Foundation pit deformation control method and system based on dynamic load compensation

Abstract

The present invention relates to a foundation pit deformation control method and system based on dynamic load compensation, and belongs to the field of foundation pit support and computer technology. The foundation pit deformation control method based on dynamic load compensation includes: S1, laying a hydraulic support unit on the inner side of the foundation pit support pile or underground continuous wall, monitoring the pressure of the hydraulic support unit and the displacement of the support pile or underground continuous wall in real time, and calculating the support force that the hydraulic support unit needs to provide; S2, monitoring the foundation pit soil strain data in real time, obtaining vehicle position, speed and load data, and constructing a dynamic load distribution model; S3, using an LSTM neural network to construct a foundation pit deformation prediction model, predict the trend of the impact of vehicle traffic on foundation pit deformation, so as to adjust the support force of the support structure in advance, and automatically optimize the vehicle path. This method actively controls deformation through a dynamic compensation mechanism to ensure construction safety and the stability of surrounding facilities. It is particularly suitable for sensitive areas near underground facilities or important buildings.

Description

Technical Field

[0001] The present invention belongs to the field of foundation pit support and computer technology, and in particular relates to a foundation pit deformation control method and system based on dynamic load compensation. Background Art

[0002] When constructing deep foundation pits in densely populated urban areas, limited space often forces heavy vehicles (such as dump trucks) to pass close to the edges of the pit. The dynamic loads from these vehicles can significantly increase the lateral pressure on the support structure, leading to excessive pit deformation. Controlling pit deformation is particularly critical in sensitive environments near underground facilities (such as subway tunnels) or important buildings. Existing methods for controlling pit deformation primarily rely on static support design and passive monitoring. However, these methods are unable to cope with real-time changes in dynamic loads and are often slow to respond, leading to the risk of excessive pit deformation.

[0003] Traditional static support system designs are unable to adapt to changes in load. Rigid, fixed support structures struggle to effectively handle sudden load fluctuations caused by vehicle traffic, especially when these changes occur rapidly. Furthermore, existing vehicle path planning techniques fail to effectively link with pit deformation data, making it impossible to predict and mitigate pit deformation through real-time monitoring and dynamic adjustments, thus increasing construction risks.

[0004] Therefore, how to solve the problem of insufficient control over deformation caused by dynamic loads during urban foundation pit construction, especially the impact of dynamic loads generated by vehicles passing around the foundation pit on the support structure, is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the present invention provides a foundation pit deformation control method and system based on dynamic load compensation, which actively controls deformation through a dynamic compensation mechanism to ensure construction safety and stability of surrounding facilities.

[0006] The present invention provides a foundation pit deformation control method based on dynamic load compensation, comprising the following steps:

[0007] S1: Install hydraulic support units inside foundation pit support piles or underground continuous walls, and monitor the pressure of the hydraulic support units in real time. , and the displacement of the supporting piles or underground continuous wall , calculate the support force that the hydraulic support unit needs to provide ;

[0008] S2 monitors the soil strain data around the foundation pit in real time, obtains the position, speed and load data of passing vehicles, and integrates these data to construct a dynamic load distribution model;

[0009] S3 uses the LSTM neural network to build a foundation pit deformation prediction model to predict the impact trend of vehicle traffic on foundation pit deformation. Based on the prediction results, the support force of the support structure is adjusted in advance, and the vehicle path is automatically optimized to avoid insufficient support force of the support structure when vehicles pass.

[0010] This technical solution actively controls deformation through a dynamic compensation mechanism to ensure construction safety and the stability of surrounding facilities. It is particularly suitable for sensitive areas near underground facilities or important buildings.

[0011] In some embodiments, in step S1, the support force required to be provided by the hydraulic support unit is calculated. The method includes: taking the surface point at the center of the foundation pit excavation area as the origin and the direction of the road along which the vehicle travels as the , perpendicular to the road direction , vertically downward Axis establishes a spatial rectangular coordinate system Set dual threshold indicators for foundation pit structure safety control, including pressure threshold and displacement threshold , when the pressure of the hydraulic support unit is monitored , or displacement of supporting piles or underground continuous wall When the pressure of the hydraulic support unit is , and the displacement of the supporting piles or underground continuous wall , through the dual-channel PID controller to integrate pressure control and displacement control, and use error feedback to calculate the support force , the calculation formula is shown in formula (1):

[0012] (1);

[0013] In formula (1), is the pressure error, ; is the displacement error, ; To adjust the weight coefficient; 、 and is the PID control parameter of the pressure channel; 、 and is the displacement channel PID control parameter; For time.

[0014] In some embodiments, in step S2, the method for constructing a dynamic load distribution model includes: using a data fusion algorithm to integrate the foundation pit deformation with the vehicle load data and motion trajectory data to construct a dynamic load distribution model, as shown in formula (2):

[0015] (2);

[0016] In formula (2), q(x, y, t) is the load distribution caused by the vehicle at position (x, y) on the foundation pit plane at time t; For the The time-varying load of the vehicle, a car; is the real-time horizontal coordinate of the vehicle, The real-time vertical coordinate of the vehicle; is the vehicle load influence radius, which is negatively correlated with vehicle speed.

[0017] In some embodiments, step S2 further includes: using optical fiber sensors around the foundation pit to sense soil strain, obtaining the relationship between soil strain and vehicle load, and converting the dynamic load distribution of the vehicle into a relationship expression between soil strain and vehicle load, thereby obtaining a strain prediction model for the foundation pit caused by the vehicle load; wherein the relationship expression between soil strain and vehicle load is shown in formula (3):

[0018] (3);

[0019] In formula (3), is the soil strain, is the vertical depth, is the elastic modulus of soil, is Poisson's ratio, is the additional stress caused by vehicle load, is the horizontal stress of soil, is the vertical stress of soil.

[0020] In some embodiments, the method for obtaining the strain prediction model of the foundation pit caused by the vehicle load includes: The foundation mechanics model is converted into , and input it into formula (3) to obtain the strain prediction model of the foundation pit caused by vehicle load, as shown in formula (4):

[0021] (4);

[0022] In formula (4), For the Vehicle load impact area, Indicates additional stress to be calculated The projection coordinates of the spatial point on the surface; is the integral variable, which represents the point of action of the vehicle load on the ground.

[0023] In some embodiments, in step S3, the method for constructing the foundation pit deformation prediction model includes: , real-time speed , the support force of the hydraulic support unit Soil strain , displacement of supporting piles or underground continuous walls , through the fusion of multi-source data, synchronous processing, normalization and time series alignment, a unified time series input vector is formed , and input it into the LSTM neural network to output the foundation pit deformation prediction model; among them, the time series input vector As shown in formula (5):

[0024]

[0025] The foundation pit deformation prediction model is shown in formula (6):

[0026] (6);

[0027] In formula (6), is the input feature at the current moment, is the size of the time window, The displacement of the supporting piles or underground continuous wall at a predicted moment in the future.

[0028] In some embodiments, step S3 further includes: performing error back propagation according to actual monitoring data to update and optimize the foundation pit deformation prediction model. The optimized foundation pit deformation prediction model expression is shown in formula (7):

[0029] (7);

[0030] In formula (7), Transformed for the future, equivalent to , is the neural network weight parameter.

[0031] In some embodiments, the LSTM neural network includes a forget gate, an input gate, and an output gate, wherein:

[0032] Forget Gate As shown in formula (8):

[0033] (8);

[0034] In formula (8), Is the output of the forget gate, with a value range between 0 and 1, where 0 means completely forgotten and 1 means completely retained; and They are the weight matrix and bias term of the forget gate respectively; is the current input, is the hidden state at the previous moment; For vector splicing, vertical vector splicing is performed on the hidden state of the previous moment and the current input; It is the Sigmoid activation function, which ensures that the output is in the range of [0,1];

[0035] Input Gate And the candidate state is shown in formula (9):

[0036] (9);

[0037] In formula (9), The gate vector output by the input gate is constrained to be in the range of [0, 1] through the Sigmoid activation function; is the weight matrix of the input gate; is the bias term of the input gate, which is used to adjust the flexibility of the model; is the bias term of the candidate cell state; is the candidate cell state, which is nonlinearly transformed by tanh; is the weight matrix of the candidate cell state; is the current cell state; is the cell state at the previous moment; is the operator symbol, which means element-by-element multiplication;

[0038] Output Gate As shown in formula (10):

[0039] (10);

[0040] In formula (10), Is the output of the output gate, which controls the output amount of the hidden state; is the weight matrix of the output gate; is the bias term of the output gate.

[0041] Based on the above-mentioned foundation pit deformation control method based on dynamic load compensation, the present invention further provides a foundation pit deformation control system based on dynamic load compensation, which adopts the above-mentioned foundation pit deformation control method based on dynamic load compensation, including:

[0042] A hydraulic support unit is arranged inside the foundation pit support pile or underground continuous wall. The hydraulic support unit includes a hydraulic support device, a pressure sensor, and a displacement meter. The hydraulic support device is used to provide support force, the pressure sensor is used to monitor the pressure of the hydraulic support unit in real time, and the displacement meter is used to monitor the displacement of the support pile or underground continuous wall in real time.

[0043] The monitoring unit includes a fiber optic sensor, a GPS locator, and a vehicle scale. The fiber optic sensor is deployed around the foundation pit to sense soil strain. The GPS locator is installed on the vehicle to obtain the vehicle's location. The vehicle scale is installed at the construction road node to weigh the vehicle.

[0044] The central control unit is communicatively connected to the hydraulic support unit and the monitoring unit, and is used to calculate the support force based on the data obtained by the hydraulic support unit and to construct a dynamic load distribution model based on the data obtained by the monitoring unit. The central control unit is also used to construct a foundation pit deformation prediction model, and to adjust the support force of the support structure according to the prediction results, while obtaining an optimized vehicle path.

[0045] This technical solution realizes the full-process automated management from data collection, analysis and prediction to active regulation, significantly improving the real-time, accuracy and reliability of foundation pit deformation control. It is particularly suitable for the safety assurance of deep foundation pit construction in sensitive environments in densely populated urban areas.

[0046] In some embodiments, the foundation pit deformation control system based on dynamic load compensation further includes an early warning unit, which is in communication with the central control unit and is activated when the pressure of the hydraulic support unit is detected. , or displacement of supporting piles or underground continuous wall When the alarm is triggered, the central control unit controls the early warning unit to sound an alarm.

[0047] Based on the above scheme, the foundation pit deformation control method based on dynamic load compensation in the embodiment of the present invention, step S1 monitors the pressure of the hydraulic support unit and the displacement of the support structure in real time, combines the dynamic load data of the vehicle, and dynamically adjusts the support force to effectively respond to the sudden load changes caused by the passage of vehicles and reduce the risk of excessive deformation of the foundation pit; step S2 integrates the soil strain data, vehicle position, speed and load data to construct a dynamic load distribution model, improves the ability to accurately analyze the stress state of the foundation pit, and provides a scientific basis for the adjustment of the support force; step S3 uses the LSTM neural network to predict the influence trend of vehicle passage on the foundation pit deformation, adjusts the support force of the support structure in advance, realizes active protection, avoids the lag problem of traditional passive monitoring, and automatically optimizes the vehicle passage path while predicting the deformation trend of the foundation pit, reduces the adverse effect of the vehicle dynamic load on the support structure, and reduces the construction risk. In summary, the foundation pit deformation control method in this embodiment actively controls deformation through a dynamic compensation mechanism to ensure construction safety and the stability of surrounding facilities. It is particularly suitable for sensitive areas adjacent to underground facilities or important buildings. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0049] Figure 1 Flowchart of a foundation pit deformation control method based on dynamic load compensation in an embodiment of the present invention;

[0050] Figure 2 Schematic diagram of the arrangement of a hydraulic support unit and a monitoring unit in a foundation pit deformation control system based on dynamic load compensation in an embodiment of the present invention.

[0051] In the picture:

[0052] 1. Hydraulic support device; 2. Pressure sensor; 3. Displacement meter; 4. Fiber optic sensor; 5. Vehicle; 6. GPS locator. DETAILED DESCRIPTION

[0053] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0054] In the description of the present invention, it should be understood that the terms "center", "transverse", "longitudinal", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the present invention.

[0055] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood broadly. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to direct connections, indirect connections through an intermediary, or internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0056] like Figure 1 As shown, in one embodiment of the foundation pit deformation control method and system based on dynamic load compensation of the present invention, the foundation pit deformation control method based on dynamic load compensation includes the following steps:

[0057] S1: Install hydraulic support units inside foundation pit support piles or underground continuous walls, and monitor the pressure of the hydraulic support units in real time. , and the displacement of the supporting piles or underground continuous wall , calculate the support force that the hydraulic support unit needs to provide ;

[0058] S2, real-time monitoring of soil strain data around the foundation pit, and obtaining the position, speed, and load data of the passing vehicle 5, integrating the data to construct a dynamic load distribution model;

[0059] S3 uses the LSTM neural network to build a foundation pit deformation prediction model to predict the impact trend of vehicle 5's passage on foundation pit deformation. Based on the prediction results, the support force of the support structure is adjusted in advance, and the path of vehicle 5 is automatically optimized to avoid insufficient support force of the support structure when vehicle 5 passes.

[0060] In the above exemplary embodiment, step S1 of the foundation pit deformation control method based on dynamic load compensation monitors the pressure of the hydraulic support unit and the displacement of the support structure in real time, combines the dynamic load data of vehicle 5, and dynamically adjusts the support force, effectively responds to the sudden load changes caused by the passage of vehicle 5, and reduces the risk of excessive foundation pit deformation; step S2 integrates soil strain data, vehicle 5 position, speed and load data to construct a dynamic load distribution model, improves the ability to accurately analyze the stress state of the foundation pit, and provides a scientific basis for adjusting the support force; step S3 uses the LSTM neural network to predict the impact trend of the passage of vehicle 5 on the foundation pit deformation, adjusts the support force of the support structure in advance, realizes active protection, avoids the lag problem of traditional passive monitoring, and automatically optimizes the passage path of vehicle 5 while predicting the deformation trend of the foundation pit, reduces the adverse impact of the dynamic load of vehicle 5 on the support structure, and reduces construction risks. In summary, the foundation pit deformation control method in this embodiment actively controls deformation through a dynamic compensation mechanism to ensure construction safety and the stability of surrounding facilities. It is particularly suitable for sensitive areas near underground facilities or important buildings.

[0061] In some embodiments, in step S1, the support force required to be provided by the hydraulic support unit is calculated. The method includes: taking the surface point at the center of the foundation pit excavation area as the origin and the direction of the road along which the vehicle 5 travels as the , perpendicular to the road direction , vertically downward Axis establishes a spatial rectangular coordinate system Set dual threshold indicators for foundation pit structure safety control, including pressure threshold and displacement threshold , when the pressure of the hydraulic support unit is monitored , or displacement of supporting piles or underground continuous wall When the pressure of the hydraulic support unit is , and the displacement of the supporting piles or underground continuous wall , through the dual-channel PID controller to integrate pressure control and displacement control, and use error feedback to calculate the support force , the calculation formula is shown in formula (1):

[0062] (1);

[0063] In formula (1), is the pressure error, ; is the displacement error, ; To adjust the weight coefficient, Adaptive adjustments can be made based on historical data, risk level, displacement rate and other factors; 、 and is the PID control parameter of the pressure channel; 、 and is the displacement channel PID control parameter; is time. This embodiment achieves precise monitoring and dynamic control of the foundation pit support structure by establishing a spatial rectangular coordinate system and setting dual threshold indicators of pressure and displacement. When the monitoring value exceeds the safety threshold, the system automatically triggers an alarm and adjusts the support force. The dual-channel PID controller integrates the pressure and displacement error feedback to calculate the optimal support force in real time (Equation 1). Among them, the weight coefficient α can be adaptively adjusted according to historical data, risk level, etc., making the control strategy more flexible, effectively balancing the control requirements of pressure and displacement, improving the support structure's adaptability to dynamic loads, and reducing the risk of excessive foundation pit deformation.

[0064] In some embodiments, in step S2, the method for constructing a dynamic load distribution model includes: using a data fusion algorithm to integrate the pit deformation with the load data and motion trajectory data of the vehicle 5 to construct a dynamic load distribution model, as shown in formula (2):

[0065] (2);

[0066] In formula (2), For vehicle 5 at time Acting on the surface Load distribution (kPa); For the The time-varying load of the vehicle, a car; is the real-time horizontal coordinate of vehicle 5, The real-time vertical coordinate of vehicle 5; is the load influence radius of vehicle 5, which is negatively correlated with vehicle speed. This embodiment integrates foundation pit deformation data with vehicle 5 load and motion trajectory information through a data fusion algorithm to construct a dynamic load distribution model (Equation 2). The model simulates the spatiotemporal influence of vehicle 5 load based on Gaussian distribution, where the load influence radius It is negatively correlated with vehicle speed and can more accurately reflect the dynamic pressure distribution of vehicle 5 on different locations of the foundation pit. The system can quantify the effect of vehicle 5 dynamic load on the support structure, provide high-precision input for subsequent deformation prediction and support force adjustment, and significantly improve the accuracy and real-time performance of dynamic load response.

[0067] In some embodiments, step S2 further includes: using the optical fiber sensor 4 around the foundation pit to sense soil strain, obtaining the relationship between the soil strain and the load of the vehicle 5, and converting the dynamic load distribution of the vehicle 5 into the relationship expression between the soil strain and the load of the vehicle 5, thereby obtaining a strain prediction model caused by the load of the vehicle 5 on the foundation pit; wherein the relationship expression between the soil strain and the load of the vehicle 5 is shown in formula (3):

[0068] (3);

[0069] In formula (3), is the soil strain, is the vertical depth, is the elastic modulus of soil (MPa), is Poisson's ratio, is the additional stress caused by vehicle 5 load (kPa), is the horizontal stress of soil, is the vertical stress of the soil. This embodiment can dynamically evaluate the stress state of the foundation pit by integrating the time-space distribution of the optical fiber monitoring data and the load of vehicle 5, providing a high-precision basis for the adjustment of the support force, and significantly improving the initiative and reliability of deformation control; Equation (3) establishes a direct physical connection between the soil strain and the load it receives, which can be used in two directions: First, by inputting the soil stress field , and combined with known soil mechanical parameters and , the stress field soil strain can be predicted; secondly, the additional stress magnitude can be inverted by the measured strain; it should be noted that the load distribution Converted into additional stress through foundation mechanics model , which is input into Equation (3) to drive strain prediction.

[0070] In some embodiments, the method for obtaining the strain prediction model of the foundation pit caused by the load of the vehicle 5 includes: The foundation mechanics model is converted into , and input it into formula (3) to obtain the strain prediction model of the foundation pit caused by the load of vehicle 5, as shown in formula (4):

[0071] (4);

[0072] In formula (4), For the Vehicle load impact area, Indicates additional stress to be calculated The projection coordinates of the spatial point on the surface of the earth, is an integral variable representing the point of action of the vehicle load on the ground. Using Equations (4) and (3), the system can predict the strain at any point in time and space and dynamically compare it with the measured strain. When the error between the two exceeds the limit, the system will immediately provide feedback and adjust the construction parameters or issue a command to suspend construction.

[0073] In some embodiments, in step S3, the method for constructing the foundation pit deformation prediction model includes: , real-time speed , the support force of the hydraulic support unit Soil strain , displacement of supporting piles or underground continuous walls , through the fusion of multi-source data, synchronous processing, normalization and time series alignment, a unified time series input vector is formed , and input it into the LSTM neural network to output the foundation pit deformation prediction model; among them, the time series input vector As shown in formula (5):

[0074]

[0075] The foundation pit deformation prediction model is shown in formula (6):

[0076] (6);

[0077] In formula (6), is the input feature at the current moment, is the size of the time window, The displacement of the supporting piles or underground continuous wall at a predicted moment in the future.

[0078] In some embodiments, step S3 further includes: performing error back propagation according to actual monitoring data to update and optimize the foundation pit deformation prediction model. The optimized foundation pit deformation prediction model expression is shown in formula (7):

[0079] (7);

[0080] In formula (7), Transformed for the future, equivalent to , is the neural network weight parameter.

[0081] In some embodiments, the LSTM neural network includes a forget gate, an input gate, and an output gate, wherein:

[0082] Forget Gate As shown in formula (8):

[0083] (8);

[0084] In formula (8), Is the output of the forget gate, with a value range between 0 and 1, where 0 means completely forgotten and 1 means completely retained; and They are the weight matrix and bias term of the forget gate respectively; is the current input, is the hidden state at the previous moment; For vector splicing, vertical vector splicing is performed on the hidden state of the previous moment and the current input; It is the Sigmoid activation function, which ensures that the output is in the range of [0,1];

[0085] Input Gate And the candidate state is shown in formula (9):

[0086] (9);

[0087] In formula (9), The gate vector output by the input gate is constrained to be in the range of [0, 1] through the Sigmoid activation function; is the weight matrix of the input gate; is the bias term of the input gate, which is used to adjust the flexibility of the model; is the bias term of the candidate cell state; is the candidate cell state, which is nonlinearly transformed by tanh; is the weight matrix of the candidate cell state; is the current cell state; is the cell state at the previous moment; is the operator symbol, which means element-by-element multiplication;

[0088] Output Gate As shown in formula (10):

[0089] (10);

[0090] In formula (10), Is the output of the output gate, which controls the output amount of the hidden state; is the weight matrix of the output gate; is the bias term of the output gate.

[0091] In some embodiments, step S3 further includes: setting a warning threshold for the prediction result ,Will and For real-time comparison, When adjusting the support force of the hydraulic support unit in advance; is the tolerance item.

[0092] In some embodiments, as Figure 1 As shown, the foundation pit deformation control method based on dynamic load compensation also includes:

[0093] S4, stores the real-time monitoring data and vehicle 5 data, and uses the self-learning mechanism to continuously optimize the foundation pit deformation prediction model.

[0094] Based on the above-mentioned foundation pit deformation control method based on dynamic load compensation, Figure 2 As shown, the present invention also provides a foundation pit deformation control system based on dynamic load compensation, which adopts the above foundation pit deformation control method based on dynamic load compensation, including:

[0095] The hydraulic support unit is arranged inside the foundation pit supporting pile or underground continuous wall. The hydraulic support unit includes a hydraulic support device 1, a pressure sensor 2 and a displacement meter 3. The hydraulic support device 1 is used to provide supporting force, the pressure sensor 2 is used to monitor the pressure of the hydraulic support unit in real time, and the displacement meter 3 is used to monitor the displacement of the supporting pile or underground continuous wall in real time.

[0096] The monitoring unit includes a fiber optic sensor 4, a GPS locator 6, and a vehicle 5 weighing device. The fiber optic sensor 4 is arranged around the foundation pit to sense soil strain. The GPS locator 6 is installed on the vehicle 5 to obtain the location of the vehicle 5. The vehicle 5 weighing device is installed at the construction road node to weigh the vehicle 5.

[0097] The central control unit is communicatively connected to the hydraulic support unit and the monitoring unit, and is used to calculate the support force based on the data obtained by the hydraulic support unit and to construct a dynamic load distribution model based on the data obtained by the monitoring unit. The central control unit is also used to construct a foundation pit deformation prediction model, and to adjust the support force of the support structure according to the prediction results, while obtaining the optimized vehicle 5 path.

[0098] In the above-mentioned exemplary embodiment, the foundation pit deformation control system realizes intelligent control of foundation pit deformation caused by dynamic loads through the coordinated cooperation of the hydraulic support unit, the monitoring unit, and the central control unit. The hydraulic support unit monitors the stress state of the support structure in real time and dynamically adjusts the support force. The monitoring unit accurately obtains soil strain and vehicle 5 load data through the fiber optic sensor 4, the GPS locator 6, and the vehicle 5 weighing device. The central control unit comprehensively processes various types of data, constructs a dynamic load distribution model and a deformation prediction model, and then optimizes the support force control strategy and the vehicle 5 passage path. The system realizes the full process of automated management from data collection, analysis and prediction to active control, significantly improving the real-time, accuracy, and reliability of foundation pit deformation control. It is particularly suitable for deep foundation pit construction safety assurance in sensitive environments in densely populated urban areas.

[0099] In some embodiments, the foundation pit deformation control system based on dynamic load compensation further includes an early warning unit, which is in communication with the central control unit and is activated when the pressure of the hydraulic support unit is detected. , or displacement of supporting piles or underground continuous wall When the alarm is triggered, the central control unit controls the early warning unit to sound an alarm.

[0100] Through the description of multiple embodiments of the foundation pit deformation control method and system based on dynamic load compensation of the present invention, it can be seen that the foundation pit deformation control method and system embodiments based on dynamic load compensation of the present invention have at least one or more of the following advantages:

[0101] 1. The foundation pit deformation control method based on dynamic load compensation provided by the present invention, S1, monitors the pressure of the hydraulic support unit and the displacement of the support structure in real time, combines the dynamic load data of the vehicle 5, and dynamically adjusts the support force to effectively cope with the sudden load changes caused by the passage of the vehicle 5, thereby reducing the risk of excessive foundation pit deformation;

[0102] 2. In the foundation pit deformation control method based on dynamic load compensation provided by the present invention, step S2 integrates soil strain data, vehicle 5 position, speed and load data to construct a dynamic load distribution model, thereby improving the ability to accurately analyze the stress state of the foundation pit and providing a scientific basis for adjusting the support force;

[0103] 3. The foundation pit deformation control method based on dynamic load compensation provided by the present invention, step S3 uses an LSTM neural network to predict the impact trend of vehicle 5 passing on foundation pit deformation, adjusts the support force of the support structure in advance, realizes active protection, and avoids the hysteresis problem of traditional passive monitoring. While predicting the foundation pit deformation trend, it automatically optimizes the vehicle 5 passage path, reduces the adverse impact of the dynamic load of vehicle 5 on the support structure, and reduces construction risks;

[0104] 4. The foundation pit deformation control system based on dynamic load compensation provided by this invention achieves intelligent control of foundation pit deformation caused by dynamic loads through the coordinated cooperation of a hydraulic support unit, a monitoring unit, and a central control unit. The hydraulic support unit monitors the stress state of the support structure in real time and dynamically adjusts the support force. The monitoring unit accurately obtains soil strain and vehicle load data through fiber optic sensors 4, GPS locators 6, and vehicle 5 weighing devices. The central control unit comprehensively processes this data to construct a dynamic load distribution model and deformation prediction model, thereby optimizing the support force control strategy and the vehicle 5 travel path.

[0105] Finally, it should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0106] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the same. Although the present invention has been described in detail with reference to preferred embodiments, persons skilled in the art should understand that the specific implementation methods of the present invention may still be modified or some technical features may be replaced by equivalents without departing from the spirit of the technical solutions of the present invention, and all of these should fall within the scope of the technical solutions claimed for protection by the present invention.

Claims

1. The foundation pit deformation control method based on dynamic load compensation is characterized by: The following steps are involved: S1: Install hydraulic support units inside foundation pit support piles or underground continuous walls, and monitor the pressure of the hydraulic support units in real time. , and the displacement of the supporting piles or underground continuous wall , calculate the support force that the hydraulic support unit needs to provide ; S2 monitors the soil strain data around the foundation pit in real time, obtains the position, speed and load data of passing vehicles, and integrates these data to construct a dynamic load distribution model; S3 uses an LSTM neural network to build a foundation pit deformation prediction model to predict the impact of vehicle traffic on foundation pit deformation. Based on the prediction results, the support force of the support structure is adjusted in advance, and the vehicle path is automatically optimized to avoid insufficient support force of the support structure when vehicles pass through. In step S1, the support force required by the hydraulic support unit is calculated The method includes: taking the surface point at the center of the foundation pit excavation area as the origin and the direction of the road along which the vehicle travels as the , perpendicular to the road direction , vertically downward Axis establishes a spatial rectangular coordinate system; sets dual threshold indicators for foundation pit structure safety control, including pressure threshold and displacement threshold , when the pressure of the hydraulic support unit is monitored , or displacement of supporting piles or underground continuous wall When the pressure of the hydraulic support unit is , and the displacement of the supporting piles or underground continuous wall , through the dual-channel PID controller to integrate pressure control and displacement control, and use error feedback to calculate the support force , the calculation formula is shown in formula (1): （1）； In formula (1), is the pressure error, ; is the displacement error, ; To adjust the weight coefficient; 、 and is the PID control parameter of the pressure channel; 、 and is the displacement channel PID control parameter; For time; In step S2, the method for constructing a dynamic load distribution model includes: using a data fusion algorithm to integrate the foundation pit deformation with the vehicle load data and motion trajectory data to construct a dynamic load distribution model, as shown in formula (2): （2）； In formula (2), For vehicles at time Acting on the surface Load distribution (kPa); For the The time-varying load of the vehicle, a car; is the real-time horizontal coordinate of the vehicle, The real-time vertical coordinate of the vehicle; is the vehicle load influence radius, which is negatively correlated with vehicle speed.

2. The foundation pit deformation control method based on dynamic load compensation according to claim 1 is characterized in that: Step S2 also includes: using optical fiber sensors around the foundation pit to sense soil strain, obtaining the relationship between soil strain and vehicle load, and converting the dynamic load distribution of the vehicle into the relationship expression between soil strain and vehicle load, thereby obtaining a strain prediction model for the foundation pit caused by the vehicle load; wherein, the relationship expression between soil strain and vehicle load is shown in formula (3): （3）； In formula (3), is the soil strain, is the vertical depth, is the elastic modulus of soil, is Poisson's ratio, is the additional stress caused by vehicle load, is the horizontal stress of soil, is the vertical stress of soil.

3. The foundation pit deformation control method based on dynamic load compensation according to claim 2 is characterized in that: The method for obtaining the strain prediction model of the foundation pit caused by vehicle load includes: The foundation mechanics model is converted into , and input it into formula (3) to obtain the strain prediction model of the foundation pit caused by vehicle load, as shown in formula (4): （4）； In formula (4), For the Vehicle load impact area; Indicates additional stress to be calculated The projection coordinates of the spatial point on the surface; is the integral variable, which represents the point of action of the vehicle load on the ground.

4. The foundation pit deformation control method based on dynamic load compensation according to claim 1 is characterized in that: In step S3, the method for constructing the foundation pit deformation prediction model includes: , real-time speed , the support force of the hydraulic support unit Soil strain , displacement of supporting piles or underground continuous walls , through the fusion of multi-source data, synchronous processing, normalization and time series alignment, a unified time series input vector is formed , and input it into the LSTM neural network to output the foundation pit deformation prediction model; among them, the time series input vector As shown in formula (5): The foundation pit deformation prediction model is shown in formula (6): （6）； In formula (6), is the input feature at the current moment, is the size of the time window, The displacement of the supporting piles or underground continuous wall at a predicted moment in the future.

5. The foundation pit deformation control method based on dynamic load compensation according to claim 4 is characterized in that: Step S3 also includes: performing error back propagation based on actual monitoring data to update and optimize the foundation pit deformation prediction model. The optimized foundation pit deformation prediction model expression is shown in formula (7): （7）； In formula (7), Transformed for the future, equivalent to , is the neural network weight parameter.

6. The foundation pit deformation control method based on dynamic load compensation according to claim 4 is characterized in that: The LSTM neural network includes a forget gate, an input gate, and an output gate, among which, Forget Gate As shown in formula (8): （8）； In formula (8), Is the output of the forget gate, with a value range between 0 and 1, where 0 means completely forgotten and 1 means completely retained; and They are the weight matrix and bias term of the forget gate respectively; is the current input, is the hidden state at the previous moment; For vector splicing, vertical vector splicing is performed on the hidden state of the previous moment and the current input; It is the Sigmoid activation function, which ensures that the output is in the range of [0,1]; Input Gate And the candidate state is shown in formula (9): （9）； In formula (9), The gate vector output by the input gate is constrained to be in the range of [0, 1] through the Sigmoid activation function; is the weight matrix of the input gate; is the bias term of the input gate, which is used to adjust the flexibility of the model; is the bias term of the candidate cell state; is the candidate cell state, which is nonlinearly transformed by tanh; is the weight matrix of the candidate cell state; is the current cell state; is the cell state at the previous moment; is the operator symbol, which means element-by-element multiplication; Output Gate As shown in formula (10): （10）； In formula (10), Is the output of the output gate, which controls the output amount of the hidden state; is the weight matrix of the output gate; is the bias term of the output gate.

7. The foundation pit deformation control system based on dynamic load compensation is characterized by: Using the foundation pit deformation control method according to any one of claims 1 to 6, the foundation pit deformation control system based on dynamic load compensation includes: A hydraulic support unit is arranged inside the foundation pit support pile or underground continuous wall. The hydraulic support unit includes a hydraulic support device, a pressure sensor, and a displacement meter. The hydraulic support device is used to provide support force, the pressure sensor is used to monitor the pressure of the hydraulic support unit in real time, and the displacement meter is used to monitor the displacement of the support pile or underground continuous wall in real time. The monitoring unit includes a fiber optic sensor, a GPS locator, and a vehicle scale. The fiber optic sensor is deployed around the foundation pit to sense soil strain. The GPS locator is installed on the vehicle to obtain the vehicle's location. The vehicle scale is installed at the construction road node to weigh the vehicle. The central control unit is communicatively connected to the hydraulic support unit and the monitoring unit, and is used to calculate the support force based on the data obtained by the hydraulic support unit and to construct a dynamic load distribution model based on the data obtained by the monitoring unit. The central control unit is also used to construct a foundation pit deformation prediction model, and to adjust the support force of the support structure according to the prediction results, while obtaining an optimized vehicle path.

8. The foundation pit deformation control system based on dynamic load compensation according to claim 7 is characterized in that: It also includes an early warning unit, which is connected to the central control unit in communication. When the pressure of the hydraulic support unit is monitored, , or displacement of supporting piles or underground continuous wall When the alarm is triggered, the central control unit controls the early warning unit to sound an alarm.

Images