Intelligent control method and system for piling depth precision of constructional engineering

Abstract

The invention relates to the technical field of crossing of artificial intelligence and intelligent construction, discloses an intelligent control method and system for piling depth precision of constructional engineering, and aims to solve the problems of insufficient depth control precision, frequent repeated correction and the like caused by geological dynamic change, hammering energy fluctuation and pile response lag in existing piling construction. The method comprises the steps that hammering force, acceleration, displacement and environment vibration signals are collected in real time through a multi-source sensing array, and instantaneous penetration displacement is calculated based on acceleration signal integration and by means of a zero-speed correction algorithm after stratigraphic profile information is combined and filtering and time alignment are conducted; a pile-soil dynamic coupling model is constructed, damping and side friction parameters are identified on line, the theoretical limit penetration depth is predicted, a hydraulic hammer energy adjusting instruction is generated according to the theoretical limit penetration depth, and closed-loop control is achieved. By fusing real-time sensing and prior geological data, centimeter-level depth control is achieved, excessive driving or insufficient driving is remarkably reduced, and the construction efficiency and the pile foundation quality are improved.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of artificial intelligence and intelligent construction, specifically relating to an intelligent control method and system for the depth and accuracy of pile driving in building engineering. Background Technology

[0002] With the upgrading of urban construction and infrastructure standards, the requirements for the accuracy of pile foundation construction in building engineering are becoming increasingly stringent. Controlling the pile driving depth is a crucial step in ensuring the bearing capacity of the pile foundation and the safety of the project. Currently, pile driving operations rely heavily on manual experience and preset mechanical parameters, lacking the comprehensive perception and adaptive control capabilities to address dynamic geological changes, real-time pile response, and equipment status. Especially under complex geological conditions, existing technologies often struggle to accurately determine the actual penetration state, easily leading to depth deviations, which in turn affect pile quality and construction efficiency.

[0003] While existing technologies have attempted to incorporate various sensors for monitoring, significant shortcomings remain: First, single sensing methods are susceptible to interference, resulting in limited data reliability; second, the failure to effectively integrate prior geological information with real-time construction data leads to poor adaptability of prediction models in actual working conditions; third, control strategies are mostly based on static thresholds or empirical formulas, making it difficult to cope with nonlinear responses caused by soil layer changes; and fourth, the lack of a closed-loop evaluation mechanism between depth error and long-term pile foundation performance restricts the level of intelligence and precision in the construction process.

[0004] Therefore, how to achieve real-time fusion, dynamic modeling, and adaptive control of multi-source information during pile driving has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] This invention provides an intelligent control method and system for pile driving depth accuracy in building engineering, aiming to solve the technical problems of insufficient pile driving depth control accuracy, frequent repeated corrections, and low construction efficiency caused by factors such as dynamic changes in geological conditions, fluctuations in hammer energy transfer efficiency, and pile response lag during existing pile driving construction. Existing technologies generally use preset penetration thresholds or empirical formulas to determine pile driving termination, lacking real-time perception and dynamic modeling capabilities of the pile-soil interaction process. This makes it impossible to effectively distinguish between instantaneous disturbances and the actual penetration state, and also makes it difficult to achieve millimeter-level depth control targets under complex geological conditions.

[0006] As one embodiment of the present invention, the intelligent control method for pile driving depth accuracy in building engineering includes the following steps:

[0007] S1: Real-time acquisition of hammer force signals, pile acceleration signals, pile top displacement signals, and environmental vibration and noise signals during the pile driving process through a multi-source sensor array installed on the pile top;

[0008] S2: Synchronously acquire the geological profile information at the target pile location;

[0009] S3: Perform timestamp alignment and bandpass filtering on the multi-source sensor signals;

[0010] S4: Based on the filtered pile acceleration signal, the instantaneous penetration displacement sequence of the pile is calculated by two numerical integrations combined with the zero-velocity correction algorithm;

[0011] S5: Compare the instantaneous penetration displacement sequence with the preset target depth threshold. If the absolute value of the deviation is greater than the preset tolerance range, then activate the depth error compensation mechanism.

[0012] Furthermore, the depth error compensation mechanism includes:

[0013] S501: Construct a dynamic coupling model of pile and soil with the current geological profile information as the boundary condition, the hammer impact signal as the input excitation, and the instantaneous penetration displacement sequence of the pile as the output response.

[0014] S502: The equivalent damping coefficient and side friction distribution parameters in the pile-soil dynamic coupling model are identified online using the recursive least squares method.

[0015] S503: Update the pile end bearing capacity prediction sub-model using the identified parameters to obtain the theoretical limit penetration depth that the pile can reach under the current hammering cycle.

[0016] S504: Compare the theoretical limit penetration depth with the target design depth to generate an energy adjustment command for the next hammering cycle.

[0017] Furthermore, in step S3, the bandpass filter employs a second-order Butterworth digital filter with a set passband range to retain the main modal frequencies of the pile structure response and filter out low-frequency drift and high-frequency electromagnetic noise.

[0018] Further, in step S4, the zero-velocity correction algorithm includes: detecting the acceleration signal during the static period of the hammering interval; if the acceleration signal remains within a preset range for more than a preset time, it is determined to be in a static state, and the velocity value obtained by integration during that period is forced to zero; at the same time, a linear offset correction is applied to the velocity and displacement sequence.

[0019] Furthermore, in step S501, the pile-soil dynamic coupling model adopts a lumped mass-spring-damping chain structure, the pile body is discretized into multiple equal-length units, and adjacent units are connected by nonlinear springs and viscous dampers.

[0020] Furthermore, in step S502, the recursive least squares method uses a preset forgetting factor, the observation vector is constructed based on the peak value and rise time of the hammer impact signal, and the estimation vector includes the equivalent damping coefficient aggregated by soil layer grouping.

[0021] Furthermore, in step S503, the pile end bearing capacity prediction sub-model is constructed based on the load transfer function theory. It obtains the theoretical ultimate penetration depth by integrating the pile body side friction described by the hyperbolic model along the depth direction and superimposing the pile end resistance calculated based on the effective overburden pressure and bearing capacity coefficient at the pile end.

[0022] Furthermore, in step S504, the rule for generating the energy adjustment command is as follows: based on the difference between the theoretical limit penetration depth and the target design depth, the hammering energy of the next hammering cycle is adjusted in stages; if the difference is less than or equal to zero, the pile driving is terminated.

[0023] Furthermore, after pile driving is completed, the construction quality score is calculated using a preset scoring formula based on the final penetration depth, total number of hammer blows, average penetration rate, and pile integrity test data.

[0024] As one embodiment of the present invention, the intelligent control system for pile driving depth accuracy in building engineering includes:

[0025] The multi-source sensor data acquisition module is used to acquire various sensor signals in real time during the piling process through a multi-source sensor array;

[0026] The stratigraphic information access module is used to synchronously acquire stratigraphic profile information at the target pile location;

[0027] The signal preprocessing module is used to perform timestamp alignment and bandpass filtering on the acquired sensor signals;

[0028] The penetration displacement calculation module is used to calculate the instantaneous penetration displacement sequence of the pile based on the processed acceleration signal;

[0029] The pile-soil coupling modeling module is used to build and update the dynamic coupling model of piles and soil online.

[0030] The bearing capacity prediction module is used to predict the theoretical ultimate penetration depth of the pile based on the updated model parameters.

[0031] The energy adjustment command generation module is used to generate energy adjustment commands based on the comparison between the theoretical limit penetration depth and the target depth.

[0032] The hydraulic actuator control module is used to receive and execute the energy regulation command to control the hammering energy of the piling equipment.

[0033] In summary, this application includes at least one of the following beneficial technical effects:

[0034] (1): This application uses a multi-source sensor array to collect hammer impact force, acceleration, displacement, and environmental vibration signals in real time, and combines them with prior geological data for signal fusion and high-precision displacement calculation, breaking through the limitations of traditional reliance on manual experience and single sensors. By constructing a dynamic coupling model of pile and soil and identifying soil parameters online, the system can dynamically predict the theoretical limit penetration depth under the current hammer impact, and generate hydraulic hammer energy adjustment commands in real time accordingly, forming a closed loop of "perception-modeling-prediction-control". This mechanism can effectively suppress depth deviation caused by abrupt changes in soil layers and fluctuations in hammer impact energy, achieve centimeter-level depth control, and significantly reduce over-hitting and under-hitting phenomena.

[0035] (2) Traditional methods for controlling pile driving depth are mostly based on static thresholds or fixed empirical formulas, which are difficult to adapt to complex strata such as alternating soft and hard layers and interlayers. This scheme updates the model parameters online using the recursive least squares method, enabling the system to dynamically identify changes in soil damping and side friction based on real-time hammer impact response, and possessing online learning and adaptive adjustment capabilities. Combining stratum profile information and load transfer theory, the system can automatically match the optimal impact strategy under different soil conditions, significantly improving the robustness of the pile driving process to geological dynamic changes and ensuring the consistency and reliability of pile quality under different site conditions.

[0036] (3): This solution not only achieves precise control of depth during the process, but also comprehensively evaluates multiple indicators such as final penetration depth, number of hammer blows, penetration rate, energy utilization rate, and pile integrity through a construction quality assessment module, outputting a quantitative quality score. This scoring system provides data support for construction quality traceability and process parameter optimization, and helps to form standardized construction guidance from single piles to the entire site. By reducing ineffective hammer blows and avoiding excessive blows, the system can also extend the service life of equipment and piles, improve construction efficiency, and has significant economic benefits and engineering applicability. Attached Figure Description

[0037] Figure 1 This is a flowchart illustrating the intelligent control method proposed in this invention;

[0038] Figure 2 This is a schematic diagram of the signal preprocessing and penetration displacement calculation process of the present invention;

[0039] Figure 3 This is a simplified flowchart illustrating the depth error compensation mechanism of the present invention.

[0040] Figure 4 This is a schematic diagram of the intelligent control system module composition of the present invention. Detailed Implementation

[0041] This invention provides an intelligent control method and system for pile driving depth accuracy in building engineering, aiming to solve the technical problems of insufficient pile driving depth control accuracy, frequent repeated corrections, and low construction efficiency caused by factors such as dynamic changes in geological conditions, fluctuations in hammer energy transfer efficiency, and pile response lag during existing pile driving construction. Existing technologies generally use preset penetration thresholds or empirical formulas to determine pile driving termination, lacking real-time perception and dynamic modeling capabilities of the pile-soil interaction process. This makes it difficult to effectively distinguish between instantaneous disturbances and the actual penetration state, and also makes it difficult to achieve millimeter-level depth control targets under complex geological conditions. To overcome these shortcomings, this embodiment constructs an intelligent control system with online learning and adaptive decision-making capabilities through multi-source sensor data fusion, high-precision displacement calculation, pile-soil dynamic coupling modeling, and an energy closed-loop adjustment mechanism.

[0042] Reference Figure 1-4 As shown, the intelligent control method for pile driving depth accuracy in building engineering of this application includes the following steps:

[0043] S1 collects hammer force signals, pile acceleration signals, pile top displacement signals, and environmental vibration and noise signals in real time during the pile driving process through a multi-source sensor array installed on the pile top.

[0044] S2, synchronously acquire the stratigraphic profile information at the target pile location provided by the underground geotechnical exploration database. The stratigraphic profile information includes the thickness, type, standard penetration test blow count, shear wave velocity, and static cone tip resistance parameters of each soil layer.

[0045] S3, perform time-stamp alignment and band-pass filtering on the multi-source sensing signals to eliminate high-frequency electromagnetic interference and low-frequency drift components, and retain the structural response characteristics within the effective frequency band;

[0046] S4, based on the filtered pile acceleration signal, calculates the instantaneous penetration displacement sequence of the pile through two numerical integrations and combined with the zero-velocity correction algorithm;

[0047] S5, compare the instantaneous penetration displacement sequence with the preset target depth threshold. If the absolute value of the deviation is greater than the preset tolerance range, then activate the depth error compensation mechanism.

[0048] Furthermore, each of the above steps will be further elaborated to ensure that the technical solution can be implemented in this embodiment, as follows:

[0049] In step S1, to achieve multi-dimensional and high-precision real-time perception of the piling process, the specific configuration and data acquisition process of the multi-source sensor array are as follows:

[0050] In step S1, to achieve multi-dimensional and high-precision real-time perception of the piling process, the specific configuration and data acquisition process of the multi-source sensor array are as follows:

[0051] S101: To accurately measure the hammer impact input excitation, a piezoelectric force sensor is placed at the contact interface between the pile cap and the hammer head. The sensor is integrated with the pile cap or hammer head via rigid mounting (e.g., fixed to a specially machined mounting interface using high-strength bolts) to ensure effective transmission of impact force and withstand high-frequency impact loads. Its signal output is connected to the data acquisition system via a shielded cable. The sensor's measurement range covers 0 to 5 Meganewtons, and its sampling frequency is no less than 5000 Hz, ensuring complete capture of the transient characteristics of the hammer impact pulse and providing accurate input force time history data for subsequent calculations.

[0052] S102: To achieve precise sensing of the pile's motion state, an industrial-grade triaxial MEMS accelerometer is securely fixed to the center of the pile top using a dedicated rigid mounting base (e.g., a metal clamp with anti-loosening structure or high-strength epoxy resin adhesive). During installation, using a digital level and axial alignment fixture, the accelerometer's attitude is adjusted so that its vertical measurement axis precisely coincides with the theoretical axis of the pile, while the two horizontal measurement axes remain orthogonal and perpendicular to the pile axis. The accelerometer's sampling frequency is set to no less than 2500 Hz to ensure coverage of the main structural response frequency components of the pile under hammer excitation (typically below 1 kHz), avoiding frequency aliasing. In the acquired synchronous triaxial acceleration signal, the vertical axial component is used for core penetration displacement calculation, while the two horizontal axial components are used to monitor possible deviations or torsional anomalies in the pile during pile driving in real time, providing data for assessing pile driving verticality.

[0053] S103: Configure and install the laser displacement meter. Position the meter on a stable pile driver frame, ensuring its emitted laser beam is perpendicularly pointed to a pre-installed reflective target on the pile top. With a measurement range of 0 to 2 meters and a resolution of 10 micrometers, it enables high-precision, non-contact measurement of the absolute displacement of the pile top. This displacement signal will serve as a benchmark for verifying and correcting the displacement results obtained from acceleration integration, thereby improving the overall reliability of the displacement calculation.

[0054] S104: Configure and install a three-component seismic geophone. This geophone is buried in the ground approximately 6 meters from the pile center to monitor environmental vibrations generated during pile driving. All three components (one vertical and two horizontal) are sampled at a frequency of 1000 Hz. The acquired environmental vibration signals serve two main purposes: first, to assess the impact of construction vibrations on the surrounding environment; and second, to serve as a noise reference channel, inputting into the subsequent signal preprocessing module to suppress the contamination of the pile's effective response signals (such as acceleration signals) by adaptive filtering technology.

[0055] In summary, step S1 completes the deployment and synchronous data acquisition of the multi-source sensor array. Through the coordinated operation of four types of sensors—force, acceleration, displacement, and environmental vibration—the system constructs a comprehensive and reliable in-situ sensing system from four dimensions: energy input, pile motion response, absolute position reference, and environmental interference. This provides an accurate and synchronous multi-source data foundation for subsequent high-precision displacement calculation, model identification, and closed-loop control, overcoming the shortcomings of single-sensor data being susceptible to interference and lacking sufficient information dimensions.

[0056] Step S1 enables real-time acquisition of multi-dimensional dynamic signals during the pile driving process, providing direct physical response data for the control system. To accurately analyze this dynamic data within a specific geological environment, step S2 needs to be executed simultaneously, namely, acquiring the stratigraphic profile information at the target pile location. This provides crucial prior knowledge and boundary conditions for subsequently constructing a pile-soil coupling model that conforms to actual geological conditions.

[0057] In step S2, to provide accurate geological environment reference for real-time sensing data, the following steps are performed simultaneously to obtain and apply prior geological information at the target pile location:

[0058] S201: Connect to the database and retrieve the stratigraphic record for the target pile location. The system accesses the project-specific engineering geological database server via a pre-defined communication protocol (such as HTTPS or a database-specific driver). The system sends a data request to the server based on the unique identifier of the target pile location (such as the pile number) or its geographic coordinates. The server responds to the request, returning the borehole exploration stratigraphic record for the corresponding pile location to the system. This data transmission can be achieved by calling the database's dedicated API interface and parsing the returned standardized data messages (such as JSON or XML format), or by reading standard-format exploration data files (such as those conforming to GEF or a custom CSV format) stored locally or on a network shared path. The stratigraphic record is organized and stored using a series of continuous soil layer units as the basic unit.

[0059] S202: In the stratigraphic record, each soil layer contains a set of structured mechanical and identification parameters. These parameters specifically include:

[0060] Spatial identification parameters: starting depth and ending depth, used to define the vertical spatial position of the soil layer.

[0061] Soil type parameters: Soil type codes (e.g., CL represents clay, ML represents silt, SP represents sand, etc.), used to qualitatively distinguish soil properties.

[0062] Key mechanical performance parameters include: standard penetration test blow count (N), shear wave velocity (Vs), static cone tip resistance (qc), and sidewall friction (fs). These parameters provide a direct data basis for subsequent quantitative analysis of soil resistance and stiffness.

[0063] S203: During the pile driving process, the system automatically compares the instantaneous penetration depth of the pile body calculated in real time in step S4 with the depth range of each soil layer in the stratum record. Once it is determined that the pile tip or a certain section of the pile body is located in a specific soil layer, the system automatically extracts the complete set of parameters corresponding to that soil layer from the database and uses them as the input boundary conditions for the pile-soil interaction model at the current moment.

[0064] S204: To ensure that the geological model is consistent with the construction progress, the matching and extraction of the above-mentioned stratum information is not completed all at once, but is carried out continuously with an update cycle that is the same as or shorter than the hammering cycle. This means that after each hammering or significant change in penetration depth, the system will reconfirm the current soil layer and update the model parameters, thereby ensuring that the boundary conditions are always synchronized with the actual penetration position of the pile.

[0065] Step S2, as described above, achieves the dynamic fusion of prior geological information and real-time construction progress. This involves not only retrieving static data from the database but also a closed-loop data service process that dynamically matches and outputs corresponding soil mechanical parameters based on the real-time penetration status. This process provides indispensable and time-updated geological input for the subsequent construction of an accurate pile-soil dynamic coupling model and is a key step in transforming general geological data into specific control parameters applicable to the current pile location and depth.

[0066] S3: To eliminate time differences and irrelevant noise interference between heterogeneous signals and extract clean and effective signals that can be used for accurate analysis, the system performs the following preprocessing operations on the raw signals acquired in step S1:

[0067] S301: Before each hammer strike, the system's central controller or a dedicated synchronization trigger module simultaneously sends a unified hardware trigger pulse signal (e.g., a rising edge pulse at TTL level) to the data acquisition units of all sensors (including piezoelectric force sensors, triaxial MEMS accelerometers, laser displacement gauges, and three-component seismic detectors). Upon receiving this trigger signal, each acquisition unit simultaneously starts or resets its internal timer and data acquisition process, ensuring that the data sequences acquired by all channels have a completely consistent absolute time starting reference. This mechanism guarantees that the hammer force, acceleration, displacement, and other signals are strictly aligned on the time axis during subsequent analysis.

[0068] S302: After timestamp alignment, the system uses a second-order Butterworth digital filter for bandpass filtering of the raw data from each channel. The passband frequency range of this filter is set from 5 Hz to 800 Hz. This range is preset based on the dynamic characteristics of the pile-soil system under hammer excitation: the lower limit of 5 Hz is used to filter out extremely low-frequency drift components caused by equipment temperature drift, slow foundation settlement, etc.; the upper limit of 800 Hz is used to filter out high-frequency electromagnetic noise generated by on-site electrical equipment, radio, etc. Through this processing, the main frequency band characteristics of the pile structure's dynamic response are effectively preserved.

[0069] S303: For the data from each channel after bandpass filtering in S302, the system performs the following processing to provide a suitable data format for subsequent calculations:

[0070] Hammering force signal: The filtered complete waveform data is directly output, and the impact peak value, rise time and pulse width characteristics contained in the waveform are completely preserved, which is used as the input excitation force time series F(t) of the pile-soil dynamic coupling model.

[0071] Pile acceleration signal: The arithmetic mean of the filtered acceleration data over the entire analysis time window is calculated, and this mean is subtracted from the original data sequence to eliminate the DC bias of the signal, thus obtaining the zero-mean dynamic acceleration data a(t). This processing provides physically consistent initial conditions for subsequent numerical integration operations.

[0072] Pile top displacement signal: For the filtered displacement data, a first-order high-pass filter (cutoff frequency is usually set to 0.5-1 Hz) or a linear / polynomial trend fitting and subtraction method is used to remove the low-frequency trend term (creep) introduced by factors such as loose equipment installation and slow structural deformation, so as to obtain displacement data that mainly reflects the instantaneous dynamic displacement changes caused by hammering.

[0073] Environmental vibration noise signal: The filtered three-component environmental vibration data n(t) is used as an adaptive noise reference channel. In the subsequent signal preprocessing module, this signal will be used as a reference input and input together with the pile top acceleration signal a(t) to an adaptive filter (such as a minimum mean square error LMS filter) to further suppress common-mode contamination of the effective response signal of the pile by environmental vibration noise.

[0074] Through steps S3, the system completes time synchronization and bandwidth purification of multi-source heterogeneous sensor signals. This solves the signal time difference problem caused by sensor response delays or different sampling start times, significantly improves the signal-to-noise ratio, and filters out interference components unrelated to the dynamic response of the pile structure. This provides reliable, high-quality data input for subsequent high-precision penetration displacement integral calculations and the construction of pile-soil coupling models.

[0075] For the high-quality data preprocessed in step S3 above, step S4 is executed. Its core objective is to use the processed acceleration signal a(t) and a precise integration algorithm and error correction technology to calculate the instantaneous penetration displacement sequence of the pile body in each hammering process in real time, thereby achieving accurate tracking of the dynamic changes in pile driving depth.

[0076] S4: To accurately reconstruct the pile's penetration motion from the preprocessed acceleration signal, the system performs the following steps, solving for the instantaneous penetration displacement sequence through a combination of numerical integration and error correction:

[0077] S401: Perform the first numerical integration to solve for the velocity sequence. Take the zero-mean vertical acceleration signal a(t) obtained after processing in step S3. Set the initial velocity v(t0) at time t0 to 0. Use the trapezoidal rule for numerical integration, and recursively calculate according to the sampling time interval Δt to obtain the velocity sequence v(t). The calculation formula is:

[0078]

[0079] This step converts acceleration information into velocity information.

[0080] S402: Perform the second numerical integration to solve for the original displacement sequence. Use the velocity sequence v(t) obtained in S401 as input, and set the initial displacement at time t0. Similarly, the trapezoidal rule for numerical integration is used for a second recursive integration to obtain the original displacement sequence. The calculation formula is as follows:

[0081]

[0082] Thus, through two integrations, the system converts the vertical acceleration signal a(t) into the original displacement sequence characterizing the pile motion. .

[0083] S403: Detect stationary periods and trigger zero-velocity correction. To eliminate displacement drift errors caused by sensor zero drift and integral accumulation, the system introduces a zero-velocity correction algorithm. At each sampling time, this algorithm discriminates the vertical acceleration signal a(t) processed in step S3 to find the period during which the pile is stationary in the hammering interval.

[0084] The determination of a stationary period is implemented through the following logic: The system sets a status monitoring window. When multiple consecutive sampling points of the acceleration signal a(t) satisfy -0.05m / s²≤a(t)≤+0.05m / s², and the total duration corresponding to these sampling points exceeds 20 milliseconds, the instantaneous velocity of the pile during that period is determined to be zero, and a zero-velocity correction flag is immediately triggered. This ensures that subsequent correction operations are only initiated when the pile is truly stationary, avoiding misjudgments.

[0085] S404: Execute speed zeroing and sequence correction. After S403 detects a static period and triggers the correction flag, the algorithm performs correction according to the following steps:

[0086] Zeroing velocity: Forces all corresponding velocity sequence values ​​v(t) during the stationary period to zero.

[0087] Determine the reference point and calculation error: the time before the start of the static period and the end of the last hammer blow. This serves as a reference point for displacement correction. Record the original displacement value at that moment. As a reference displacement. Because in At any given moment, the pile should be stationary and its displacement should be stable. The drift error contained within is the amount that needs to be corrected.

[0088] Perform linear offset correction:

[0089] Calculate the unit time offset: for the entire hammering cycle that needs correction (from the start of this hammering moment) until the end of the static period ), calculate the linear rate of change k of the displacement error. The calculation formula is: This formula assumes that the drift error accumulates linearly over time from the moment the hammer strike begins.

[0090] Generate and apply the correction sequence: for any time within the correction period ( ≤ ≤ ), calculate the corresponding displacement correction amount. Then, the original displacement and velocity sequences are corrected respectively:

[0091] Corrected displacement: ;

[0092] Corrected speed: (Because the derivative of the displacement correction is a constant k);

[0093] This operation ensures that the reference point is reached. Corrected displacement (It satisfies the physical constraint that the displacement is continuous and returns to zero), and eliminates the drift of the linear trend throughout the entire period.

[0094] S405: Output and Verify High-Precision Displacement Sequence. After the zero-velocity correction processing in S404, the system will generate the corrected instantaneous ingress displacement sequence. The data is output as a structured time-series data block. This data block typically contains an array of timestamps and corresponding arrays of displacement values, which are stored in real time in the system's circular buffer or sent to the central processing unit, serving as the core state variables characterizing the dynamic penetration process of the pile within the current hammering cycle.

[0095] Through the aforementioned integration and correction process, the sequence effectively suppresses low-frequency drift errors, enabling it to more realistically reflect the penetration dynamics of the pile under each hammer blow. The final value (i.e., the cumulative penetration at the end of this hammering) and the independent absolute displacement reference signal at the pile top obtained in step S3 The reliability of the solution can be verified by comparison, and its accuracy can reach the millimeter level. This high-precision displacement sequence is the direct input for depth deviation judgment and decision-making in the subsequent step S5.

[0096] In summary, step S4 of the system achieves a reliable solution from acceleration to high-precision displacement. Its core lies in the dual mechanism of "numerical integration + zero-velocity correction": integration completes the conversion of physical quantities, while zero-velocity correction utilizes the inherent intermittent static characteristics of the pile driving process to provide physical constraints and error correction benchmarks for the integration results. This provides accurate and reliable deep feedback information—i.e., instantaneous penetration displacement sequence—for the entire control system. This displacement sequence is the direct basis for subsequent depth deviation judgment and triggering of closed-loop control (step S5).

[0097] After completing step S4, the system has obtained an instantaneous penetration displacement sequence that can accurately reflect the real-time motion state of the pile. This sequence precisely depicts the dynamic penetration process of the pile body with each hammer blow, and is the core basis for determining whether the current pile driving depth meets the design requirements.

[0098] Based on this, the system proceeds to step S5. The core task of step S5 is to compare the real-time calculated penetration displacement with the preset target depth and determine whether to activate the depth error compensation mechanism based on the deviation. This step constitutes a key judgment node from "state perception" to "decision control," aiming to ensure that the final pile depth is accurately and stably controlled within the design allowable range.

[0099] In step S5, the system uses the instantaneous intrusion displacement sequence output in S4. The final value (i.e., the current cumulative penetration depth) is compared with the preset target design depth. The target design depth is manually entered before construction through the system's human-computer interface, or automatically imported through a standard data interface (such as reading CAD / BIM model files or structured databases containing pile foundation design information). The system sets a preset tolerance range (e.g., ±30 mm), which is usually determined according to the depth allowable deviation requirements of the pile foundation engineering design code.

[0100] If the absolute value of the deviation between the current penetration depth and the target depth exceeds this tolerance range, a significant deviation in depth control is determined, and a depth error compensation mechanism is triggered. The core of this mechanism lies in constructing and updating the pile-soil dynamic coupling model online to predict the theoretical limit penetration depth that the pile can achieve under the current geological and hammering conditions, and based on this, generating instructions to adjust the subsequent hammering energy.

[0101] In step S5, after determining that the depth error compensation mechanism needs to be activated, the system executes the specific steps of the depth error compensation mechanism. S501: Construct a dynamic coupling model of the pile and soil. This model uses the current geological profile information as boundary conditions, the hammer impact signal as input excitation, and the instantaneous penetration displacement sequence of the pile as the output response. S502: Use the recursive least squares method to identify the equivalent damping coefficient and side friction distribution parameters in the model online. S503: Update the pile tip bearing capacity prediction sub-model using the identified parameters. This sub-model is based on the load transfer function theory, integrating the side friction of each micro-segment of the pile along the depth direction and superimposing the pile tip resistance term to obtain the theoretical limit penetration depth that the pile can reach under the current hammer impact cycle. S504: Perform a difference calculation between the theoretical limit penetration depth and the target design depth to generate the energy adjustment command for the next hammer impact cycle.

[0102] Step S501 aims to construct a dynamic model that can reflect the real-time interaction between the pile and the soil, and is implemented as follows:

[0103] S501: To predict the penetration behavior of piles under different geological conditions, a lumped mass-spring-damping chain model is constructed to simulate the dynamic response of the pile-soil system. This model is established through the following steps:

[0104] S5011: Discretize the pile body along its axial direction into 50 equal-length elements. The length of each element is... It is one-fiftieth of the total pile length. For the i-th element, its mass... The calculation formula is determined based on the actual material and dimensions of the pile: ,in The density of the pile material. This represents the cross-sectional area of ​​the pile.

[0105] S5012: Adjacent discrete units are connected in parallel by nonlinear springs and viscous dampers to simulate the elasticity of the pile material and the damping effect of the soil on the pile vibration.

[0106] Nonlinear springs: their stiffness It is not a fixed value, but depends on the soil properties at the location of the spring. The system is based on the depth corresponding to the center point of the spring. Query the pre-generated "soil layer type-depth-secant modulus" correspondence table to obtain the secant modulus of the current soil layer. ( Then calculate the spring stiffness. This relationship table is pre-compiled based on the stratigraphic profile information and soil mechanical parameters (such as the standard penetration test blow count N and the cone tip resistance qc) obtained in step S2.

[0107] Viscous damper: its damping coefficient This is one of the core parameters to be identified by the model. To initiate the identification process, the system sets an initial value for it. The initial value is based on the type of soil layer in which the current element is located, taking 70% of its standard damping ratio ξ, and estimated using the following formula: .

[0108] S5013: The hammer impact force time history signal F(t) is applied as an external excitation to the top of the model (representing the pile top). The bottom of the model is set as a fixed or elastic constraint according to the actual situation. The motion equation of the entire discrete system can be expressed as:

[0109]

[0110] Where [M], [C], and [K] are the system mass matrix, damping matrix, and stiffness matrix, respectively, assembled from element mass, damping coefficient, and spring stiffness; {a(t)}, {v(t)}, and {x(t)} are the acceleration, velocity, and displacement vectors of each node, respectively. The system uses the explicit central difference method to solve the equation through stepwise integration in the time domain. Specifically, given the current time... displacement and speed Then, the next time step is solved recursively using the following formula. acceleration, velocity, and displacement:

[0111] Calculate the current acceleration: ;

[0112] Update speed: ;

[0113] Update displacement: ;

[0114] Update complete step size velocity: (in{ The updated displacement needs to be used in the next step. (Obtained after recalculating the force terms).

[0115] The model output is the displacement time history of the pile top node. .

[0116] In step S501, a parameterized and computable dynamic coupling numerical model of pile and soil is constructed. This model discretizes the mass and elastic distribution of the pile body and, through a nonlinear spring associated with real-time geological data and a damper to be identified online, preliminarily describes the soil's reaction to the pile's motion. The establishment of this model links geological profile information (S2), hammer excitation (S1), and the measured pile response (S4) within the same physical framework, providing an operational mathematical model foundation for subsequent online identification of soil dynamic parameters (S502). The predicted displacement obtained from the model solution... Will be compared with the measured displacement The error between the comparisons will drive the update of the model parameters.

[0117] S502: To update the soil dynamic parameters in the pile-soil dynamic coupling model in real time and ensure it tracks changes in actual working conditions, the equivalent damping coefficient of the model is identified online using the recursive least squares method. This process is performed after each hammering cycle, and the specific steps are as follows:

[0118] A forgetting factor λ = 0.98 is set for the recursive least squares method to balance the weights of historical data and new observation data, enabling the algorithm to maintain stability by utilizing historical information while responding quickly to new changes. An observation vector φ is defined, consisting of two observations: the peak value of the current hammer impact force signal F(t). and the time it takes for it to rise from the 10% peak to the 90% peak. ,Right now Define an estimation vector θ, which contains the equivalent damping coefficients corresponding to each soil layer group in the pile-soil dynamic coupling model. To reduce computational dimensionality and improve identification efficiency, the damping coefficients of multiple discrete elements within the same soil layer are merged into a single group parameter to be identified. Initialize the covariance matrix P and the parameter estimation vector θ. Typically, P is initialized as a large diagonal matrix (e.g., P0 = αI, where α is a large positive number and I is the identity matrix), and θ is initialized to the initial values ​​of the damping coefficients set in S501.

[0119] After each hammer blow, the system extracts the peak force of the blow from the preprocessed hammer force signal F(t) using a peak detection algorithm (such as finding local maxima). Simultaneously, the force signal was determined from 0.1 using a threshold comparison method. Rising to 0.9 The time interval experienced is referred to as the rise time. .Will and Combine the vectors of this observation (k represents the kth hammer blow).

[0120] The constructed pile-soil dynamic coupling model was solved under the excitation of the hammer impact force F(t) to obtain the model prediction sequence of pile top displacement. The model predicts the pile top displacement at the main end of the hammering action (e.g., when the force signal drops to 10% of its peak value). The pile top displacement at the corresponding time obtained by actual measurement and correction in step S4. Compare and calculate the model prediction error for this hammer strike. : ;

[0121] This error reflects the deviation between the current model parameters (mainly the damping coefficient) and the actual dynamic characteristics of the soil.

[0122] Using newly observed and the calculated error Update the covariance matrix P and the parameter estimation vector θ according to the standard formula of recursive least squares:

[0123] Calculate the gain matrix : ;

[0124] Update parameter estimation vector : ;

[0125] Update covariance matrix : ;

[0126] Updated This includes the latest, grouped estimates of the equivalent damping coefficients for different soil layers. The system immediately uses... Update the parameters of the corresponding damper in the pile-soil dynamic coupling model.

[0127] Through the above steps, the key dynamic parameters (equivalent damping coefficient) of pile-soil interaction are identified online and adaptively. This mechanism continuously corrects the model parameters using the input (force characteristics) and output (displacement response) data generated by each hammer blow. The introduction of the forgetting factor allows the model to gradually "forget" outdated geological information and respond quickly to newly entered soil layers (such as abrupt changes from soft soil to hard rock) or abnormal changes in pile condition (such as local damage), thereby maintaining the model's ability to represent the current actual working conditions. The updated and more accurate model parameters lay the core foundation for the next step (S503) to reliably predict the pile tip bearing capacity and theoretical ultimate penetration depth.

[0128] S503: To accurately predict the final depth the pile can reach under current geological conditions and hammering energy, the system uses the updated model parameters from step S502 to perform pile tip bearing capacity prediction. This prediction is based on load transfer function theory and is completed through the following steps:

[0129] S5031: Calculate the distribution of side skin friction in the pile. The unit side skin friction τ(z) at different depths of the pile is described using a hyperbolic model, and its expression is as follows: ,in:

[0130] This is the soil shear modulus. This value is derived from the shear wave velocity in the stratigraphic profile information obtained in step S2. The calculation formula is as follows: ρ represents the natural unit weight of the corresponding soil layer, also derived from the stratigraphic profile information. γ represents the soil unit weight, directly taken from the stratigraphic profile information. β is a dimensionless attenuation coefficient used to describe the attenuation characteristics of side friction with increasing depth; based on engineering experience, its value is usually taken as 0.02. z represents the depth of the calculation point from the pile top.

[0131] S5032: Total resistance at pile tip The calculation is performed using the Meyerhof formula, and its expression is: ,in: This is the pile tip bearing capacity coefficient. This coefficient is determined by consulting a general geotechnical bearing capacity coefficient table (such as one based on Terzaghi or Meyerhof theory) based on the soil type and effective internal friction angle φ' of the soil layer where the pile tip is located. The value of φ' can be obtained from stratigraphic profile information or associated geotechnical test data. This is the effective vertical overburden pressure at the pile tip. Its value is equal to the sum of the soil weights of all soil layers covering the pile tip plane (using natural unit weight above the groundwater level and buoyant unit weight below the groundwater level). Let be the cross-sectional area of ​​the pile tip, which is a known design constant.

[0132] S5033: Theoretical Limit Penetration Depth Defined as: when the pile penetrates to this depth, its total resistance (the sum of side friction and end resistance) is exactly equal to the current hammering energy. The maximum resistance that can be overcome after efficiency reduction. This is solved through the following iterative process:

[0133] Assume a trial depth .

[0134] Calculate total resistance: Adjust the pile length from 0 to... The depth is discretized into several micro-segments. For each micro-segment, the side friction τ(z) is calculated using the formula S5031 based on its center depth z and the soil layer it is located in. This value is then multiplied by the side surface area of ​​the micro-segment to obtain its side friction contribution. The side friction contributions of all micro-segments are summed to obtain the total side friction. In addition, the pile end resistance corresponding to depth (Calculated according to S5032), the total resistance is obtained. .

[0135] Comparison and judgment: Compare the total resistance Q_total(d_trial) with the equivalent failure load R_u corresponding to the current hammer energy (R_u can be estimated by the energy method using E_hammer and the pile penetration-resistance relationship model).

[0136] Iterative correction: If This indicates that the pile can continue to penetrate, thus increasing the risk. Then return to step 2; if If , then decrease d_trial. Repeat this process until... and The difference satisfies the preset convergence tolerance (e.g., the difference is less than 1%). This is the predicted theoretical limit of penetration depth. .

[0137] In summary, step S503 completes the conversion from microscopic soil parameters to macroscopic bearing capacity prediction. This process deeply integrates real-time updated soil layer information (S2), online identified soil damping parameters (S502), and the current hammer impact energy state. Through rigorous geotechnical mechanics models and numerical iterations, it achieves a quantitative and refined prediction of future single-hammer penetration potential. It is a key control variable that will serve as a benchmark and be compared with the target design depth in the next step (S504) to generate precise energy regulation commands, thus forming the decision core of the entire intelligent control closed loop.

[0138] S504: The theoretical limit penetration depth predicted based on step S503 With respect to the preset target design depth Calculate the remaining depth Based on the value of Δd, the energy command for adjusting the next hammer cycle is generated according to the following rules:

[0139] If Δd > 0.05 meters, it indicates that the target depth is still far away, and the current hydraulic system hammer energy control parameters (such as system working pressure and main pump displacement) should be kept unchanged.

[0140] If 0.01 m < Δd ≤ 0.05 m, it indicates that the target depth has been approached. To control the penetration speed and prevent over-penetration, the hammering energy is reduced to 80% of the current reference value. This instruction is achieved by proportionally reducing key parameters controlling the hydraulic hammer (such as the set current of the proportional relief valve).

[0141] If 0 < Δd ≤ 0.01 meters, it indicates that the design depth is about to be reached. In order to carry out the final precise penetration and prepare for termination, the hammering energy will be significantly reduced to 50% of the current benchmark value.

[0142] If Δd≤0, it indicates that the predicted limit depth has reached or exceeded the design depth, and a termination piling command is immediately generated, triggering a depth over-limit alarm signal.

[0143] The thresholds (0.05 meters, 0.01 meters) in the above rules are designed to achieve a graded, smooth approach control process. The generated energy regulation commands (including maintain, proportional down, or terminate) are encapsulated as standard digital control signals (such as specific messages or setpoints) and sent in real time to the hydraulic actuator control module to drive it to adjust the corresponding valves and pump control units.

[0144] On the other hand, the intelligent control system for piling depth accuracy in building engineering disclosed in this application includes the following core functional modules and an auxiliary evaluation module in order to realize the aforementioned intelligent control system method for piling depth accuracy in building engineering:

[0145] Core processing and control module:

[0146] Multi-source sensor data acquisition module: Connects piezoelectric force sensor, triaxial MEMS accelerometer, laser displacement meter and three-component seismic detector via shielded twisted pair cable. Built-in high-precision analog-to-digital converter and anti-aliasing filter, responsible for high-fidelity synchronous acquisition and transmission of raw signals.

[0147] Stratigraphic Information Access Module: Communicates with the engineering geological database via industrial Ethernet, and has data parsing (such as JSON format), caching and disconnection reconnection functions. It is responsible for acquiring and accessing the stratigraphic profile information of the target pile location in real time.

[0148] Signal preprocessing module: Deployed on an embedded processor, it is responsible for performing preprocessing algorithms such as timestamp alignment, bandpass filtering, and adaptive noise cancellation on the acquired raw signals.

[0149] Penetration displacement calculation module: Based on the preprocessed acceleration signal, it runs a numerical integration algorithm including zero velocity correction to calculate and output a high-precision instantaneous penetration displacement sequence of the pile in real time.

[0150] The pile-soil coupling modeling module is responsible for building and updating the dynamic coupling model of piles and soil online based on real-time data (such as through recursive least squares method).

[0151] Bearing capacity prediction module: Based on the updated model parameters, predicts the theoretical ultimate penetration depth of the pile under the current hammering cycle according to the load transfer function theory.

[0152] Energy adjustment command generation module: Based on the comparison between the theoretical limit penetration depth and the target depth, it generates corresponding energy adjustment or termination commands according to preset grading rules.

[0153] Hydraulic actuator control module: Receives energy regulation commands and precisely controls the striking energy of the hydraulic hammer and the start and stop of operation through components such as proportional relief valve (controlling the current and system pressure linearly), variable pump controller (adjusting pump displacement) and stroke counter.

[0154] Construction quality assessment module (auxiliary module):

[0155] This module runs automatically after single pile driving is completed. By integrating and analyzing multi-dimensional data such as final penetration depth, total number of hammer blows, average penetration, energy utilization rate, and pile integrity index, it calculates a quantitative score according to a preset weighted scoring formula, providing a basis for construction quality traceability and subsequent process parameter optimization.

[0156] This embodiment achieves high-precision closed-loop control of pile driving depth through the aforementioned method and system. Multi-source sensor data fusion ensures the reliability of penetration state perception, the pile-soil dynamic coupling model provides physically interpretable predictive capabilities, and the energy regulation mechanism enables adaptive optimization of hammering parameters. The entire process requires no manual intervention, significantly improving construction efficiency and quality consistency.

[0157] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, the embodiments should be regarded as exemplary and non-limiting in all respects.

[0158] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for intelligent control of pile driving depth accuracy in building engineering, characterized in that, Includes the following steps: S1: Real-time acquisition of hammer force signals, pile acceleration signals, pile top displacement signals, and environmental vibration and noise signals during the pile driving process through a multi-source sensor array installed on the pile top; S2: Synchronously acquire the geological profile information at the target pile location; S3: Perform timestamp alignment and bandpass filtering on the multi-source sensor signals; S4: Based on the filtered pile acceleration signal, the instantaneous penetration displacement sequence of the pile is calculated by two numerical integrations combined with the zero-velocity correction algorithm; S5: Compare the instantaneous penetration displacement sequence with the preset target depth threshold. If the absolute value of the deviation is greater than the preset tolerance range, then activate the depth error compensation mechanism.

2. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, in, The depth error compensation mechanism includes: S501: Construct a dynamic coupling model of pile and soil with the current geological profile information as the boundary condition, the hammer impact signal as the input excitation, and the instantaneous penetration displacement sequence of the pile as the output response. S502: The equivalent damping coefficient and side friction distribution parameters in the pile-soil dynamic coupling model are identified online using the recursive least squares method. S503: Update the pile end bearing capacity prediction sub-model using the identified parameters to obtain the theoretical limit penetration depth that the pile can reach under the current hammering cycle. S504: Compare the theoretical limit penetration depth with the target design depth to generate an energy adjustment command for the next hammering cycle.

3. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, In step S3, the bandpass filter uses a second-order Butterworth digital filter with a set passband range to retain the main modal frequencies of the pile structure response and filter out low-frequency drift and high-frequency electromagnetic noise.

4. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, In step S4, the zero-velocity correction algorithm includes: detecting the acceleration signal during the static period of the hammering interval; if the acceleration signal remains within a preset range for more than a preset time, it is determined to be in a static state, and the velocity value obtained by integration during that period is forced to zero; at the same time, a linear offset correction is applied to the velocity and displacement sequence.

5. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, In step S501, the pile-soil dynamic coupling model adopts a lumped mass-spring-damping chain structure. The pile body is discretized into multiple equal-length units, and adjacent units are connected by nonlinear springs and viscous dampers.

6. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, In step S502, the recursive least squares method uses a preset forgetting factor, the observation vector is constructed based on the peak value and rise time of the hammer impact signal, and the estimation vector includes the equivalent damping coefficient aggregated by soil layer grouping.

7. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, In step S503, the pile end bearing capacity prediction sub-model is constructed based on the load transfer function theory. The theoretical ultimate penetration depth is obtained by integrating the pile body side friction described by the hyperbolic model along the depth direction and superimposing the pile end resistance calculated based on the effective overburden pressure and bearing capacity coefficient at the pile end.

8. The intelligent control method for pile driving depth accuracy in building engineering according to claim 1, characterized in that, In step S504, the rule for generating the energy adjustment command is as follows: based on the difference between the theoretical limit penetration depth and the target design depth, the hammering energy of the next hammering cycle is adjusted in stages; if the difference is less than or equal to zero, the pile driving is terminated.

9. The intelligent control method for pile driving depth accuracy in building engineering according to any one of claims 1 to 8, characterized in that, After pile driving is completed, the construction quality score is calculated based on the final penetration depth, total number of hammer blows, average penetration degree, and pile integrity test data using a preset scoring formula.

10. An intelligent control system for pile driving depth accuracy in building engineering, used to implement the method described in any one of claims 1 to 9, characterized in that, include: The multi-source sensor data acquisition module is used to acquire various sensor signals in real time during the piling process through a multi-source sensor array; The stratigraphic information access module is used to synchronously acquire stratigraphic profile information at the target pile location; The signal preprocessing module is used to perform timestamp alignment and bandpass filtering on the acquired sensor signals; The penetration displacement calculation module is used to calculate the instantaneous penetration displacement sequence of the pile based on the processed acceleration signal; The pile-soil coupling modeling module is used to build and update the dynamic coupling model of piles and soil online. The bearing capacity prediction module is used to predict the theoretical ultimate penetration depth of the pile based on the updated model parameters. The energy adjustment command generation module is used to generate energy adjustment commands based on the comparison between the theoretical limit penetration depth and the target depth. The hydraulic actuator control module is used to receive and execute the energy regulation command to control the hammering energy of the piling equipment.

Images