] This algorithm innovatively integrates the dynamic safety factor λ with a data-driven prediction model. λ is not a fixed value, but a dynamic variable that integrates real-time monitoring information from multiple sources, including hydrology (Cw), geology (Cs), and construction (Rd), enabling the prediction system to possess both the high adaptability of data-driven approaches and the strong robustness of theoretical mechanisms.

[0087] Designed specifically for complex structures such as three-circle interlocking jacking pipes, it shifts the water inrush point from the inside corner of the three-circle interlocking jacking pipe to the sealed entrance of the pilot pipe and traditional waterproofing methods. Its sealing chamber structure and control logic can effectively cope with the unique contour changes and leakage risks at the multi-pipe interlocking interface of this type of construction method.

[0088] Based on the same technical concept, in a second aspect, the present invention also proposes a water and soil pressure prediction device, comprising: The dataset modeling module is used to construct the current dataset of soil and water pressure based on the collected historical sample data. The current dataset includes multiple sample pairs, each of which includes a feature vector X and a corresponding target value Y. The feature vector X is the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, as well as the soil and water pressure values ​​and the current speed of the pipe jacking machine over the past n time steps. The target value Y is the soil and water pressure value over the next k time steps. A vector construction module is used to construct the current state vector; wherein, the current state vector includes the soil and water pressure values ​​of the current time step and the past n time steps, and the current speed of the pipe jacking machine; The calculation module is used to calculate the current distance between the current state vector and each feature vector X in the current dataset; The selection module is used to select k samples with the smallest distance from the current dataset as the most similar historical working condition sequence; The prediction module is used to predict future soil and water pressure based on the most similar historical working condition sequence.

[0089] The soil and water pressure prediction device provided in this application, employing the soil and water pressure prediction method described in the above embodiments, can solve the problem in the prior art where the feedback control system cannot respond in advance to sudden pressure changes when the tunneling speed of the pipe jacking machine, geological conditions, or external hydrological environment changes, and there is still a risk of instability. Compared with the prior art, the beneficial effects of the soil and water pressure prediction device provided in this application are the same as those of the soil and water pressure prediction method provided in the above embodiments, and other technical features in the soil and water pressure prediction device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0090] Based on the same technical concept, in a third aspect, the present invention also proposes a soil and water pressure prediction device, which includes a processor and a memory. The memory stores a soil and water pressure prediction program. When the soil and water pressure prediction program is executed by the processor, it implements the soil and water pressure prediction method described in the first aspect.

[0091] The soil and water pressure prediction device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), vehicle terminals (such as vehicle control terminals), and fixed terminals such as digital TVs and desktop computers.

[0092] The soil and water pressure prediction device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the soil and water pressure prediction device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the soil and water pressure prediction device to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows soil and water pressure prediction devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0093] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0094] The soil and water pressure prediction device provided in this application, employing the soil and water pressure prediction method described in the above embodiments, can solve the problem in the prior art where the feedback control system cannot respond in advance to sudden pressure changes when the tunneling speed of the pipe jacking machine, geological conditions, or external hydrological environment changes, and there is still a risk of instability. Compared with the prior art, the beneficial effects of the soil and water pressure prediction device provided in this application are the same as those of the soil and water pressure prediction method provided in the above embodiments, and other technical features of this soil and water pressure prediction device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0095] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0096] Based on the same technical concept, in a fourth aspect, the present invention also proposes a soil and water pressure prediction system, comprising: A pipe jacking machine, the pipe jacking machine having a driver's cab and a machine head, the driver's cab being installed inside the machine head; The water and soil pressure prediction device described in the third aspect is installed inside the cockpit; and... A mud pump is mounted on the machine head and is communicatively connected to the soil and water pressure prediction device.

[0097] The soil and water pressure prediction system provided in this application, employing the soil and water pressure prediction method described in the above embodiments, can solve the problem in the prior art where the feedback control system cannot respond in advance to sudden changes in pressure when the tunneling speed of the pipe jacking machine, geological conditions, or external hydrological environment changes, and there is still a risk of instability. Compared with the prior art, the beneficial effects of the soil and water pressure prediction system provided in this application are the same as those of the soil and water pressure prediction method provided in the above embodiments, and other technical features of this soil and water pressure prediction system are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0098] Based on the same technical concept, in a fifth aspect, the present invention also proposes a computer-readable storage medium storing a computer program, which, when executed by one or more processors, implements the water and soil pressure prediction method described in the first aspect.

[0099] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0100] The aforementioned computer-readable storage medium may be included in the soil and water pressure prediction device; or it may exist independently and not be assembled into the soil and water pressure prediction device.

[0101] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the soil and water pressure prediction device, enable the soil and water pressure prediction device to implement the soil and water pressure prediction method described above.

[0102] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0103] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0104] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0105] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described soil and water pressure prediction method. This addresses the problem in existing technologies where the feedback control system cannot anticipate sudden pressure changes when the tunneling speed of the pipe jacking machine, geological conditions, or the external hydrological environment changes, thus still facing the risk of instability. Compared to existing technologies, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the soil and water pressure prediction method provided in the above embodiments, and will not be elaborated upon here.

[0106] The above description is merely an exemplary embodiment of the present invention and does not limit the scope of the present invention. Any equivalent structural transformations made based on the technical concept of the present invention and the contents of the specification and drawings of the present invention, or direct / indirect applications in other related technical fields, are included within the protection scope of the present invention.

Claims

1. A method of predicting water and soil pressure, characterized by, Includes the following steps: Based on the collected historical sample data, a current dataset of soil and water pressure is constructed; wherein, the current dataset includes multiple sample pairs, each sample pair includes a feature vector X and a corresponding target value Y, wherein the feature vector X is the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, as well as the soil and water pressure value and the current speed of the pipe jacking machine at the past n time steps, and the target value Y is the soil and water pressure value at the future k time steps. Construct the current state vector; wherein the current state vector includes the soil and water pressure values ​​of the current time step and the past n time steps, and the current speed of the pipe jacking machine; Calculate the current distance between the current state vector and each of the feature vectors X in the current dataset; Select k samples with the smallest distance from the current dataset as the most similar historical working condition sequence; Based on the most similar historical working condition sequence, predict future soil and water pressure.

2. The water and soil pressure prediction method according to Claim 1, wherein Before the step of constructing the current dataset of soil and water pressure based on the collected historical sample data, the method further includes: Construct the feature vector X; wherein the feature vector X is represented by Formula 1, which is: Xi=[P(t),P(t 1),P(t 2),...,P(t n),V(t)] P(t) represents the soil and water pressure value at time step t, and V(t) represents the speed of the pipe jacking machine at time step t.

3. The water and soil pressure prediction method according to claim 2, wherein Before the step of constructing the current dataset of soil and water pressure based on the collected historical sample data, the method further includes: Construct a second equation to describe the target value Y; wherein the second equation is: Yi = P(t+k) Where P(t+k) represents the predicted soil and water pressure for the next k steps.

4. The water and soil pressure prediction method according to claim 3, wherein The step of calculating the current distance between the current state vector and each feature vector X in the current dataset includes: Formula 3 is used to calculate the current distance between the current state vector and each feature vector X in the current dataset; wherein, Formula 3 is: represents the distance of the i-th sample to the current state vector, represents the current state vector, represents the feature vector of the i-th sample, m represents the dimension of the feature vector.

5. The water and soil pressure prediction method according to claim 4, characterized by, The step of predicting future soil and water pressure based on the most similar historical working condition sequence includes: Based on the most similar historical operating condition sequence, the target value Y corresponding to all the historical operating condition sequences is weighted and averaged using Formula 4 to obtain the target value of the i-th sample; wherein, Formula 4 is: Let i be the weight of the i-th sample. Let be the distance between the i-th sample and the current state vector. To prevent the denominator from being zero; Based on the target value of the i-th sample, future soil and water pressure is predicted using Formula 5; wherein, Formula 5 is: Yi is the target value of the i-th sample, and k is the number of nearest neighbor samples selected.

6. The method for predicting soil and water pressure as described in any one of claims 1 to 5, characterized in that, Following the step of predicting future soil and water pressure based on the most similar historical working condition sequence, the method further includes: The output pressure of the external mud pump is obtained based on the predicted future earth pressure.

7. A water and soil pressure prediction device, characterized in that, include: The dataset modeling module is used to construct the current dataset of soil and water pressure based on the collected historical sample data. The current dataset includes multiple sample pairs, each of which includes a feature vector X and a corresponding target value Y. The feature vector X is the soil and water pressure value and the current speed of the pipe jacking machine at the current time step, as well as the soil and water pressure values ​​and the current speed of the pipe jacking machine over the past n time steps. The target value Y is the soil and water pressure value over the next k time steps. A vector construction module is used to construct the current state vector; wherein, the current state vector includes the soil and water pressure values ​​of the current time step and the past n time steps, and the current speed of the pipe jacking machine; The calculation module is used to calculate the current distance between the current state vector and each feature vector X in the current dataset; The selection module is used to select k samples with the smallest distance from the current dataset as the most similar historical working condition sequence; The prediction module is used to predict future soil and water pressure based on the most similar historical working condition sequence.

8. A water and soil pressure prediction device, characterized in that, The soil and water pressure prediction device includes a processor and a memory. The memory stores a soil and water pressure prediction program. When the processor executes the soil and water pressure prediction program, it implements the soil and water pressure prediction method as described in any one of claims 1 to 6.

9. A water and soil pressure prediction system, characterized in that, include: A pipe jacking machine, the pipe jacking machine having a driver's cab and a machine head, the driver's cab being installed inside the machine head; The soil and water pressure prediction device as described in claim 8 is installed inside the cockpit; as well as, A mud pump is mounted on the machine head and is communicatively connected to the soil and water pressure prediction device.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by one or more processors, implements the water and soil pressure prediction method as described in any one of claims 1 to 6.

Images