Monitoring and Analysis Method for Time Effect of Soil Displacement Characteristics of Jacked Piles in Soft Soil Areas

Abstract

The present invention relates to the technical field of construction engineering construction monitoring in soft soil areas, and discloses a monitoring and analysis method for the time effect of the soil extrusion characteristics of static pressure piles in soft soil areas. This method arranges a multi-source sensor array in the construction area to collect signals such as soil displacement and pore water pressure. After preprocessing such as noise filtering, time series alignment, and data normalization, a finite element model of layered soil is established. A mathematical model associating the soil extrusion effect with time is constructed by combining time series analysis. A prediction model based on machine learning is used to predict the soil response, and an optimization algorithm is used to adjust the model parameters. Whether to converge is judged according to the model prediction error, and then the sensor array is dynamically adjusted to form a closed-loop iterative analysis process. This method can comprehensively and accurately monitor and analyze the time effect of the soil extrusion characteristics of static pressure piles in soft soil areas, provide a scientific basis for construction decision-making, and ensure the safety and quality of the project.

Description

Technical Field

[0001] The present invention relates to the technical field of construction engineering construction monitoring in soft soil areas, and specifically to a method for monitoring and analyzing the time effect of the soil squeezing characteristics of static pressure piles in soft soil areas. Background Art

[0002] When carrying out construction in soft soil areas, static pressure piles are widely used due to their advantages such as low noise, no vibration, and high construction efficiency. However, during the construction process of static pressure piles, significant soil squeezing effects will be generated, which have an adverse impact on the stability of the surrounding soil and facilities such as adjacent buildings and underground pipelines. Soft soil has characteristics such as high water content, large compressibility, low strength, and poor permeability, which makes the soil squeezing effect of static pressure piles more complex in soft soil areas. During construction, the penetration of the pile body will cause displacement and deformation of the surrounding soil, and the pore water pressure will rise sharply and dissipate slowly. This soil displacement may cause settlement and inclination of the foundation of adjacent buildings, and even structural damage; the change of pore water pressure may also change the effective stress state of the soil, reduce the shear strength of the soil, and increase the construction safety risk.

[0003] At present, there are many deficiencies in the monitoring and analysis methods for the soil squeezing characteristics of static pressure piles in soft soil areas. Traditional monitoring means often rely on single-type sensors and cannot comprehensively obtain information on multiple aspects such as soil displacement, pore water pressure, pile penetration resistance, and vibration, making it difficult to accurately grasp the overall picture of the soil squeezing effect. Moreover, the collected data is easily interfered by noise, the time series of data from different sensors is inconsistent, and the data processing method is simple and rough, resulting in poor accuracy and reliability of the monitoring data.

[0004] In terms of the analysis method, existing models are mostly simplified models and do not fully consider the complexity and dynamic change characteristics of soil layers in soft soil areas. These models cannot accurately simulate the dynamic changes of parameters such as elastic modulus, Poisson's ratio, and permeability coefficient of soil layers at different depths, and the description of the correlation between the soil squeezing effect and time is not accurate enough, making it difficult to effectively predict the soil response in the subsequent construction stage.

[0005] In addition, during the construction process, due to the lack of effective dynamic monitoring and analysis means, it is impossible to adjust construction parameters and monitoring strategies in a timely manner according to the actual construction situation. Once problems occur, often only remedial measures can be taken afterwards, which not only increases the project cost but also may delay the construction period, seriously affecting the project quality and construction safety. With the continuous advancement of urban construction in soft soil areas, the scale and complexity of construction projects are increasing day by day, and the need for accurate monitoring and analysis of the soil squeezing characteristics of static pressure piles in soft soil areas is becoming more urgent. There is an urgent need for a more advanced and effective monitoring and analysis method to solve these problems. Summary of the Invention

[0006] The object of the present invention is to provide a method for monitoring and analyzing the time effect of the soil squeezing characteristics of static pressure piles in soft soil areas, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solution: A method for monitoring and analyzing the time effect of the soil squeezing characteristics of static pressure piles in soft soil areas, the method comprising:

[0008] STEP1: Arrange a multi-source sensor array in the construction area of the static pressure pile, and collect the soil displacement, pore water pressure, pile penetration resistance and vibration signals during the construction process in real time;

[0009] STEP2: Preprocess the collected original data, including noise filtering, time series alignment and data normalization, to generate a standardized monitoring data set;

[0010] STEP3: Based on the standardized monitoring data set, establish a finite element model of layered soil, and the model includes the dynamic parameters of the elastic modulus, Poisson's ratio and permeability coefficient of different depth soil layers in soft soil areas;

[0011] STEP4: Combine the time series analysis method to extract the time-varying characteristics of soil displacement and pore water pressure, and construct a mathematical model of the correlation between soil squeezing effect and time;

[0012] STEP5: Adopt a prediction model based on machine learning, input the time-varying characteristics of STEP4 and the parameters of the layered soil model of STEP3, and predict the soil response in the subsequent construction stage;

[0013] STEP6: Adjust the dynamic parameters of the layered soil model through an optimization algorithm to minimize the error between the prediction result and the actual monitoring data;

[0014] STEP7: Judge whether the prediction error of the model after parameter optimization meets the convergence condition; if it meets, output the optimized soil parameters and the time evolution law of the soil squeezing effect; if it does not meet, return to STEP6 to re-optimize;

[0015] STEP8: Dynamically adjust the sampling frequency and spatial distribution of the sensor array according to the output result of STEP7;

[0016] STEP9: Feed the updated monitoring data back into the prediction model of STEP5 to form a closed-loop iterative analysis process.

[0017] Preferably, in the STEP1, the multi-source sensor array includes distributed optical fiber sensors, piezoelectric pore water pressure gauges, strain displacement gauges and acceleration sensors; the sensors are arranged in layers along the axis of the pile, and the radial arrangement density changes with the gradient of the soil layer permeability coefficient.

[0018] Preferably, in the STEP2, wavelet threshold denoising algorithm is adopted for noise filtering, dynamic time warping algorithm is adopted for time series alignment, and piecewise standardization method based on soil layer depth is adopted for data normalization.

[0019] Preferably, in the STEP3, the dynamic parameters of the layered soil finite element model are corrected in real time through a backpropagation neural network. The input is the standardized monitoring data of STEP2, and the output is the elastic modulus correction coefficient and permeability coefficient attenuation factor of each soil layer.

[0020] Preferably, in the STEP4, the mathematical model is a partial differential equation system coupling soil elastoplastic deformation and seepage field, and its boundary conditions are dynamically updated according to the pile penetration rate and vibration signal spectrum characteristics.

[0021] Preferably, in the STEP5, the prediction model is a long short-term memory neural network. Its input features include historical displacement gradient, pore water pressure change rate, and vibration energy distribution, and the output is the extreme value of soil displacement and the peak value of pore water pressure in the next three construction stages.

[0022] Preferably, in the STEP6, the optimization algorithm adopts an adaptive particle swarm optimization algorithm. The optimization variables are the cohesion correction factor and permeability coefficient attenuation rate of the layered soil, and the constraint conditions are the engineering experience thresholds of soil layer physical parameters.

[0023] Preferably, in the STEP7, the convergence condition is that the root mean square value decline rate of the prediction error for three consecutive iterations is less than the set threshold, and the maximum absolute error does not exceed the allowable engineering deviation.

[0024] Preferably, in the STEP8, the sampling frequency adjustment strategy of the sensor array is as follows: According to the displacement change rate and pore water pressure gradient, the sampling interval is dynamically adjusted according to an exponential function; the spatial distribution optimization adopts a monitoring blind area filling point algorithm based on Kriging interpolation.

[0025] Preferably, in the STEP9, in the closed-loop iterative analysis process, the weight parameters of the prediction model are updated through an online learning mechanism. The online learning mechanism adopts an incremental stochastic gradient descent algorithm, and the learning rate has a negative correlation with the construction stage progress.

[0026] Compared with the prior art, the beneficial effects of the present invention are:

[0027] A multi-source sensor array is adopted, including distributed fiber optic sensors, piezoelectric pore water pressure gauges, strain displacement gauges and acceleration sensors. They are arranged in layers along the pile axis and the radial arrangement density changes with the gradient of the soil layer permeability coefficient. This layout method can collect soil displacement, pore water pressure, pile penetration resistance and vibration signals in real time from multiple dimensions, comprehensively reflecting the impact of jacked pile construction on the soil. Compared with traditional single-sensor monitoring, the obtained data is richer and more accurate, providing a solid data foundation for subsequent precise analysis.

[0028] In the data preprocessing stage, the wavelet threshold denoising algorithm is used to remove noise, the dynamic time warping algorithm is used to align time series, and the piecewise normalization method based on soil layer depth is used to normalize data. These advanced data processing techniques effectively improve the data quality, eliminate noise interference and problems of inconsistent time series, make the data of different sensors comparable, and the generated standardized monitoring dataset can more truly reflect the actual state of the soil, improving the reliability of the analysis results.

[0029] The established layered soil finite element model includes dynamic parameters such as elastic modulus, Poisson's ratio and permeability coefficient of soil layers at different depths in soft soil areas, and is corrected in real time through a backpropagation neural network. At the same time, a partial differential equation system coupling soil elastoplastic deformation and seepage field is constructed as a mathematical model of the relationship between soil compaction effect and time, and its boundary conditions are dynamically updated according to the pile penetration rate and the spectral characteristics of vibration signals. This model construction and update method fully considers the complexity of soil layers in soft soil areas and the dynamic changes in the construction process, and can more accurately describe the variation law of soil compaction effect with time, providing strong model support for accurate analysis and prediction.

[0030] The long short-term memory neural network (LSTM) is used as a prediction model. By inputting features such as historical displacement gradient, pore water pressure change rate and vibration energy distribution, it can effectively learn the time series characteristics of soil state and accurately predict the extreme values of soil displacement and the peak values of pore water pressure in the next three construction stages. Compared with traditional prediction methods, the LSTM model has stronger processing ability for complex non-linear relationships, and the prediction results are more reliable, which helps construction personnel to master the trend of soil response in advance and take corresponding measures in time to ensure construction safety and project quality.

[0031] The adaptive particle swarm optimization algorithm is used to adjust the dynamic parameters of the layered soil model to minimize the error between the prediction results and the actual monitoring data. The optimization variables are the cohesion correction factor and the permeability coefficient attenuation rate of the layered soil, and the constraint conditions are the engineering experience thresholds of soil physical parameters. This optimization algorithm can quickly search for the optimal solution within a reasonable parameter range, making the model continuously approach the actual situation, improving the prediction accuracy and adaptability of the model, and ensuring the accuracy of the analysis results.

[0032] According to the model prediction error and analysis results, dynamically adjust the sampling frequency and spatial distribution of the sensor array. By dynamically adjusting the sampling interval according to the displacement change rate and pore water pressure gradient according to the exponential function, and optimizing the spatial distribution by using the blind area filling point algorithm based on Kriging interpolation, the sensor layout can be made more reasonable, and the monitoring efficiency and data representativeness can be improved. Increase the sampling frequency and the number of sensors in the area where the soil changes violently, and reasonably reduce them in the area where the change is gentle, so as to ensure the integrity and accuracy of the monitoring data while avoiding resource waste.

[0033] Feed the updated monitoring data back into the prediction model to form a closed-loop iterative analysis process. The weight parameters of the prediction model are updated through an online learning mechanism (incremental stochastic gradient descent algorithm), and the learning rate is negatively correlated with the progress of the construction stage. As the construction progresses, the model continuously absorbs new monitoring data, continuously optimizes its own performance, realizes the dynamic update of the prediction of soil body response, makes the monitoring and analysis results always closely combined with the actual construction situation, provides timely and accurate basis for construction decision-making, and effectively guarantees the smooth progress of the construction process and the overall quality of the project. Brief Description of the Drawings

[0034] Figure 1 It is the working principle diagram of the monitoring and analysis method for the time effect of the soil squeezing characteristics of static pressure piles in soft soil areas described in the present invention;

[0035] Figure 2 It is the step diagram of the layout of the multi-source sensor array;

[0036] Figure 3 It is the step diagram of the correction of the dynamic parameters of the stratified soil finite element model;

[0037] Figure 4 It is the flow chart of the soil body response prediction based on LSTM;

[0038] Figure 5 It is the dynamic adjustment diagram of the sensor array. Detailed Embodiment

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0040] Please refer to Figures 1 - 5 , the present invention provides a technical solution: a monitoring and analysis method for the time effect of the soil squeezing characteristics of static pressure piles in soft soil areas, the method includes:

[0041] STEP1: Deploy a multi-source sensor array in the construction area of static pressure piles to collect soil displacement, pore water pressure, pile penetration resistance, and vibration signals during construction in real time. By reasonably arranging the sensors, comprehensive and accurate on-site data can be obtained, providing a basis for subsequent analysis.

[0042] STEP2: Preprocess the collected raw data, including noise filtering, time series alignment, and data normalization, to generate a standardized monitoring data set. The preprocessing process removes noise interference in the data, unifies the time series, and standardizes the data range, making the data more suitable for subsequent modeling and analysis.

[0043] STEP3: Based on the standardized monitoring data set, establish a layered soil finite element model, which includes the dynamic parameters of elastic modulus, Poisson's ratio, and permeability coefficient of soil layers at different depths in soft soil areas. This model can simulate the mechanical response of soil during the construction of static pressure piles.

[0044] STEP4: Combine time series analysis methods to extract the time-varying characteristics of soil displacement and pore water pressure, and construct a mathematical model that correlates the soil compaction effect with time. By analyzing the time-varying characteristics, reveal the variation law of the soil compaction effect over time.

[0045] STEP5: Adopt a prediction model based on machine learning, input the time-varying characteristics of STEP4 and the parameters of the layered soil model in STEP3, and predict the soil response in the subsequent construction stage. Utilize the powerful prediction ability of machine learning to estimate the impact of construction on soil in advance.

[0046] STEP6: Adjust the dynamic parameters of the layered soil model through an optimization algorithm to minimize the error between the prediction results and the actual monitoring data. Continuously optimize the model parameters to improve the accuracy of model prediction.

[0047] STEP7: Determine whether the prediction error of the model after parameter optimization meets the convergence condition; if it meets, output the optimized soil parameters and the time evolution law of the soil compaction effect; if it does not meet, return to STEP6 for re-optimization. Ensure that the model reaches a satisfactory prediction accuracy.

[0048] STEP8: According to the output results of STEP7, dynamically adjust the sampling frequency and spatial distribution of the sensor array. Make the monitoring more targeted and improve the monitoring efficiency and accuracy.

[0049] STEP9: Feed the updated monitoring data back into the prediction model in STEP5 to form a closed-loop iterative analysis process. Continuously update the data and continuously improve the performance of the prediction model.

[0050] The present invention will be further described below in conjunction with Examples 1 to 5:

[0051] Example 1:

[0052] The multi-source sensor array includes a distributed fiber optic sensor, a piezoelectric pore water pressure gauge, a strain displacement gauge, and an acceleration sensor. The distributed fiber optic sensor has the advantages of high precision, long distance, and distributed measurement, and can continuously monitor the strain change of the soil along the pile axis direction, thereby obtaining the soil displacement information. The piezoelectric pore water pressure gauge uses the piezoelectric effect to quickly and accurately measure the change of pore water pressure in the soil. It has high measurement accuracy and fast response speed, and can effectively capture the dynamic change of pore water pressure during construction. The strain displacement gauge calculates the soil displacement by measuring its own strain, and has the characteristics of simple structure and relatively high measurement accuracy, and can work stably in a complex construction environment. The acceleration sensor is used to monitor the vibration condition during the pile construction process, providing data support for analyzing the influence of vibration on the soil.

[0053] The sensors are arranged in layers along the pile axis direction. This is because there are differences in the properties of soil layers at different depths in soft soil areas, and arranging in layers can obtain the response information of the soil at different depths. For example, in the shallow soft soil, the soil is greatly disturbed by the pile construction, and the displacement and pore water pressure change significantly; while in the deep soil, although the disturbance is relatively small, there will still be a certain response. Arranging in layers can comprehensively capture these changes. Moreover, the radial arrangement density of the sensors changes with the gradient of the soil permeability coefficient. In the soil layer with a large permeability coefficient, the pore water in the soil flows relatively fast, and the propagation and dissipation of the soil compaction effect are also relatively fast. Therefore, increasing the radial arrangement density of the sensors in these soil layers can more accurately monitor the changes in pore water pressure and soil displacement; on the contrary, in the soil layer with a small permeability coefficient, the arrangement density of the sensors is relatively small, which can not only meet the monitoring requirements but also reasonably control the cost.

[0054] In actual construction, assume that static pressure pile construction is carried out in a certain soft soil area, and the soil layers on the site from top to bottom are silty clay, silty clay, and sandy silt. According to the soil permeability coefficient, in the silty clay, the radial arrangement spacing of the sensors is set to 0.5 meters; in the silty clay, the spacing is set to 1 meter; in the sandy silt, the spacing is set to 1.5 meters. Along the pile axis direction, a group of sensors is arranged every 1 meter, and each group of sensors includes a distributed fiber optic sensor, a piezoelectric pore water pressure gauge, a strain displacement gauge, and an acceleration sensor, ensuring that the soil information at different depths and radial positions can be comprehensively and accurately collected.

[0055] Example 2:

[0056] Noise filtering adopts the wavelet threshold denoising algorithm. During the construction of static pressure piles, the data collected by sensors will inevitably be interfered by various noises, such as environmental noise, noise generated by construction machinery vibration, etc. The wavelet threshold denoising algorithm is based on the multi-resolution analysis characteristics of wavelet transform, decomposing the signal into different frequency sub-bands. In each sub-band, a threshold is set according to the statistical characteristics of the noise. The wavelet coefficients smaller than the threshold are regarded as noise and processed, while the wavelet coefficients larger than the threshold are retained. Then, the signal is reconstructed through inverse wavelet transform to achieve the purpose of denoising. This algorithm can effectively remove noise while retaining the main features of the signal. For example, when processing soil displacement signals, it can not only remove the fluctuations caused by noise but also accurately retain the trend of displacement changes.

[0057] Time series alignment adopts the dynamic time warping algorithm. Since multiple sensors start collecting data at different positions and times, the collected time series may be inconsistent. The dynamic time warping algorithm aligns the asynchronous time series by finding the optimal matching path between two time series. Specifically, it calculates the cumulative distance of all possible matching paths between two time series and selects the path with the minimum cumulative distance as the optimal matching path, thus achieving the alignment of time series. Taking the soil displacement and pore water pressure data as an example, after being processed by the dynamic time warping algorithm, the data collected by different sensors can be made corresponding in time, facilitating the subsequent simultaneous analysis of the relationship between the two.

[0058] Data normalization adopts the piecewise standardization method based on soil layer depth. The physical properties of soil layers at different depths are different, and there are also differences in the data range and magnitude collected by sensors. The piecewise standardization method based on soil layer depth divides the soil layer according to depth and standardizes the data within each soil layer segment. Suppose the soil layer is divided into three segments: shallow layer (0 - 5 meters), middle layer (5 - 10 meters), and deep layer (below 10 meters). For the data of the shallow soil layer, the formula is used:

[0059]

[0060] where, is the original data, and are the minimum and maximum values of this type of data within the shallow soil layer respectively, is the normalized data. Through this piecewise standardization method, the influence of soil layer depth difference on data can be eliminated, making the data of soil layers at different depths comparable and providing unified standard data for subsequent modeling and analysis.

[0061] In practical applications, data collected over a week at a construction site is processed. After being processed by the wavelet threshold denoising algorithm, the data curve becomes significantly smoother, and the burr phenomenon caused by noise is greatly reduced; the dynamic time warping algorithm precisely aligns the time series of different sensors, and when analyzing the relationship between soil displacement and pore water pressure over time, the data corresponds accurately; after being processed by the piecewise normalization method based on soil layer depth, the data of soil layers at different depths fluctuates within the range of 0 - 1, providing a good data basis for establishing a layered soil finite element model in the follow-up.

[0062] Example 3:

[0063] The dynamic parameters of the layered soil finite element model are corrected in real time through a backpropagation neural network. The backpropagation neural network is a neural network model with powerful learning ability. It adjusts the weights and thresholds of the network through the backpropagation of errors, so that the output of the network is as close as possible to the true value. In the present invention, the normalized monitoring data of STEP2 is used as the input of the backpropagation neural network, and the output is the elastic modulus correction coefficient and permeability coefficient attenuation factor of each soil layer.

[0064] The specific process is as follows: information such as soil displacement and pore water pressure in the normalized monitoring data is input into the backpropagation neural network, and multiple forward propagation and backpropagation calculations are performed inside the network. During the forward propagation process, the data is processed through weighted summation and activation functions by neurons in the network, and the output of the network is gradually obtained; during the backpropagation process, according to the error between the network output and the actual monitoring data, the gradients of the error with respect to the network weights and thresholds are calculated, and the weights and thresholds are adjusted using the gradient descent method. After multiple iterative trainings, the elastic modulus correction coefficient and permeability coefficient attenuation factor output by the network can more accurately reflect the actual situation of the soil layer.

[0065] For example, in a certain construction stage, it is found through monitoring data that there is a deviation between the predicted soil displacement of the model and the actual monitored value. The normalized monitoring data of this stage is input into the backpropagation neural network, and after training, the elastic modulus correction coefficient of a certain soil layer is 1.2, and the permeability coefficient attenuation factor is 0.8. This means that the actual elastic modulus of this soil layer is larger than the original model setting value, and the permeability coefficient needs to be appropriately reduced. By real-time correcting the dynamic parameters of the layered soil finite element model, the model can be made to fit the mechanical behavior of the soil during the actual construction process more closely, improve the simulation accuracy of the soil compaction effect of the model, and further provide a more reliable basis for subsequent prediction and analysis. In the subsequent construction stage, predictions are made based on the corrected model, and the error between the prediction results and the actual monitoring data is significantly reduced, proving the effectiveness of this dynamic parameter correction method.

[0066] Example 4:

[0067] In STEP4, the constructed mathematical model is a partial differential equation system that couples the elastoplastic deformation of the soil and the seepage field. For the elastoplastic deformation of the soil, the Mohr-Coulomb yield criterion is adopted, and the yield function is:

[0068]

[0069] where and are the maximum and minimum principal stresses respectively, is the cohesion of the soil, is the internal friction angle.

[0070] The equilibrium equation of the soil is:

[0071]

[0072] where is the stress tensor, is the body force component; in this formula represents the spatial coordinate variable, which is used to describe the variation direction of the soil stress tensor . In three-dimensional space, j usually takes 1, 2, 3, corresponding to the x, y, and z directions respectively. By taking the partial derivative of with respect to , the variation of stress in each spatial direction is reflected, and further the equilibrium relationship between the internal stress of the soil and the body force component is reflected.

[0073] For the seepage field, according to Darcy's law, the seepage velocity is:

[0074]

[0075] where is the permeability coefficient tensor, is the hydraulic head. In this formula is also the spatial coordinate variable, which is used to determine the variation direction and position of the hydraulic head . j corresponds to different spatial directions. By taking the partial derivative of the hydraulic head h with respect to , the variation rate of the hydraulic head in this direction is obtained, and then combined with the permeability coefficient tensor to calculate the component of the seepage velocity in the i direction.

[0076] The continuity equation is:

[0077]

[0078] where is the porosity, is related to the seepage velocity The spatial coordinate variable corresponding to the direction, and the value of i determines which component of the seepage velocity in the spatial direction. is the time variable.

[0079] Couple the elastoplastic deformation of the soil mass and the seepage field equation to form a partial differential equation system describing the soil compaction effect.

[0080] The boundary conditions are dynamically updated according to the pile penetration rate and the spectral characteristics of the vibration signal. Assume the pile penetration rate is , on the pile-soil contact boundary, the displacement boundary condition of the soil can be expressed as:

[0081]

[0082] Among them, is the soil displacement component, is the boundary normal vector, is the time variable.

[0083] For the vibration signal, the main frequency components are obtained through spectral analysis , and the pore water pressure boundary condition considering the vibration effect on the boundary can be expressed as:

[0084]

[0085] Among them, is the pore water pressure, is the initial pore water pressure, is the amplitude, is the phase. By dynamically updating the boundary conditions in this way, the mathematical model can more accurately reflect the change of the soil compaction effect with time during the construction of the static pressure pile.

[0086] During the construction of the static pressure pile, the soil mass undergoes both elastoplastic deformation and is accompanied by the seepage of pore water, and the two affect each other. This partial differential equation system couples the elastoplastic mechanics equation of the soil mass and the seepage equation to comprehensively describe the mechanical response of the soil mass during the construction process. Its boundary conditions are dynamically updated according to the pile penetration rate and the spectral characteristics of the vibration signal. The pile penetration rate directly affects the deformation rate of the soil mass and the change of the pore water pressure, and the spectral characteristics of the vibration signal reflect the vibration energy distribution during the construction process. These factors will all affect the mechanical behavior of the soil mass. For example, when the pile penetration rate increases, the extrusion effect on the soil mass is enhanced, and the pore water pressure rises rapidly. At this time, the boundary conditions of the mathematical model will be adjusted accordingly to more accurately simulate the response of the soil mass.

[0087] In terms of the prediction model, the long short-term memory neural network (LSTM) is adopted. LSTM has the ability to process long sequence data and capture the long-term dependence relationships of data, and is very suitable for predicting the changes of soil displacement and pore water pressure over time. Its input features include historical displacement gradient, pore water pressure change rate, and vibration energy distribution. The historical displacement gradient reflects the change trend of soil displacement over time, the pore water pressure change rate reflects the dynamic change of pore water pressure, and the vibration energy distribution describes the influence degree of construction vibration on the soil. By inputting these features into the LSTM, the model can learn the relationships between them and the future soil response, and the output is the extreme values of soil displacement and the peak values of pore water pressure in the next three construction stages.

[0088] Inside the LSTM cell, there are input gates , forget gates , output gates and memory cells . The input features are selected as the historical displacement gradient, pore water pressure change rate, and vibration energy distribution, and these features can fully reflect the change trend of soil state over time and the vibration influence during the construction process. Let the historical displacement gradient be , the pore water pressure change rate be , the vibration energy distribution be , and the current input vector is composed of these three features, that is , and the hidden state at the previous moment is .

[0089] The calculation formula of the input gate is:

[0090]

[0091] Among them, and are the weight matrices from the input and the hidden state to the input gate respectively, is the bias of the input gate, is the Sigmoid activation function, which maps the input value to the interval and is used to control the degree of new information entering the memory cell.

[0092] The calculation formula of the forget gate is:

[0093]

[0094] and are the corresponding weight matrices, is the bias. The forget gate determines which information in the memory cell needs to be retained and which needs to be discarded, and outputs values between through the Sigmoid function. Values close to 1 indicate retaining the corresponding information, and values close to 0 indicate discarding it.

[0095] Calculate the candidate memory cell :

[0096]

[0097] Here, and are weight matrices, is the bias, The function maps the input to the interval (-1, 1) to generate new information that may be updated to the memory cell.

[0098] Update the memory cell :

[0099]

[0100] That is, according to the control of the forget gate and the input gate, the information in the memory cell at the previous moment is fused with the new information in the candidate memory cell .

[0101] Output gate The calculation formula is:

[0102]

[0103] and are weight matrices, is the bias. The output gate determines which information in the memory cell will be used to generate the output at the current moment.

[0104] Finally, calculate the hidden state at the current moment:

[0105]

[0106] The output of the LSTM is the extreme values of soil displacement and the peak pore water pressure in the next three construction stages. These predicted values provide an important basis for the construction party to understand the soil response in advance.

[0107] In practical applications, take a large construction project in a soft soil area as an example. During the construction process, historical displacement gradients, pore water pressure change rates, and vibration energy distribution data are collected in real time and input into the trained LSTM prediction model. The model predicts that the extreme values of soil displacement in the next three construction stages are 5 cm, 6 cm, and 7 cm respectively, and the peak values of pore water pressure are 50 kPa, 55 kPa, and 60 kPa respectively. According to these prediction results, construction workers take corresponding measures in advance, such as adjusting the construction sequence and speed of subsequent piles, to avoid the impact of excessive soil displacement and pore water pressure on surrounding buildings and underground pipelines. At the same time, by comparing the actual monitoring data with the prediction results, it is found that the prediction results are relatively close to the actual situation, verifying the effectiveness and practicality of the mathematical model and prediction model.

[0108] Example 5:

[0109] In terms of the optimization algorithm, the adaptive particle swarm optimization algorithm is adopted. This algorithm simulates the foraging behavior of bird flocks and searches for the optimal solution by continuously searching and updating particles in the solution space. In the present invention, the optimization variables are the cohesion correction factor of the layered soil and the attenuation rate of the permeability coefficient. During the optimization process, each particle represents a set of values of the cohesion correction factor and the permeability coefficient attenuation rate. The particle continuously adjusts its position according to its own historical optimal position and the global optimal position of the group to find the parameter values that minimize the error between the prediction result and the actual monitoring data. The constraint condition is the engineering experience threshold of the soil layer physical parameters to ensure that the optimized parameters are within a reasonable engineering range. For example, in a certain optimization iteration, after calculation, the cohesion correction factor corresponding to a certain particle is 1.1, and the permeability coefficient attenuation rate is 0.05, which meets the requirements of the engineering experience threshold, and the subsequent iteration continues until the optimal solution is found.

[0110] The convergence condition is that the root mean square value decline rate of the prediction error for three consecutive iterations is less than the set threshold, and the maximum absolute error does not exceed the allowable engineering deviation. The root mean square value of the prediction error can comprehensively reflect the error degree between the prediction result and the actual data. The decline rate less than the set threshold indicates that the optimization effect of the model is gradually stabilizing; the maximum absolute error not exceeding the allowable engineering deviation ensures the usability of the model prediction result in engineering. Suppose the set threshold for the root mean square value decline rate of the prediction error is 0.05, and the allowable engineering deviation is 10%. During the model optimization process, after multiple iterative calculations, when the root mean square value decline rates of the prediction error for three consecutive iterations are 0.04, 0.03, 0.02 respectively, and the maximum absolute error is 8%, the convergence condition is met, the optimization is stopped, and the optimized soil parameters and the time evolution law of the soil compaction effect are output.

[0111] The sampling frequency adjustment strategy for the sensor array is as follows: According to the displacement change rate and pore water pressure gradient, the sampling interval is dynamically adjusted according to an exponential function. When the displacement change rate is large or the pore water pressure gradient is large, it indicates that the soil state changes rapidly, and it is necessary to increase the sampling frequency and shorten the sampling interval; on the contrary, the sampling frequency is reduced and the sampling interval is increased. For example, when the displacement change rate reaches 5 mm / h and the pore water pressure gradient is 10 kPa / m, the sampling interval is adjusted to once every 5 minutes according to the exponential function; when the displacement change rate decreases to 1 mm / h and the pore water pressure gradient decreases to 5 kPa / m, the sampling interval is adjusted to once every 15 minutes. This can reasonably reduce the data acquisition volume and processing pressure while ensuring the accuracy of the monitoring data.

[0112] The spatial distribution optimization adopts a monitoring blind area filling point algorithm based on Kriging interpolation. During the construction process, due to the limitations of sensor layout, there may be monitoring blind areas. The Kriging interpolation algorithm uses the data of known monitoring points, estimates the values of the monitoring blind areas through the spatial autocorrelation function, and reasonably fills points in the monitoring blind areas according to the estimation results. For example, in a certain construction area, through the analysis of the Kriging interpolation algorithm, it is found that there is a monitoring blind area in a certain area. The soil displacement and pore water pressure in this area are estimated based on the data of surrounding monitoring points, and sensors are added in this area to improve the comprehensiveness and accuracy of the monitoring.

[0113] In the closed-loop iterative analysis process, the weight parameters of the prediction model are updated through an online learning mechanism, and the online learning mechanism adopts an incremental stochastic gradient descent algorithm. As the construction progresses, new monitoring data is continuously generated. The incremental stochastic gradient descent algorithm uses these new data and updates the model weights with only one sample each time, which can quickly adapt to the changes in the data. The learning rate has a negative correlation with the progress of the construction stage. In the initial stage of construction, the learning rate is large, and the model can quickly learn the characteristics of the new data; as the construction progresses, the learning rate gradually decreases, and the model becomes more stable, avoiding overfitting. By continuously updating the weight parameters of the prediction model, the model can better adapt to the changes during the construction process and improve the accuracy of the prediction.

[0114] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "including", "comprising" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0115] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring and analyzing the time effect of soil squeezing characteristics of static piles in soft soil areas, characterized in that: The method comprises: STEP 1: Deploy a multi-source sensor array in the static pile construction area to collect soil displacement, pore water pressure, pile penetration resistance and vibration signals in real time during the construction process; STEP 2: Preprocess the collected raw data, including noise filtering, time series alignment and data normalization, to generate a standardized monitoring data set; STEP3: Based on the standardized monitoring data set, a layered soil finite element model is established, wherein the model includes dynamic parameters of elastic modulus, Poisson's ratio and permeability coefficient of soil layers at different depths in the soft soil area; STEP4: Combine the time series analysis method to extract the time-varying characteristics of soil displacement and pore water pressure, and construct a mathematical model of the correlation between soil squeezing effect and time; STEP5: Use a prediction model based on machine learning, input the time-varying characteristics of STEP4 and the layered soil model parameters of STEP3 to predict the soil response in the subsequent construction stage; STEP 6: Adjust the dynamic parameters of the layered soil model through the optimization algorithm to minimize the error between the prediction results and the actual monitoring data; STEP7: Determine whether the model prediction error after parameter optimization meets the convergence condition; if so, output the optimized soil parameters and the time evolution law of soil squeezing effect; if not, return to STEP6 for re-optimization; STEP8: According to the output results of STEP7, dynamically adjust the sampling frequency and spatial distribution of the sensor array; STEP9: Feed the updated monitoring data back to the prediction model in STEP5 to form a closed-loop iterative analysis process.

2. The method according to claim 1, characterized in that: In STEP1, the multi-source sensor array includes distributed optical fiber sensors, piezoelectric pore water pressure gauges, strain gauges and acceleration sensors; the sensors are arranged in layers along the axis of the pile body, and the radial arrangement density varies with the gradient of the soil permeability coefficient.

3. The method according to claim 1, characterized in that In the STEP 2, the noise filtering adopts a wavelet threshold denoising algorithm, the time series alignment adopts a dynamic time warping algorithm, and the data normalization adopts a segmented standardization method based on soil layer depth.

4. The method according to claim 2, characterized in that: In the STEP3, the dynamic parameters of the layered soil finite element model are corrected in real time through a back propagation neural network, the input is the standardized monitoring data of STEP2, and the output is the elastic modulus correction coefficient and permeability attenuation factor of each soil layer.

5. The method according to claim 1, characterized in that In the STEP 4, the mathematical model is a set of partial differential equations that couple the elastic-plastic deformation of the soil and the seepage field, and its boundary conditions are dynamically updated according to the pile penetration rate and the frequency spectrum characteristics of the vibration signal.

6. The method according to claim 1, characterized in that In STEP 5, the prediction model is a long short-term memory neural network, whose input features include historical displacement gradients, pore water pressure change rates and vibration energy distribution, and the output is the soil displacement extremes and pore water pressure peaks in the next three construction stages.

7. The method according to claim 1, characterized in that: In the STEP 6, the optimization algorithm adopts an adaptive particle swarm optimization algorithm, the optimization variables are the cohesion correction factor and the permeability attenuation rate of the layered soil, and the constraint conditions are the engineering experience thresholds of the physical parameters of the soil layer.

8. The method according to claim 1, characterized in that: In STEP 7, the convergence condition is that the rate of decrease of the root mean square value of the prediction error for three consecutive iterations is less than a set threshold, and the maximum absolute error does not exceed the allowable engineering deviation.

9. The method according to claim 1, characterized in that: In STEP 8, the sampling frequency adjustment strategy of the sensor array is: dynamically adjusting the sampling interval according to the displacement change rate and the pore water pressure gradient according to the exponential function; the spatial distribution optimization adopts the monitoring blind area filling algorithm based on Kriging interpolation.

10. The method according to claim 1, characterized in that In STEP 9, in the closed-loop iterative analysis process, the weight parameters of the prediction model are updated through an online learning mechanism. The online learning mechanism adopts an incremental stochastic gradient descent algorithm, and the learning rate is negatively correlated with the progress of the construction phase.

Images