A frozen soil pipe pile construction parameter monitoring method and system

Abstract

The application provides a frozen soil pipe pile construction parameter monitoring method and system, the method comprises the following steps: a corresponding digital twin benchmark model library is constructed, before construction, based on the actual environmental parameters of the target pile, a benchmark sampling strategy and a power consumption control baseline for the whole construction period are generated; during the construction process, the low-frequency benchmark sampling of the target pile is performed according to the benchmark sampling strategy and the power consumption control baseline, so as to obtain the real-time state data of the target pile and output the corresponding actual construction parameters; the real-time inversion reconstruction processing of the target pile is performed according to the actual environmental parameters and the real-time state data, so as to output the corresponding target construction parameters; the implicit correlation law between the target construction parameters and the frozen soil thermal disturbance and the construction process is detected, so as to generate the corresponding targeted construction optimization strategy, and the actual construction parameters are dynamically adjusted according to the targeted construction optimization strategy. The application can adapt to the working conditions of low temperature and insufficient power supply in remote frozen soil areas, and guarantee the construction quality.

Description

Technical Field

[0001] This invention relates to the field of parameter monitoring technology, and in particular to a method and system for monitoring construction parameters of pipe piles in frozen soil. Background Technology

[0002] Permafrost regions are the core scenarios for the construction of major infrastructure projects in high-altitude and cold areas such as the Qinghai-Tibet Plateau and the Greater Khingan Mountains in Northeast my country. Pipe pile foundations, with their advantages of high construction efficiency, stable pile formation, and controllable permafrost disturbance, have become the most widely used foundation type in these regions. Accurate, continuous, and real-time monitoring of core parameters in permafrost pipe pile construction is crucial for controlling permafrost thermal disturbance, ensuring pile quality, and avoiding frost heave and thaw settlement problems. Intelligent monitoring systems based on distributed fiber optic sensing are currently the mainstream technical solution in this field.

[0003] However, existing intelligent systems for monitoring construction parameters of permafrost pipe piles suffer from a core bottleneck: a severe mismatch between power consumption and on-site power supply capabilities. The core equipment, such as the fiber optic demodulator, edge computing unit, and data transmission gateway, has extremely high requirements for continuous power supply. However, permafrost pipe pile construction is often located in remote, uninhabited areas with very low grid power coverage. The commonly used off-grid power supply solution of solar power + lithium batteries suffers from power outages during the winter polar night, and lithium battery capacity decays to less than 30% of its normal capacity at extreme temperatures of -20℃ and below, making it impossible to support continuous system operation. This easily leads to equipment downtime, data disconnection, and complete failure of real-time monitoring of construction parameters.

[0004] To address the aforementioned bottlenecks, existing conventional low-power solutions all revolve around hardware optimization, only achieving power reduction by lowering the sampling frequency and reducing the data acquisition and transmission frequency. This comes at the cost of sacrificing monitoring time resolution, data continuity, and accuracy, directly resulting in the lack of capture of key construction conditions and dynamic changes in permafrost. This creates a core contradiction between "low-power stable operation" and "high-precision real-time monitoring," which cannot be fundamentally resolved through hardware optimization. This severely restricts the large-scale application of intelligent monitoring technology and poses hidden dangers to the construction quality and long-term safe operation of projects in cold regions. Summary of the Invention

[0005] Based on this, the purpose of the present invention is to provide a method and system for monitoring construction parameters of pipe piles in frozen soil, so as to solve the problem that the existing technology can only reduce power consumption by reducing sampling frequency and reducing data acquisition and transmission frequency in low temperature environment, which easily leads to data capture loss.

[0006] The first aspect of the present invention proposes:

[0007] A method for monitoring construction parameters of pipe piles in frozen soil, the method comprising:

[0008] Historical big data on frozen soil pipe pile construction in multiple scenarios is collected and frozen soil water-heat-force coupling prior constraints are embedded simultaneously to construct a corresponding digital twin benchmark model library. Simultaneously, before construction, based on the actual environmental parameters of the target pile, a dedicated benchmark model is matched from the digital twin benchmark model library and distributed to the edge computing node to generate a benchmark sampling strategy and power consumption control baseline for the entire construction cycle.

[0009] During construction, low-frequency reference sampling of the target pile is performed according to the reference sampling strategy and the power consumption control baseline to obtain the real-time status data of the target pile. Simultaneously, the current construction progress, real-time monitoring risk level, and computing power load of the edge computing node are combined and input into the pre-trained power consumption control model to output the corresponding actual construction parameters.

[0010] Based on the actual environmental parameters and the real-time status data, the target pile is inverted and reconstructed in real time to output the corresponding target construction parameters;

[0011] The implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures is detected to generate a corresponding targeted construction optimization strategy. Simultaneously, the actual construction parameters are dynamically adjusted according to the targeted construction optimization strategy to complete the corresponding construction parameter monitoring and processing.

[0012] The beneficial effects of this invention are as follows: This technical solution effectively solves the problem of missing data capture caused by simply reducing the sampling frequency and data acquisition and transmission frequency to reduce power consumption in low-temperature environments in existing technologies; by constructing a digital twin benchmark model library and matching it with a dedicated benchmark model, combined with low-frequency benchmark sampling and real-time inversion reconstruction, the accurate capture and complete acquisition of target pile construction parameters are achieved while strictly controlling power consumption; with the help of power consumption management models and targeted construction optimization strategies, the actual construction parameters are dynamically adjusted, taking into account both low-power operation and high-precision monitoring requirements, avoiding data disconnection and distortion, improving the continuity and reliability of frozen soil pipe pile construction parameter monitoring, adapting to the working conditions of low temperature and insufficient power supply in remote frozen soil areas, and ensuring construction quality and safety.

[0013] Furthermore, the step of performing low-frequency reference sampling of the target pile according to the reference sampling strategy and the power consumption control baseline during construction to obtain real-time status data of the target pile includes:

[0014] Based on the exclusive benchmark model, the inherent sparse features of pile stress, frozen soil temperature field and freeze-thaw phase change corresponding to the entire construction cycle of the target pile are extracted, and the frozen soil water-thermal-mechanical coupling prior constraints are embedded simultaneously to construct an exclusive sparse characterization dictionary.

[0015] Based on the dedicated sparse representation dictionary, the entire pile body and the full parameter field of the frozen soil around the pile of the target pile are sparsely reconstructed to output the corresponding full data.

[0016] Based on the benchmark sampling strategy and the power consumption control baseline, the full data is physically compliantly verified, and the real-time status data of the target pile is output synchronously after the verification is passed.

[0017] Furthermore, the step of performing physical compliance verification on the full data according to the benchmark sampling strategy and the power consumption control baseline, and simultaneously outputting the real-time status data of the target pile after the verification is passed, includes:

[0018] Based on the aforementioned benchmark sampling strategy, combined with the current power consumption control baseline and the freeze-thaw risk level of the frozen soil around the piles, a hierarchical physical compliance verification framework with risk-power consumption dual drive is constructed, and verification domains of different risk levels are simultaneously divided.

[0019] The full data is matched to the corresponding verification domain to perform hierarchical differentiated verification, and corresponding compliant verification data and non-compliant verification data are output.

[0020] The non-compliant verification data is differentiated and corrected using the dedicated sparse representation dictionary, and then embedded into the compliant verification data to output the real-time status data of the target pile.

[0021] Furthermore, the step of performing real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data to output the corresponding target construction parameters includes:

[0022] Based on the actual environmental parameters, the real-time status data is processed to complete the spatiotemporal features of the phase transition sensitive domain. Simultaneously, according to the current construction progress and the real-time monitoring risk level, the embedded frozen soil water-thermal-mechanical coupling prior constraints are dynamically weighted to construct the corresponding real-time inversion objective function.

[0023] Based on the real-time inversion objective function, an inversion grid adapted to the target pile is constructed. Simultaneously, the iteration step size and convergence threshold of the inversion grid are dynamically adjusted according to the real-time computing power load of the edge computing nodes to output multi-field distribution reconstruction results.

[0024] Based on the multi-field distribution reconstruction results, and combined with the permissible threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction process, the target construction parameters are output accordingly.

[0025] Furthermore, the step of outputting the target construction parameters based on the multi-field distribution reconstruction results, combined with the allowable threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction procedure, includes:

[0026] Based on the multi-field distribution reconstruction results, the real-time dynamic boundary of the phase transition sensitive domain is delineated, and the benchmark multi-field distribution characteristics generated by the dedicated benchmark model under the same working conditions are compared simultaneously to reconstruct the time-varying risk quantification map of the phase transition sensitive domain.

[0027] Based on the time-varying risk quantification map, with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process as rigid constraints, and combined with the power consumption control baseline, the corresponding time-varying extrapolation is performed to output the corresponding feasible domain of construction parameters.

[0028] With the output objectives of minimizing thermal disturbance of frozen soil, maximizing construction efficiency, and optimizing monitoring power consumption, the corresponding optimal solution set is obtained based on the feasible region of the construction parameters, and the optimal solution set is simultaneously set as the target construction parameters.

[0029] Furthermore, the step of detecting the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures to generate a corresponding targeted construction optimization strategy includes:

[0030] The corresponding sensitive feature set is extracted from the target construction parameters, and the corresponding historical source domain data is matched in the digital twin benchmark model library simultaneously. The corresponding adaptive alignment processing is also completed to construct an incremental learning training set.

[0031] Based on the incremental learning training set, a corresponding association mining model is trained to uncover the implicit association between the target construction parameters, the frozen soil thermal disturbance, and the construction procedures.

[0032] The optimization target points contained in the implicit correlation patterns are detected, and the targeted construction optimization strategy is generated based on the optimization target points.

[0033] Furthermore, the step of detecting the optimization target points contained in the implicit correlation patterns, and generating the targeted construction optimization strategy based on the optimization target points, includes:

[0034] The entire link of the implicit correlation pattern is decomposed into causal hierarchy to distinguish optimization targets at different levels, and invalid targets are eliminated simultaneously to output the corresponding target set.

[0035] A coupled simulation model is constructed based on the target point set, and a full-cycle time-varying simulation is performed simultaneously to determine the optimal adjustable range of each target point.

[0036] Based on the optimal adjustable range, combined with the current construction progress and the real-time monitoring risk level, the targeted construction optimization strategy is integrated accordingly.

[0037] The second aspect of the present invention proposes:

[0038] A monitoring system for construction parameters of pipe piles in frozen soil, the system comprising:

[0039] The data acquisition module is used to collect historical big data on frozen soil pipe pile construction in multiple scenarios and simultaneously embed the prior constraints of frozen soil water-thermal-mechanical coupling to build a corresponding digital twin benchmark model library. Simultaneously, before construction, based on the actual environmental parameters of the target pile, a dedicated benchmark model is matched from the digital twin benchmark model library and sent to the edge computing node to generate a benchmark sampling strategy and power consumption control baseline for the entire construction cycle.

[0040] The output module is used to perform low-frequency reference sampling of the target pile according to the reference sampling strategy and the power consumption control baseline during the construction process, so as to obtain the real-time status data of the target pile, and simultaneously combine the current construction progress, real-time monitoring risk level, and computing power load of the edge computing node to input the pre-trained power consumption control model to output the corresponding actual construction parameters.

[0041] The processing module is used to perform real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data, so as to output the corresponding target construction parameters;

[0042] The adjustment module is used to detect the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures, so as to generate a corresponding targeted construction optimization strategy. Simultaneously, the actual construction parameters are dynamically adjusted according to the targeted construction optimization strategy to complete the corresponding construction parameter monitoring and processing.

[0043] Furthermore, the output module is specifically used for:

[0044] Based on the exclusive benchmark model, the inherent sparse features of pile stress, frozen soil temperature field and freeze-thaw phase change corresponding to the entire construction cycle of the target pile are extracted, and the frozen soil water-thermal-mechanical coupling prior constraints are embedded simultaneously to construct an exclusive sparse characterization dictionary.

[0045] Based on the dedicated sparse representation dictionary, the entire pile body and the full parameter field of the frozen soil around the pile of the target pile are sparsely reconstructed to output the corresponding full data.

[0046] Based on the benchmark sampling strategy and the power consumption control baseline, the full data is physically compliantly verified, and the real-time status data of the target pile is output synchronously after the verification is passed.

[0047] Furthermore, the output module is specifically used for:

[0048] Based on the aforementioned benchmark sampling strategy, combined with the current power consumption control baseline and the freeze-thaw risk level of the frozen soil around the piles, a hierarchical physical compliance verification framework with risk-power consumption dual drive is constructed, and verification domains of different risk levels are simultaneously divided.

[0049] The full data is matched to the corresponding verification domain to perform hierarchical differentiated verification, and corresponding compliant verification data and non-compliant verification data are output.

[0050] The non-compliant verification data is differentiated and corrected using the dedicated sparse representation dictionary, and then embedded into the compliant verification data to output the real-time status data of the target pile.

[0051] Furthermore, the processing module is specifically used for:

[0052] Based on the actual environmental parameters, the real-time status data is processed to complete the spatiotemporal features of the phase transition sensitive domain. Simultaneously, according to the current construction progress and the real-time monitoring risk level, the embedded frozen soil water-thermal-mechanical coupling prior constraints are dynamically weighted to construct the corresponding real-time inversion objective function.

[0053] Based on the real-time inversion objective function, an inversion grid adapted to the target pile is constructed. Simultaneously, the iteration step size and convergence threshold of the inversion grid are dynamically adjusted according to the real-time computing power load of the edge computing nodes to output multi-field distribution reconstruction results.

[0054] Based on the multi-field distribution reconstruction results, and combined with the permissible threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction process, the target construction parameters are output accordingly.

[0055] Furthermore, the processing module is specifically used for:

[0056] Based on the multi-field distribution reconstruction results, the real-time dynamic boundary of the phase transition sensitive domain is delineated, and the benchmark multi-field distribution characteristics generated by the dedicated benchmark model under the same working conditions are compared simultaneously to reconstruct the time-varying risk quantification map of the phase transition sensitive domain.

[0057] Based on the time-varying risk quantification map, with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process as rigid constraints, and combined with the power consumption control baseline, the corresponding time-varying extrapolation is performed to output the corresponding feasible domain of construction parameters.

[0058] With the output objectives of minimizing thermal disturbance of frozen soil, maximizing construction efficiency, and optimizing monitoring power consumption, the corresponding optimal solution set is obtained based on the feasible region of the construction parameters, and the optimal solution set is simultaneously set as the target construction parameters.

[0059] Furthermore, the adjustment module is specifically used for:

[0060] The corresponding sensitive feature set is extracted from the target construction parameters, and the corresponding historical source domain data is matched in the digital twin benchmark model library simultaneously. The corresponding adaptive alignment processing is also completed to construct an incremental learning training set.

[0061] Based on the incremental learning training set, a corresponding association mining model is trained to uncover the implicit association between the target construction parameters, the frozen soil thermal disturbance, and the construction procedures.

[0062] The optimization target points contained in the implicit correlation patterns are detected, and the targeted construction optimization strategy is generated based on the optimization target points.

[0063] Furthermore, the adjustment module is specifically used for:

[0064] The entire link of the implicit correlation pattern is decomposed into causal hierarchy to distinguish optimization targets at different levels, and invalid targets are eliminated simultaneously to output the corresponding target set.

[0065] A coupled simulation model is constructed based on the target point set, and a full-cycle time-varying simulation is performed simultaneously to determine the optimal adjustable range of each target point.

[0066] Based on the optimal adjustable range, combined with the current construction progress and the real-time monitoring risk level, the targeted construction optimization strategy is integrated accordingly.

[0067] The third aspect of the present invention proposes:

[0068] A computer includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the frozen soil pipe pile construction parameter monitoring method as described above.

[0069] The fourth aspect of the present invention proposes:

[0070] A readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method for monitoring construction parameters of frozen soil pipe piles as described above.

[0071] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0072] Figure 1 A flowchart of the method for monitoring construction parameters of pipe piles in frozen soil provided in the first embodiment of the present invention;

[0073] Figure 2 The structural block diagram of the frozen soil pipe pile construction parameter monitoring system provided in the third embodiment of the present invention is shown.

[0074] The following detailed description, in conjunction with the accompanying drawings, will further illustrate the present invention. Detailed Implementation

[0075] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Several embodiments of the invention are illustrated in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete.

[0076] It should be noted that when a component is said to be "fixed to" another component, it can be directly on the other component or there may be an intervening component. When a component is said to be "connected to" another component, it can be directly connected to the other component or there may be an intervening component. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0077] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0078] Please see Figure 1 The image shows a method for monitoring construction parameters of permafrost pipe piles provided in the first embodiment of the present invention. This method can dynamically adjust actual construction parameters by using a power consumption control model and a targeted construction optimization strategy. It takes into account both low power consumption operation and high-precision monitoring requirements, avoids data disconnection and distortion, improves the continuity and reliability of permafrost pipe pile construction parameter monitoring, adapts to the working conditions of low temperature and insufficient power supply in remote permafrost areas, and ensures construction quality and safety.

[0079] Specifically, this embodiment provides:

[0080] A method for monitoring construction parameters of pipe piles in frozen soil, the method comprising:

[0081] Step S10: Collect historical big data on frozen soil pipe pile construction in multiple scenarios, and simultaneously embed the frozen soil water-thermal-mechanical coupling prior constraints to construct a corresponding digital twin benchmark model library. Simultaneously, before construction, based on the actual environmental parameters of the target pile, match the exclusive benchmark model from the digital twin benchmark model library and send it to the edge computing node to generate a benchmark sampling strategy and power consumption control baseline for the entire construction cycle.

[0082] It is important to note that the core risk source for pipe pile engineering in permafrost regions lies in the thermal disturbance to the permafrost surrounding the pile caused by construction activities such as pile driving friction heat, pile splicing welding heat input, and concrete hydration heat. This disturbance disrupts the inherent water-thermal-mechanical balance of the permafrost, triggering freeze-thaw phase transitions and leading to irreversible engineering problems such as soil settlement around the pile, weakening of the pile-soil interface strength, and insufficient pile bearing capacity. Furthermore, the water-thermal-mechanical coupling behavior of permafrost exhibits clear physical laws and strong regional variations. The evolution patterns and risk characteristics of construction thermal disturbance differ fundamentally between perennial and seasonally frozen soils, different ground temperature zones, and different pile types and construction techniques. Traditional monitoring methods typically employ fixed sampling frequencies and general control strategies, which cannot adapt to the differentiated needs of different engineering scenarios, nor can they meet the long-term, low-power operation requirements of edge monitoring equipment in field construction scenarios. This step first involves collecting historical construction big data covering various scenarios with different permafrost types, pile parameters, construction techniques, and environmental conditions. Simultaneously, it embeds mature water-thermal-mechanical coupling theories from permafrost mechanics and physics as hard constraints, constructing a digital twin benchmark model library with clear physical interpretability, rather than a purely data-driven black-box model. This fundamentally ensures the model's adaptability to permafrost construction scenarios. Before target pile construction, based on the actual environmental parameters of the target pile (including permafrost type, annual average ground temperature, geological stratification, pile design parameters, bearing capacity requirements, and ambient temperature characteristics), a dedicated benchmark model highly compatible with the target pile scenario is matched from the model library. After lightweight processing, this model is distributed to edge computing nodes at the construction site, eliminating the need for remote cloud transmission. Furthermore, based on the dedicated benchmark model, a benchmark sampling strategy and power consumption control baseline adapted to the entire construction cycle of the target pile are generated in advance. This clarifies the sampling frequency, sensor wake-up strategy, and computing power allocation rules for different construction stages and risk levels, providing a clear benchmark for low-power monitoring during construction and achieving differentiated, proactive control for each pile.

[0083] Step S20: During construction, low-frequency reference sampling of the target pile is performed according to the reference sampling strategy and the power consumption control baseline to obtain the real-time status data of the target pile. Simultaneously, the current construction progress, real-time monitoring risk level, and computing power load of the edge computing node are combined and input into the pre-trained power consumption control model to output the corresponding actual construction parameters.

[0084] It should be noted that most of the pipe pile construction sites in permafrost areas are located in remote fields with extremely limited power supply. Monitoring equipment and edge computing nodes usually rely on solar energy and batteries for power supply. Power consumption control is the core prerequisite for ensuring the stable operation of the monitoring system throughout the entire construction cycle. The traditional high-frequency continuous sampling mode not only has extremely high power consumption, but also generates a large amount of redundant data, occupying limited edge computing power and transmission bandwidth. This step, based on the pre-issued benchmark sampling strategy and power consumption control baseline, prioritizes low-frequency benchmark sampling during construction. This minimizes equipment power consumption while ensuring no core monitoring data is lost. Simultaneously, the construction process is dynamic; different construction procedures (pile driving, pile splicing, concrete pouring, curing) have varying impacts on the thermal disturbance of the frozen soil, the risk level of the frozen soil around the piles changes in real time, and the computing load of edge computing nodes fluctuates with the amount of data processed. Therefore, the current construction progress, real-time monitoring risk level, and edge node computing load—three core dynamic variables—are input into the pre-trained power consumption control model. This dynamically adjusts the sampling frequency, sensor wake-up strategy, and computing power allocation rules, while outputting actual construction parameters adapted to the current working conditions. This achieves adaptive control of "low-risk, low-power sampling; high-risk, high-frequency encryption," ensuring both long-term stable operation of the monitoring system and real-time capture of construction risks and dynamic adaptation of construction parameters.

[0085] Step S30: Perform real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data to output the corresponding target construction parameters;

[0086] It should be noted that the real-time status data obtained by low-frequency benchmark sampling is discrete monitoring point data of the pile body and the frozen soil around the pile. It cannot fully present the global distribution characteristics of the temperature field, stress field and seepage field of the frozen soil around the pile, nor can it accurately delineate the dynamic boundary of the sensitive area of ​​freeze-thaw phase change. This global multi-field distribution information is the core basis for judging the risk of thermal disturbance of frozen soil and determining the optimal construction parameters. This step, based on the actual environmental parameters of the target pile and discrete real-time status monitoring data, combined with embedded frozen soil water-thermal-mechanical coupling prior constraints, performs real-time inversion and reconstruction of the multiphysics field of the entire pile body and the frozen soil around the pile. It transforms discrete monitoring point data into continuous multi-field distribution results across the entire domain, accurately restoring the entire process of thermal disturbance evolution of the frozen soil around the pile, freeze-thaw phase change development, and pile stress change. At the same time, combined with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process, it outputs target construction parameters that meet engineering safety constraints, providing clear quantitative targets for precise control of the construction process, and realizing the core upgrade from "discrete point monitoring" to "full-domain status perception".

[0087] Step S40: Detect the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures to generate a corresponding targeted construction optimization strategy. Simultaneously, dynamically adjust the actual construction parameters according to the targeted construction optimization strategy to complete the corresponding construction parameter monitoring and processing.

[0088] It should be noted that during the construction of pipe piles in frozen soil, there are not only explicit causal relationships between construction parameters, construction procedures, and thermal disturbance of frozen soil, but also a large number of implicit correlations involving multiple factors. For example, the combination of pile driving rate and interval time affects the cumulative effect of frictional heat generation; the coupling effect of pile splicing welding time and ambient temperature affects the depth of frozen soil thawing; and the connection sequence of different procedures affects the superposition effect of thermal disturbance. These implicit correlations cannot be accurately captured by traditional theoretical formulas and human experience, and are one of the core reasons for the uncontrolled thermal disturbance and frequent engineering defects during the construction of pipe piles in frozen soil areas. This step, based on real-time monitoring data, inversion reconstruction results, and construction parameter records throughout the construction process, uses machine learning algorithms to uncover the implicit correlations between target construction parameters, permafrost thermal disturbance evolution, and construction procedures. It accurately identifies optimization targets that can effectively control permafrost thermal disturbance and improve construction efficiency, generates targeted construction optimization strategies, and dynamically adjusts actual construction parameters in a closed loop based on these strategies. This achieves continuous optimization of construction parameters and proactive control of permafrost thermal disturbance, ultimately completing a closed loop of full-cycle monitoring and dynamic control of permafrost pipe pile construction parameters. This ensures both the safety and controllability of the construction process and the synergistic optimization of construction efficiency and project quality.

[0089] Second Embodiment

[0090] Furthermore, the step of performing low-frequency reference sampling of the target pile according to the reference sampling strategy and the power consumption control baseline during construction to obtain real-time status data of the target pile includes:

[0091] Based on the exclusive benchmark model, the inherent sparse features of pile stress, frozen soil temperature field and freeze-thaw phase change corresponding to the entire construction cycle of the target pile are extracted, and the frozen soil water-thermal-mechanical coupling prior constraints are embedded simultaneously to construct an exclusive sparse characterization dictionary.

[0092] Based on the dedicated sparse representation dictionary, the entire pile body and the full parameter field of the frozen soil around the pile of the target pile are sparsely reconstructed to output the corresponding full data.

[0093] Based on the benchmark sampling strategy and the power consumption control baseline, the full data is physically compliantly verified, and the real-time status data of the target pile is output synchronously after the verification is passed.

[0094] It should be noted that during the construction of frozen soil pipe piles, the evolution of pile stress, frozen soil temperature field, and freeze-thaw phase transition is not a random change, but follows the physical laws of frozen soil water-thermal-mechanical coupling. It has significant sparsity characteristics in the spatiotemporal dimension. Specifically, for example, the change of frozen soil temperature field is mainly concentrated in the phase transition sensitive area within a limited range around the pile, and the change of pile stress is mainly concentrated at the pile tip and pile-soil contact surface. The parameter changes in most areas have strong continuity and predictability, which is the core premise for achieving sparse sampling and reconstruction. This step, based on the target pile's dedicated benchmark model, extracts the inherent sparse characteristics of pile stress, frozen soil temperature field, and freeze-thaw phase change throughout the entire construction cycle. It identifies which areas and time periods are the core sensitive zones for parameter changes and embeds prior constraints of frozen soil water-thermal-mechanical coupling to ensure that the extraction of sparse features conforms to the physical laws of frozen soil engineering, rather than purely mathematical unconstrained feature decomposition. Finally, it constructs a sparse representation dictionary adapted to the target pile's specific scenario, providing a core benchmark for subsequent sparse reconstruction and realizing a fundamental shift from "indiscriminate general sampling" to "targeted sampling for sensitive features."

[0095] The dedicated sparse characterization dictionary has clearly defined the core sparse features and primitives of the target pile parameter field. Therefore, only a small amount of discrete monitoring point data needs to be obtained through low-frequency benchmark sampling. The discrete sampling data can be mapped to the primitive space of the sparse characterization dictionary through the sparse reconstruction algorithm to restore the continuous full data of the entire pile body and the frozen soil around the pile. There is no need to obtain full-domain data through high-frequency dense sampling, which fundamentally reduces the sampling power consumption and data transmission volume of the monitoring equipment. It achieves the core goal of "small amount of sampling, full restoration" and perfectly adapts to the low power consumption operation requirements of field construction scenarios.

[0096] The full dataset obtained from sparse reconstruction must conform to the physical laws of water-heat-mechanical coupling in permafrost. Otherwise, false data may mislead risk assessments, such as reconstructed permafrost temperature changes that do not conform to the basic laws of heat conduction or pile stress exceeding the material strength limit. Such non-compliant data will directly affect subsequent risk assessments and parameter control. This step, based on a benchmark sampling strategy and power consumption control baseline, performs physical compliance verification on the full dataset obtained from reconstruction. It verifies whether the data conforms to the physical laws of permafrost engineering and whether it is within a reasonable parameter threshold range. Only compliant data that passes the verification can be used as the real-time status data output for the target pile. This mechanism eliminates interference from false reconstruction data and ensures the authenticity, reliability, and physical interpretability of the real-time status data.

[0097] Furthermore, the step of performing physical compliance verification on the full data according to the benchmark sampling strategy and the power consumption control baseline, and simultaneously outputting the real-time status data of the target pile after the verification is passed, includes:

[0098] Based on the aforementioned benchmark sampling strategy, combined with the current power consumption control baseline and the freeze-thaw risk level of the frozen soil around the piles, a hierarchical physical compliance verification framework with risk-power consumption dual drive is constructed, and verification domains of different risk levels are simultaneously divided.

[0099] The full data is matched to the corresponding verification domain to perform hierarchical differentiated verification, and corresponding compliant verification data and non-compliant verification data are output.

[0100] The non-compliant verification data is differentiated and corrected using the dedicated sparse representation dictionary, and then embedded into the compliant verification data to output the real-time status data of the target pile.

[0101] It is important to note that the freeze-thaw risk levels vary significantly across different areas of the permafrost surrounding the pile. For example, the phase change sensitive zone within 0.5m of the pile is the area with the most intense thermal disturbance and the highest freeze-thaw risk. The accuracy of the data in this zone directly impacts the reliability of risk assessment and requires rigorous, multi-dimensional verification. Conversely, the stable permafrost zone far from the pile exhibits minimal parameter changes and a very low risk level, requiring only basic threshold verification instead of high-computational-power verification. Traditional uniform verification methods apply the same verification standard to all areas, leading to either insufficient verification in high-risk areas resulting in missed risk assessments or excessive verification in low-risk areas, wasting computational and power resources. This step takes the freeze-thaw risk level of the frozen soil around the pile as the core driver and the current power consumption control baseline as a hard constraint to construct a risk-power dual-driven hierarchical physical compliance verification framework. According to the level of freeze-thaw risk, the entire pile body and the frozen soil around the pile are divided into multiple different verification domains. Different verification rules, verification dimensions and verification accuracy are set for different verification domains, realizing differentiated control of "strict verification in high-risk areas and broad verification in low-risk areas". The framework balances the reliability of verification and the computing power and power consumption.

[0102] After completing the verification domain division, the full dataset obtained from sparse reconstruction is matched to the corresponding verification domain according to its spatial location and risk level, and differentiated hierarchical verification is performed: For data in high-risk verification domains, full-dimensional physical compliance verification is performed, including multiple dimensions such as heat conduction law verification, water-heat-force coupling balance verification, parameter threshold verification, and temporal continuity verification, to ensure the absolute reliability of the data; for data in medium-risk verification domains, core-dimensional verification is performed, including parameter threshold verification and temporal continuity verification; for data in low-risk verification domains, only basic threshold verification is performed to filter out obviously abnormal false data. Through hierarchical differentiated verification, compliant verification data and non-compliant verification data are accurately distinguished, which not only ensures the reliability of core data in high-risk areas, but also minimizes the computing power and power consumption of the verification process, perfectly adapting to the limited computing power resources of edge computing nodes.

[0103] Traditional methods for handling non-compliant data identified during verification typically involve direct rejection, which disrupts data continuity and may even result in the loss of crucial risk change information. This step abandons the direct rejection approach. Based on a dedicated sparse representation dictionary for the target stake, and combined with compliance verification data from adjacent areas and historical monitoring data over time, non-compliant data undergoes differentiated correction processing to restore true data that conforms to physical laws. The corrected data is then embedded into the compliance verification data, ultimately forming complete, continuous, and compliant real-time status data for the target stake. This approach ensures data integrity and continuity while eliminating interference from anomalous data, providing a high-quality data foundation for subsequent inversion reconstruction and risk assessment.

[0104] Furthermore, the step of performing real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data to output the corresponding target construction parameters includes:

[0105] Based on the actual environmental parameters, the real-time status data is processed to complete the spatiotemporal features of the phase transition sensitive domain. Simultaneously, according to the current construction progress and the real-time monitoring risk level, the embedded frozen soil water-thermal-mechanical coupling prior constraints are dynamically weighted to construct the corresponding real-time inversion objective function.

[0106] Based on the real-time inversion objective function, an inversion grid adapted to the target pile is constructed. Simultaneously, the iteration step size and convergence threshold of the inversion grid are dynamically adjusted according to the real-time computing power load of the edge computing nodes to output multi-field distribution reconstruction results.

[0107] Based on the multi-field distribution reconstruction results, and combined with the permissible threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction process, the target construction parameters are output accordingly.

[0108] It should be noted that the freeze-thaw phase transition sensitive region is the core area of ​​the thermal disturbance evolution of permafrost and the key to judging construction risks. Its temperature, moisture content and stress change most drastically, and discrete monitoring data are prone to the problem of missing phase transition boundary characteristics, which directly affects the accuracy of the inversion results. This step first completes the spatiotemporal characteristics of the phase change sensitive domain based on the actual environmental parameters of the target pile and the real-time status data. It focuses on completing the spatiotemporal change characteristics of the freeze-thaw boundary area, accurately capturing the dynamic changes of the phase change boundary, and avoiding the loss of features in the core risk area. At the same time, the different components of the frozen soil water-thermal-mechanical coupling prior constraints have different importance at different construction stages and risk levels. For example, during the pile driving construction stage, the heat conduction constraint of pile-soil friction heat generation has the highest weight, while during the pile splicing welding stage, the heat boundary constraint of heat source input has the highest weight. The higher the freeze-thaw risk level, the higher the weight of the phase change equilibrium constraint. Therefore, this step dynamically assigns weights to the different components of the prior constraints according to the current construction progress and the real-time monitoring risk level. This allows the inversion objective function to accurately adapt to the core characteristics of the current working condition, rather than a general function with fixed weights, which greatly improves the adaptability and accuracy of the inversion results to the current working condition.

[0109] Traditional multi-field inversion of permafrost typically employs fixed-precision grids and iterative rules, resulting in extremely high computational demands. This requires high-performance cloud servers and cannot achieve real-time inversion at edge computing nodes. Furthermore, the computational load on edge nodes is dynamic. When computational power is sufficient, high-precision grids can be used to improve inversion accuracy, while when computational power is limited, the computational load needs to be appropriately reduced to ensure the real-time nature of the inversion. This step, based on the real-time inversion objective function, constructs an inversion grid adapted to the target pile type and geological conditions. Simultaneously, it dynamically adjusts the iteration step size and convergence threshold of the inversion grid according to the real-time computing load of the edge nodes: when the edge nodes have sufficient computing power, a smaller iteration step size and a stricter convergence threshold are used to improve the spatial and computational accuracy of the inversion; when the edge nodes have limited computing power, while ensuring the inversion accuracy of the phase transition sensitive domain, the iteration step size of the non-sensitive region is appropriately increased and the convergence threshold is relaxed to reduce the computational load and ensure real-time output of the inversion results. This achieves a dynamic balance between inversion accuracy and computational efficiency, enabling multi-field inversions that originally required high-performance cloud computing to run in real-time at the edge nodes of the construction site, completely eliminating dependence on remote cloud environments and avoiding management lag issues caused by network transmission delays.

[0110] The multi-field distribution reconstruction results accurately restored the temperature field, stress field, and seepage field distribution of the frozen soil around the pile under the current working conditions, clarifying the development range of freeze-thaw phase change and the distribution state of pile stress. Based on this, this step combines the allowable threshold of frozen soil thermal disturbance (the upper limit of frozen soil thawing depth and the upper limit of ground temperature change specified in the standard) and the design threshold of pile bearing capacity corresponding to the current construction procedure to clarify the safety boundary of the current construction parameters, and output the target construction parameters that meet the rigid constraints of engineering safety and are suitable for the current working conditions. This provides a clear quantitative target for the precise control of the construction process and realizes a complete logical closed loop of "inverting to perceive risks, threshold constraining boundaries, and precise parameter calibration".

[0111] Furthermore, the step of outputting the target construction parameters based on the multi-field distribution reconstruction results, combined with the allowable threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction procedure, includes:

[0112] Based on the multi-field distribution reconstruction results, the real-time dynamic boundary of the phase transition sensitive domain is delineated, and the benchmark multi-field distribution characteristics generated by the dedicated benchmark model under the same working conditions are compared simultaneously to reconstruct the time-varying risk quantification map of the phase transition sensitive domain.

[0113] Based on the time-varying risk quantification map, with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process as rigid constraints, and combined with the power consumption control baseline, the corresponding time-varying extrapolation is performed to output the corresponding feasible domain of construction parameters.

[0114] With the output objectives of minimizing thermal disturbance of frozen soil, maximizing construction efficiency, and optimizing monitoring power consumption, the corresponding optimal solution set is obtained based on the feasible region of the construction parameters, and the optimal solution set is simultaneously set as the target construction parameters.

[0115] It should be noted that the multi-field distribution reconstruction results have already presented the full-domain multi-field distribution of the frozen soil around the piles. Based on this, this step accurately delineates the real-time dynamic boundary of the sensitive area where the frozen soil undergoes freeze-thaw phase change, clarifying the influence range and thawing depth of thermal disturbance. At the same time, the real-time reconstructed multi-field distribution results are compared with the benchmark multi-field distribution characteristics of the dedicated benchmark model under the same working conditions to clarify the deviation between the current thermal disturbance evolution and the benchmark prediction, quantify the freeze-thaw risk level of different regions, and finally reconstruct a time-varying risk quantification map of the phase change sensitive area. This intuitively and accurately presents the spatiotemporal distribution and evolution trend of the risk of the frozen soil around the piles, providing a precise risk basis for subsequent construction parameter constraints and optimization.

[0116] The adjustment of construction parameters for frozen soil pipe piles must first and foremost meet the rigid constraints of engineering safety. The allowable thresholds for frozen soil thermal disturbance and the design thresholds for pile bearing capacity must not be exceeded; these are inviolable red lines. Simultaneously, the control requirements for construction efficiency and monitoring power consumption must also be considered. This step, based on a time-varying risk quantification map, uses the two rigid constraints as insurmountable boundary conditions. Combined with the current power consumption control baseline, it performs time-varying extrapolations on the evolution of frozen soil thermal disturbance and pile stress changes under different combinations of construction parameters. This clarifies which combinations of construction parameters can consistently meet the rigid constraint requirements, ultimately outputting the feasible domain of the construction parameters. This defines a safe range for subsequent optimal parameter solutions, fundamentally ensuring the safety and compliance of the construction parameters.

[0117] Within the defined feasible region of construction parameters, there are countless parameter combinations that satisfy safety constraints. However, engineering practice requires optimal parameter combinations that balance multiple objectives. This step takes "minimizing thermal disturbance of frozen soil" to ensure the long-term service safety of the project, "maximizing construction efficiency" to ensure project progress, and "optimizing monitoring power consumption" to ensure the stable operation of the monitoring system as three core optimization objectives. Through a multi-objective optimization algorithm, a Pareto optimal solution set is obtained within the feasible region of construction parameters. Then, combined with the current construction progress and risk level, the optimal parameter combination that best suits the current working conditions is selected from the optimal solution set and set as the final target construction parameters. This achieves the synergistic optimization of the three core objectives of construction safety, construction efficiency, and monitoring power consumption, and completely solves the problem of traditional construction parameter determination that prioritizes either "safety over efficiency" or "progress over risk," resulting in a trade-off.

[0118] Furthermore, the step of detecting the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures to generate a corresponding targeted construction optimization strategy includes:

[0119] The corresponding sensitive feature set is extracted from the target construction parameters, and the corresponding historical source domain data is matched in the digital twin benchmark model library simultaneously. The corresponding adaptive alignment processing is also completed to construct an incremental learning training set.

[0120] Based on the incremental learning training set, a corresponding association mining model is trained to uncover the implicit association between the target construction parameters, the frozen soil thermal disturbance, and the construction procedures.

[0121] The optimization target points contained in the implicit correlation patterns are detected, and the targeted construction optimization strategy is generated based on the optimization target points.

[0122] It's important to note that the real-time data generated during the target pile construction process is incremental and small-sample, while the historical big data in the digital twin benchmark model library is source domain data. Due to differences in permafrost type, geological conditions, and construction techniques, there is an inherent domain offset problem between the two. Directly mixing historical data with real-time data for training would lead to model bias, and the discovered correlation patterns would not match the actual scenario of the target pile. This step first extracts the core feature set sensitive to permafrost thermal disturbance and changes in construction procedures from the real-time output target construction parameters, clarifying the core variables for correlation mining. Simultaneously, in the digital twin benchmark model library, the historical source domain data with the highest similarity to the target pile scenario is matched. Through a domain adaptive alignment algorithm, the distribution differences between the historical source domain data and the target pile target domain data are eliminated, solving the domain offset problem. Then, the aligned historical data is combined with the real-time incremental data of the target pile to construct an incremental learning training set. This fully utilizes the prior information of historical big data while ensuring the adaptability of the training set to the target pile scenario, providing a high-quality dataset for subsequent correlation mining model training.

[0123] Based on the constructed incremental learning training set, an association mining model adapted to permafrost construction scenarios is trained. The model can simultaneously capture explicit causal relationships between variables and implicit correlation patterns of multi-factor coupling. For example, the influence of the timing of different construction procedures on the superposition effect of thermal disturbance, the lag effect of construction parameter combinations on the development of freeze-thaw phase change, and the amplification effect of the coupling effect of ambient temperature and construction parameters on thermal disturbance of permafrost. These implicit patterns, which cannot be accurately captured by human experience and traditional theoretical formulas, can be accurately identified and quantified through the association mining model. At the same time, the model training process embeds prior constraints of permafrost water-thermal-mechanical coupling to ensure that the mined correlation patterns conform to the physical laws of permafrost engineering and have clear physical interpretability, rather than purely data-driven spurious correlations, thus avoiding the uncontrollable risks brought about by "black box models".

[0124] After uncovering the complete implicit correlation patterns, this step further identifies the core variables within these patterns that have a significant impact on frozen soil thermal disturbance control, construction efficiency improvement, and power consumption optimization. These are the optimization targets, clarifying which construction parameters and process control nodes can be actively adjusted to achieve significant optimization results. For example, discovering that "the combination of pile driving interval time and the penetration depth of each pile section has a 62% weighting in terms of cumulative thermal disturbance of the frozen soil around the pile" allows us to identify pile driving interval time and the penetration depth of each pile section as core optimization targets. This provides a clear direction for the subsequent generation of targeted optimization strategies, realizing the practical transformation from "pattern discovery" to "target location."

[0125] Furthermore, the step of detecting the optimization target points contained in the implicit correlation patterns, and generating the targeted construction optimization strategy based on the optimization target points, includes:

[0126] The entire link of the implicit correlation pattern is decomposed into causal hierarchy to distinguish optimization targets at different levels, and invalid targets are eliminated simultaneously to output the corresponding target set.

[0127] A coupled simulation model is constructed based on the target point set, and a full-cycle time-varying simulation is performed simultaneously to determine the optimal adjustable range of each target point.

[0128] Based on the optimal adjustable range, combined with the current construction progress and the real-time monitoring risk level, the targeted construction optimization strategy is integrated accordingly.

[0129] It is important to note that the implicit correlation patterns obtained through association mining exhibit clear causal hierarchies between different variables. Some are core causal targets that directly influence the results, some are secondary targets that indirectly influence the results, and some are invalid targets that only have correlations but no causal relationship. If all variables are treated as optimization targets without distinction, it will lead to scattered optimization strategies, failure to solve core problems, and even the risk of exacerbating risks through reverse adjustments. This step decomposes the entire link of implicit correlation patterns into causal hierarchies. Based on causal inference algorithms, it distinguishes different levels of root cause targets, intermediate targets, and apparent targets, clarifies the causal transmission path of different targets, and eliminates invalid targets that only have spurious correlations and no causal relationship. Finally, it selects a set of target targets with clear causal relationships, which can be actively adjusted and have significant optimization effects. This ensures the effectiveness and relevance of the optimization targets from the root, avoiding the waste of resources and security risks caused by ineffective optimization.

[0130] After selecting the target point set, it is necessary to clarify the safe adjustable range of each target point to avoid exceeding the rigid constraints of engineering safety during the adjustment process. Furthermore, different target points have coupling effects; adjusting a single target point can affect the optimization effect of other target points. Therefore, it is essential to determine the adjustable range for collaborative optimization through coupled simulation. This step, based on the target point set and combined with the prior constraints of frozen soil water-thermal-mechanical coupling, constructs a multi-target coupled simulation model. It performs time-varying simulations of different adjustment combinations of different target points throughout the entire construction cycle, simulating the evolution of frozen soil thermal disturbance, pile stress changes, and construction efficiency changes under different parameter adjustments. This clarifies the adjustable range of each target point without exceeding the rigid constraints of safety, and simultaneously identifies the optimal adjustable range for multi-target collaborative optimization. This provides a clear quantitative boundary for subsequent optimization strategy generation, ensuring that the optimization strategy is always executed within a safe and controllable range.

[0131] After determining the optimal adjustable range, this step combines the current construction progress at each stage and the real-time monitored risk level to generate targeted construction optimization strategies: When construction is in the pile driving stage and the freeze-thaw risk level is high, the optimization strategy focuses on core targets such as pile driving rate and interval time, selecting the parameter combination that minimizes thermal disturbance within the optimal adjustable range to prioritize control of construction risks; when construction is in the pile splicing stage and the risk level is low, the optimization strategy focuses on targets such as welding time and post-weld cooling time, selecting the parameter combination that maximizes construction efficiency while ensuring safety; at the same time, the optimization strategy clarifies the adjustment range, adjustment sequence, and verification nodes for each target, forming a closed-loop management and control scheme that can be directly implemented, ultimately achieving full-cycle dynamic optimization and proactive risk management of frozen soil pipe pile construction parameters, perfectly adapting to the collaborative management and control needs of construction safety, quality, and progress in frozen soil pipe pile projects.

[0132] Please see Figure 2 The third embodiment of the present invention provides:

[0133] A monitoring system for construction parameters of pipe piles in frozen soil, the system comprising:

[0134] The data acquisition module is used to collect historical big data on frozen soil pipe pile construction in multiple scenarios and simultaneously embed the prior constraints of frozen soil water-thermal-mechanical coupling to build a corresponding digital twin benchmark model library. Simultaneously, before construction, based on the actual environmental parameters of the target pile, a dedicated benchmark model is matched from the digital twin benchmark model library and sent to the edge computing node to generate a benchmark sampling strategy and power consumption control baseline for the entire construction cycle.

[0135] The output module is used to perform low-frequency reference sampling of the target pile according to the reference sampling strategy and the power consumption control baseline during the construction process, so as to obtain the real-time status data of the target pile, and simultaneously combine the current construction progress, real-time monitoring risk level, and computing power load of the edge computing node to input the pre-trained power consumption control model to output the corresponding actual construction parameters.

[0136] The processing module is used to perform real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data, so as to output the corresponding target construction parameters;

[0137] The adjustment module is used to detect the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures, so as to generate a corresponding targeted construction optimization strategy. Simultaneously, the actual construction parameters are dynamically adjusted according to the targeted construction optimization strategy to complete the corresponding construction parameter monitoring and processing.

[0138] Furthermore, the output module is specifically used for:

[0139] Based on the exclusive benchmark model, the inherent sparse features of pile stress, frozen soil temperature field and freeze-thaw phase change corresponding to the entire construction cycle of the target pile are extracted, and the frozen soil water-thermal-mechanical coupling prior constraints are embedded simultaneously to construct an exclusive sparse characterization dictionary.

[0140] Based on the dedicated sparse representation dictionary, the entire pile body and the full parameter field of the frozen soil around the pile of the target pile are sparsely reconstructed to output the corresponding full data.

[0141] Based on the benchmark sampling strategy and the power consumption control baseline, the full data is physically compliantly verified, and the real-time status data of the target pile is output synchronously after the verification is passed.

[0142] Furthermore, the output module is specifically used for:

[0143] Based on the aforementioned benchmark sampling strategy, combined with the current power consumption control baseline and the freeze-thaw risk level of the frozen soil around the piles, a hierarchical physical compliance verification framework with risk-power consumption dual drive is constructed, and verification domains of different risk levels are simultaneously divided.

[0144] The full data is matched to the corresponding verification domain to perform hierarchical differentiated verification, and corresponding compliant verification data and non-compliant verification data are output.

[0145] The non-compliant verification data is differentiated and corrected using the dedicated sparse representation dictionary, and then embedded into the compliant verification data to output the real-time status data of the target pile.

[0146] Furthermore, the processing module is specifically used for:

[0147] Based on the actual environmental parameters, the real-time status data is processed to complete the spatiotemporal features of the phase transition sensitive domain. Simultaneously, according to the current construction progress and the real-time monitoring risk level, the embedded frozen soil water-thermal-mechanical coupling prior constraints are dynamically weighted to construct the corresponding real-time inversion objective function.

[0148] Based on the real-time inversion objective function, an inversion grid adapted to the target pile is constructed. Simultaneously, the iteration step size and convergence threshold of the inversion grid are dynamically adjusted according to the real-time computing power load of the edge computing nodes to output multi-field distribution reconstruction results.

[0149] Based on the multi-field distribution reconstruction results, and combined with the permissible threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction process, the target construction parameters are output accordingly.

[0150] Furthermore, the processing module is specifically used for:

[0151] Based on the multi-field distribution reconstruction results, the real-time dynamic boundary of the phase transition sensitive domain is delineated, and the benchmark multi-field distribution characteristics generated by the dedicated benchmark model under the same working conditions are compared simultaneously to reconstruct the time-varying risk quantification map of the phase transition sensitive domain.

[0152] Based on the time-varying risk quantification map, with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process as rigid constraints, and combined with the power consumption control baseline, the corresponding time-varying extrapolation is performed to output the corresponding feasible domain of construction parameters.

[0153] With the output objectives of minimizing thermal disturbance of frozen soil, maximizing construction efficiency, and optimizing monitoring power consumption, the corresponding optimal solution set is obtained based on the feasible region of the construction parameters, and the optimal solution set is simultaneously set as the target construction parameters.

[0154] Furthermore, the adjustment module is specifically used for:

[0155] The corresponding sensitive feature set is extracted from the target construction parameters, and the corresponding historical source domain data is matched in the digital twin benchmark model library simultaneously. The corresponding adaptive alignment processing is also completed to construct an incremental learning training set.

[0156] Based on the incremental learning training set, a corresponding association mining model is trained to uncover the implicit association between the target construction parameters, the frozen soil thermal disturbance, and the construction procedures.

[0157] The optimization target points contained in the implicit correlation patterns are detected, and the targeted construction optimization strategy is generated based on the optimization target points.

[0158] Furthermore, the adjustment module is specifically used for:

[0159] The entire link of the implicit correlation pattern is decomposed into causal hierarchy to distinguish optimization targets at different levels, and invalid targets are eliminated simultaneously to output the corresponding target set.

[0160] A coupled simulation model is constructed based on the target point set, and a full-cycle time-varying simulation is performed simultaneously to determine the optimal adjustable range of each target point.

[0161] Based on the optimal adjustable range, combined with the current construction progress and the real-time monitoring risk level, the targeted construction optimization strategy is integrated accordingly.

[0162] The fourth embodiment of the present invention provides a computer, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the frozen soil pipe pile construction parameter monitoring method as described above.

[0163] The fifth embodiment of the present invention provides a readable storage medium on which a computer program is stored, which, when executed by a processor, implements the method for monitoring construction parameters of frozen soil pipe piles as described above.

[0164] In summary, the frozen soil pipe pile construction parameter monitoring method and system provided in the above embodiments of the present invention can dynamically adjust the actual construction parameters by means of a power consumption control model and a targeted construction optimization strategy, taking into account both low power consumption operation and high precision monitoring requirements, avoiding data disconnection and distortion, improving the continuity and reliability of frozen soil pipe pile construction parameter monitoring, adapting to the working conditions of low temperature and insufficient power supply in remote frozen soil areas, and ensuring construction quality and safety.

[0165] It should be noted that the above modules can be functional modules or program modules, and can be implemented through software or hardware. For modules implemented through hardware, the above modules can reside in the same processor; or the above modules can be located in different processors in any combination.

[0166] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0167] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0168] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0169] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0170] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for monitoring construction parameters of pipe piles in frozen soil, characterized in that, The method includes: Historical big data on frozen soil pipe pile construction in multiple scenarios is collected and frozen soil water-heat-force coupling prior constraints are embedded simultaneously to construct a corresponding digital twin benchmark model library. Simultaneously, before construction, based on the actual environmental parameters of the target pile, a dedicated benchmark model is matched from the digital twin benchmark model library and distributed to the edge computing node to generate a benchmark sampling strategy and power consumption control baseline for the entire construction cycle. During construction, low-frequency reference sampling of the target pile is performed according to the reference sampling strategy and the power consumption control baseline to obtain the real-time status data of the target pile. Simultaneously, the current construction progress, real-time monitoring risk level, and computing power load of the edge computing node are combined and input into the pre-trained power consumption control model to output the corresponding actual construction parameters. Based on the actual environmental parameters and the real-time status data, the target pile is inverted and reconstructed in real time to output the corresponding target construction parameters; The implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures is detected to generate a corresponding targeted construction optimization strategy. Simultaneously, the actual construction parameters are dynamically adjusted according to the targeted construction optimization strategy to complete the corresponding construction parameter monitoring and processing. The step of performing real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data to output the corresponding target construction parameters includes: Based on the actual environmental parameters, the real-time status data is processed to complete the spatiotemporal features of the phase transition sensitive domain. Simultaneously, according to the current construction progress and the real-time monitoring risk level, the embedded frozen soil water-thermal-mechanical coupling prior constraints are dynamically weighted to construct the corresponding real-time inversion objective function. Based on the real-time inversion objective function, an inversion grid adapted to the target pile is constructed. Simultaneously, the iteration step size and convergence threshold of the inversion grid are dynamically adjusted according to the real-time computing power load of the edge computing nodes to output multi-field distribution reconstruction results. Based on the multi-field distribution reconstruction results, and combined with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process, the target construction parameters are output accordingly. The step of detecting the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures to generate a corresponding targeted construction optimization strategy includes: The corresponding sensitive feature set is extracted from the target construction parameters, and the corresponding historical source domain data is matched in the digital twin benchmark model library simultaneously. The corresponding adaptive alignment processing is also completed to construct an incremental learning training set. Based on the incremental learning training set, a corresponding association mining model is trained to uncover the implicit association between the target construction parameters, the frozen soil thermal disturbance, and the construction procedures. The optimization target points contained in the implicit correlation patterns are detected, and the targeted construction optimization strategy is generated based on the optimization target points.

2. The method for monitoring construction parameters of pipe piles in frozen soil according to claim 1, characterized in that, The step of performing low-frequency reference sampling of the target pile according to the reference sampling strategy and the power consumption control baseline during construction to obtain real-time status data of the target pile includes: Based on the exclusive benchmark model, the inherent sparse features of pile stress, frozen soil temperature field and freeze-thaw phase change corresponding to the entire construction cycle of the target pile are extracted, and the frozen soil water-thermal-mechanical coupling prior constraints are embedded simultaneously to construct an exclusive sparse characterization dictionary. Based on the dedicated sparse representation dictionary, the entire pile body and the full parameter field of the frozen soil around the pile of the target pile are sparsely reconstructed to output the corresponding full data. Based on the benchmark sampling strategy and the power consumption control baseline, the full data is physically compliantly verified, and the real-time status data of the target pile is output synchronously after the verification is passed.

3. The method for monitoring construction parameters of frozen soil pipe piles according to claim 2, characterized in that, The step of performing physical compliance verification on the full data according to the benchmark sampling strategy and the power consumption control baseline, and simultaneously outputting the real-time status data of the target pile after the verification is passed, includes: Based on the aforementioned benchmark sampling strategy, combined with the current power consumption control baseline and the freeze-thaw risk level of the frozen soil around the piles, a hierarchical physical compliance verification framework with risk-power consumption dual drive is constructed, and verification domains of different risk levels are simultaneously divided. The full data is matched to the corresponding verification domain to perform hierarchical differentiated verification, and corresponding compliant verification data and non-compliant verification data are output. The non-compliant verification data is differentiated and corrected using the dedicated sparse representation dictionary, and then embedded into the compliant verification data to output the real-time status data of the target pile.

4. The method for monitoring construction parameters of pipe piles in frozen soil according to claim 1, characterized in that, The step of outputting the target construction parameters based on the multi-field distribution reconstruction results, combined with the permissible threshold for frozen soil thermal disturbance and the design threshold for pile bearing capacity of the current construction procedure, includes: Based on the multi-field distribution reconstruction results, the real-time dynamic boundary of the phase transition sensitive domain is delineated, and the benchmark multi-field distribution characteristics generated by the dedicated benchmark model under the same working conditions are compared simultaneously to reconstruct the time-varying risk quantification map of the phase transition sensitive domain. Based on the time-varying risk quantification map, with the permissible threshold of frozen soil thermal disturbance and the design threshold of pile bearing capacity for the current construction process as rigid constraints, and combined with the power consumption control baseline, the corresponding time-varying extrapolation is performed to output the corresponding feasible domain of construction parameters. With the output objectives of minimizing thermal disturbance of frozen soil, maximizing construction efficiency, and optimizing monitoring power consumption, the corresponding optimal solution set is obtained based on the feasible region of the construction parameters, and the optimal solution set is simultaneously set as the target construction parameters.

5. The method for monitoring construction parameters of frozen soil pipe piles according to claim 1, characterized in that, The step of detecting the optimization target points contained in the implicit correlation patterns and generating the targeted construction optimization strategy based on the optimization target points includes: The entire link of the implicit correlation pattern is decomposed into causal hierarchy to distinguish optimization targets at different levels, and invalid targets are eliminated simultaneously to output the corresponding target set. A coupled simulation model is constructed based on the target point set, and a full-cycle time-varying simulation is performed simultaneously to determine the optimal adjustable range of each target point. Based on the optimal adjustable range, combined with the current construction progress and the real-time monitoring risk level, the targeted construction optimization strategy is integrated accordingly.

6. A monitoring system for construction parameters of pipe piles in frozen soil, characterized in that, For implementing the method for monitoring construction parameters of frozen soil pipe piles as described in any one of claims 1 to 5, the system comprises: The data acquisition module is used to collect historical big data on frozen soil pipe pile construction in multiple scenarios and simultaneously embed the prior constraints of frozen soil water-thermal-mechanical coupling to build a corresponding digital twin benchmark model library. Simultaneously, before construction, based on the actual environmental parameters of the target pile, a dedicated benchmark model is matched from the digital twin benchmark model library and sent to the edge computing node to generate a benchmark sampling strategy and power consumption control baseline for the entire construction cycle. The output module is used to perform low-frequency reference sampling of the target pile according to the reference sampling strategy and the power consumption control baseline during the construction process, so as to obtain the real-time status data of the target pile, and simultaneously combine the current construction progress, real-time monitoring risk level, and computing power load of the edge computing node to input the pre-trained power consumption control model to output the corresponding actual construction parameters. The processing module is used to perform real-time inversion and reconstruction processing of the target pile based on the actual environmental parameters and the real-time status data, so as to output the corresponding target construction parameters; The adjustment module is used to detect the implicit correlation between the target construction parameters, frozen soil thermal disturbance, and construction procedures, so as to generate a corresponding targeted construction optimization strategy. Simultaneously, the actual construction parameters are dynamically adjusted according to the targeted construction optimization strategy to complete the corresponding construction parameter monitoring and processing.

7. A computer comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for monitoring construction parameters of frozen soil pipe piles as described in any one of claims 1 to 5.

8. A readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the method for monitoring construction parameters of frozen soil pipe piles as described in any one of claims 1 to 5.

Images