Static pressure pile driver pile sinking posture dynamic compensation method and system

Abstract

The present application relates to the field of static pile sinking compensation, more particularly, the present application relates to a static pile sinking posture dynamic compensation method and system. The method comprises: obtaining a standardized data sequence; constructing a yaw inclination recursive index; constructing a soil-water interaction compensation coefficient; constructing a torsion decoupling deflection correction factor; constructing a hydraulic servo flow feedforward distribution matrix; generating a leveling oil cylinder control instruction based on the hydraulic servo flow feedforward distribution matrix, and executing dynamic compensation combined with the determination result of the yaw inclination recursive index by the state machine. Through the technical scheme of the present application, the dynamic response bandwidth and action real-time performance of the system can be improved, so that the pile machine can exhibit stronger anti-interference toughness and rapid recovery ability when facing various sudden mechanical disturbances in the penetration process.

Description

Technical Field

[0001] This invention relates to the field of pile driving compensation for static pressure pile drivers. More specifically, this invention relates to a method and system for dynamic compensation of the pile driving posture of a static pressure pile driver. Background Technology

[0002] During static pressure pile driving, real-time control of the pile's attitude primarily relies on a closed-loop feedback system consisting of tilt sensors and leveling cylinders mounted on the pile frame. A common control strategy is proportional-integral-derivative (PID) regulation based on tilt deviation. When a tilt is detected along the X-axis or Y-axis, the controller calculates the correction amount based on the instantaneous angle error and drives the corresponding leveling cylinder to extend or retract to counteract the unidirectional tilt. This method is simple to implement and has a certain correction capability for slow tilting in stable strata.

[0003] However, in construction practice, it has been found that the control architecture relying solely on instantaneous tilt angle feedback has several limitations. First, this architecture does not extract and model the yaw trend associated with continuous settlement during penetration. When the pile tip encounters a local hard layer or a soft-hard interface, the tilt angle deviation often accumulates to a significant degree before triggering correction, resulting in a lag in adjustment and a tendency to overshoot and oscillation. Second, dynamic changes in pore water pressure cause changes in the effective stress of the soil around the pile and lead to fluctuations in penetration resistance. Existing feedback control lacks direct quantification and compensation for this soil-hydraulic disturbance, making it difficult for control parameters to adapt to geological conditions with drastic changes in pore pressure, thus reducing the robustness of attitude maintenance. Furthermore, when the pile body simultaneously exhibits tilt and torsion in orthogonal directions, the tilt angle changes in the X and Y axes are coupled with torsional components. If each degree of freedom cylinder operates independently based on the tilt angle error of its corresponding axis, repeated pulling will occur due to the cross-effects caused by torsion, resulting in coupling interference in multi-axis correction, prolonging the convergence time and reducing control accuracy. These problems are exacerbated, especially when there are frequent changes in the strata or when ultra-long piles are driven into the ground, significantly affecting the quality of pile formation and construction efficiency. Summary of the Invention

[0004] To address the aforementioned problems, the present invention provides solutions in several aspects.

[0005] In a first aspect, this invention discloses a dynamic compensation method for the driving posture of a static pressure pile driver, comprising: collecting multi-source sensor data during the pile driving process and preprocessing it to obtain a standardized data sequence, the standardized data sequence including longitudinal settlement increment, longitudinal filtered tilt angle data, pore water pressure spatial gradient, projected penetration rate, lateral filtered tilt angle data, and lateral displacement increment; constructing a yaw angle recursive index based on the longitudinal settlement increment and longitudinal filtered tilt angle data; constructing a soil-water interaction compensation coefficient based on the pore water pressure spatial gradient and projected penetration rate; constructing a torsional decoupling skew correction factor based on the lateral filtered tilt angle data, longitudinal filtered tilt angle data, longitudinal settlement increment, and lateral displacement increment; constructing a hydraulic servo flow feedforward allocation matrix according to the yaw angle recursive index, the soil-water interaction compensation coefficient, and the torsional decoupling skew correction factor; generating a leveling cylinder control command based on the hydraulic servo flow feedforward allocation matrix, and performing dynamic compensation based on the determination result of the yaw angle recursive index using a state machine.

[0006] Preferably, the acquisition and preprocessing of multi-source sensor data during pile driving includes: acquiring the original tilt angle sequence of the pile body in two orthogonal directions, the change in vertical distance at the pile top, the offset of the horizontal light spot at the pile top, the original pore fluid pressure values ​​at different soil depths, and the advance distance of the pile driving cylinder; filtering the original tilt angle sequence and displacement data to obtain a filtered sequence; using the control cycle as the time reference, periodically truncating and arithmetically averaging the filtered sequence, the original pore fluid pressure values, and the advance distance of the pile driving cylinder to obtain the average attitude value, average displacement value, and pore pressure value for any control cycle. The average pressure gradient and average propulsion velocity are calculated. The average attitude values ​​include lateral and longitudinal attitude values, and the average displacement values ​​include lateral and longitudinal displacement values. The lateral attitude values ​​are normalized to obtain lateral filtered dip angle data, the longitudinal attitude values ​​are normalized to obtain longitudinal filtered dip angle data, the longitudinal displacement values ​​are normalized to obtain longitudinal settlement increments, the lateral displacement values ​​are normalized to obtain lateral displacement increments, the pore pressure gradient values ​​are normalized to obtain the pore water pressure spatial gradient, and the average propulsion velocity is normalized to obtain the projected penetration rate.

[0007] Preferably, the construction of the yaw angle recursive index includes: taking any control cycle as the target cycle and the preceding control cycle adjacent to the target cycle as the control cycle; calculating the product of the longitudinal settlement increment and the longitudinal filtered tilt data of the target cycle to obtain the horizontal offset trend; calculating the ratio of the change in the longitudinal filtered tilt data of the target cycle and the control cycle to obtain the tilt change rate; weighting and summing the horizontal offset trend and the tilt change rate, and performing time-series accumulation within a preset time window to obtain the yaw angle recursive index.

[0008] Preferably, the construction of the soil-water interaction compensation coefficient includes: calculating the instantaneous disturbance intensity by multiplying the spatial gradient of pore water pressure and the projected penetration rate; obtaining a preset dynamic viscosity coefficient of the soil layer and a pressure dissipation characteristic constant; calculating the product of the pressure dissipation characteristic constant and the cumulative pile driving time as a first product; performing a negative correlation mapping on the first product using an exponential function to obtain a mapping result; calculating the difference between constant 1 and the mapping result; calculating the sum of the difference and a preset zero-prevention constant as a first sum; multiplying the first sum and the preset dynamic viscosity coefficient of the soil layer as a second product; and using the ratio of the instantaneous disturbance intensity term to the second product as the soil-water interaction compensation coefficient.

[0009] Preferably, the construction of the torsional decoupling skew correction factor includes: calculating the product of the lateral filtered tilt angle data and the longitudinal settlement increment as a third product, and calculating the product of the longitudinal filtered tilt angle data and the lateral displacement increment as a fourth product; calculating the difference between the third product and the fourth product to obtain a cross-cancellation term; performing lower limit protection processing on the projection penetration rate to obtain a lower limit clamping projection penetration rate; calculating the sum of the square of the lateral filtered tilt angle data, the square of the longitudinal filtered tilt angle data, and a preset zero-prevention constant as a second sum, calculating the square root of the second sum, and multiplying the square root by the lower limit clamping projection penetration rate as a smooth scaling term; and using the ratio of the cross-cancellation term to the smooth scaling term as the torsional decoupling skew correction factor.

[0010] Preferably, the construction of the hydraulic servo flow feedforward allocation matrix includes: multiplying the calculated two-dimensional identity matrix with the soil-water interaction compensation coefficient as the global reference flow term; constructing a two-dimensional diagonal matrix with the torsional decoupling skew correction factor and its negative values ​​as the main diagonal elements, and multiplying the two-dimensional diagonal matrix with the yaw angle recursive index to obtain the skew differential flow term; and superimposing the global reference flow term and the skew differential flow term to obtain the hydraulic servo flow feedforward allocation matrix.

[0011] Preferably, the dynamic compensation of the yaw angle recursive index determination result by combining the state machine includes: setting a preset stability threshold and a preset warning threshold; when the absolute value of the yaw angle recursive index is less than the preset stability threshold for a first preset number of control cycles, it is determined to be a stable penetration state, and the synchronous oil supply of the dual-side leveling cylinders is maintained; when the absolute value of the yaw angle recursive index is not less than the preset stability threshold and not greater than the preset warning threshold for a first preset number of control cycles, it is determined to be a warning state, the current benchmark is locked and the cylinder preheating is started; when the absolute value of the yaw angle recursive index is greater than the preset warning threshold for a second preset number of control cycles, it is determined to be a high-risk state, and the main pump frequency reduction and forced reset command are executed; under different determination states, the hydraulic servo flow feedforward allocation matrix is ​​called, and according to the positive and negative symmetry structure of the main diagonal elements of the matrix, the deviation polarity of the yaw angle recursive index is converted into the differential flow ratio command of the dual-side leveling cylinders.

[0012] Secondly, the present invention discloses a dynamic compensation system for the driving posture of a static pressure pile driver, comprising: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a dynamic compensation method for the driving posture of a static pressure pile driver as described in any one of the present invention is implemented.

[0013] The beneficial effects of this invention are: 1. By recursively fusing settlement increment and tilt angle data, a yaw tilt angle recursive index is generated that can characterize the yaw evolution trend of the pile. Compared with traditional feedback control that only responds to instantaneous tilt angle deviation, this index provides predictive attitude evolution information for the control system, enabling the hydraulic leveling system to intervene and correct the yaw trend in advance before it expands. This eliminates overshoot and oscillation caused by feedback lag at the root, improving the dynamic stability and vertical accuracy of the pile attitude during pile driving.

[0014] 2. A soil-water interaction compensation coefficient, constructed based on pore water pressure gradient and penetration rate, enables real-time quantification of changes in the mechanical state of the soil surrounding the pile. This coefficient can dynamically correct for changes in penetration resistance caused by fluctuations in pore water pressure, injecting environmental mechanical disturbances into the control model in a feedforward manner. This gives the leveling strategy the ability to adapt to geological environments with changes in pore pressure, effectively overcoming the problems of sudden attitude deviation and slow convergence caused by alternating strata, and ensuring the construction quality and efficiency of continuous long pile driving.

[0015] 3. By utilizing the torsional decoupling correction factor generated through decoupling calculations of orthogonal inclination angles and displacement increments, precise separation of the torsional and skew motion components of the pile body is achieved. This factor provides an independent correction basis for each degree of freedom, blocking the mechanical coupling and mutual interference paths between multi-axis leveling actions from a control mechanism perspective. It avoids actuator oscillations caused by mutual tension during simultaneous multi-directional correction, thereby significantly accelerating the overall convergence speed of the pile body attitude and improving the control accuracy under complex correction conditions.

[0016] 4. By integrating the above three innovative indicators to construct a hydraulic servo flow feedforward allocation matrix, an end-to-end feedforward control architecture is established that directly maps multi-source sensor features to flow commands for each leveling cylinder. This matrix, combined with state machine decision-making, efficiently transforms multi-dimensional sensing information into physical actions of the actuator, shortening the control link of "perception-decision-execution," improving the system's dynamic response bandwidth and real-time performance, and enabling the piling machine to exhibit stronger anti-interference resilience and rapid recovery capability when facing various sudden mechanical disturbances during the penetration process. Attached Figure Description

[0017] Figure 1 This is a flowchart of a dynamic compensation method for the driving posture of a static pressure pile driver according to an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0019] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0020] Reference Figure 1 A dynamic compensation method for the driving posture of a static pressure pile driver includes steps S1-S6, which are described in detail below.

[0021] S1: Collect multi-source sensor data during the pile driving process and preprocess it to obtain a standardized data sequence. The standardized data sequence includes longitudinal settlement increment, longitudinal filtered tilt angle data, pore water pressure spatial gradient, projected penetration rate, lateral filtered tilt angle data, and lateral displacement increment.

[0022] In one embodiment, during the static pressure pile driving operation, multi-source sensor data reflecting the pile-soil interaction are acquired in real time through a multi-dimensional sensor network. Specifically, bidirectional attitude sensor nodes arranged at the pile top collect the original tilt angle sequence of the pile body in two orthogonal directions at high frequency to characterize the initial alignment accuracy and the micro-oscillation and torsional trends caused by lateral compression of the stratum; a planar laser ranging module mounted on the main beam, in conjunction with a high-reflectivity target, synchronously records the change in vertical distance at the pile top and the horizontal spot offset, thereby tracking the cumulative pile driving stroke and the lateral drift trajectory of the axis; a multi-hole monitoring probe array uniformly distributed around the pile reads the original pore fluid pressure at different soil depths at regular intervals to reflect the generation and dissipation process of the penetration-induced excess static pressure; simultaneously, a hydraulic stroke encoder collects the advance distance of the pile driving cylinder in real time and differentially derives the advance speed, mapping the main pump power output and the mechanical response characteristics of the soft-hard interface of the stratum. These multi-source sensor data together constitute the underlying information foundation for dynamic compensation of pile driving attitude.

[0023] To address the instantaneous signal jumps introduced by construction machinery vibration and electromagnetic coupling, a moving average filtering algorithm is first used to denoise the original tilt angle and displacement data, eliminating spurious extreme values ​​caused by occasional sensor jitter or target reflection interference, and retaining effective temporal components that conform to the continuity of physical motion. Subsequently, to eliminate time misalignment caused by differences in sampling frequencies and response delays among various sensor nodes, multi-source time series aggregation is performed using a fixed control period as a unified time reference. By periodically extracting and arithmetically averaging the filtered tilt angle and displacement sequences acquired at high frequencies, the average lateral and longitudinal attitude and displacement values ​​within that control period are obtained. Simultaneously, the original pore fluid pressure acquired at low frequencies is extracted within the same time window, the spatial pressure gradient of adjacent monitoring holes is calculated, and averaged to obtain the average pore pressure gradient. Combined with the average propulsion speed within the period, the average propulsion speed is finally obtained. This aggregation mechanism ensures that all input data within the same decision step have a strictly consistent time reference.

[0024] To eliminate the differences in magnitude and physical units between multi-source sensor data, a physical reference standard is introduced for dimensionless transformation, achieving dimensional unification and normalization mapping. Specifically, the mean values ​​of lateral and longitudinal attitudes are divided by the maximum allowable dip angle reference value to obtain lateral and longitudinal filtered dip angle data, respectively. The difference in the mean longitudinal displacement between adjacent control cycles is divided by the standard length reference value of a single pile section to obtain the longitudinal settlement increment; the difference in the mean lateral displacement between adjacent control cycles is divided by the maximum allowable lateral offset reference value to obtain the lateral displacement increment; the mean pore pressure gradient is divided by the standard pore pressure gradient reference value of the formation to obtain the pore water pressure spatial gradient; and the average advance velocity is divided by the design penetration rate reference value to obtain the projected penetration rate. After this rigorous preprocessing, the final output is a standardized data sequence containing the longitudinal settlement increment, longitudinal filtered dip angle data, pore water pressure spatial gradient, projected penetration rate, lateral filtered dip angle data, and lateral displacement increment.

[0025] S2: Based on longitudinal settlement increment and longitudinal filtered tilt angle data, construct a yaw tilt angle recursive index.

[0026] In one embodiment, during pile driving in strata where soft and hard soils intersect, as the pile penetrates vertically, a slow-changing yaw tilt occurs due to factors such as stratum heterogeneity and variations in pile end resistance distribution. This yaw evolution is not only reflected in the instantaneous tilt angle value but also hidden in the continuous settlement accumulation process. Relying solely on instantaneous tilt angle data for judgment fails to capture the progressive development of the yaw trend under the coupled effects of settlement and tilt angle, leading to a lack of forward-looking assessment of yaw risk. Therefore, this embodiment innovatively constructs a yaw tilt angle recursive index, combining single-step penetration action with attitude change depth to provide quantitative information reflecting the yaw evolution trend for attitude feedforward compensation.

[0027] In the specific calculation, the longitudinal settlement increment and longitudinal filtered dip angle data of a specific control period are first selected as the basic inputs to construct a product term. The engineering intention is to directly calculate the horizontal offset trend of a single period. The positive or negative sign of this product represents the direction of deviation, and the magnitude of the value reflects the degree of deviation. When the pile is basically vertical, the dip angle value approaches zero, and the product term automatically returns to zero, effectively avoiding malfunctions in response to minor vibrations. When the dip angle persists, the product terms remain superimposed in the same direction, truly reflecting the deviation direction of the pile trajectory. Simultaneously, a differential weighting coefficient is introduced, multiplied by the ratio of the change in longitudinal filtered dip angle data of adjacent control periods to the control period, to sensitively capture the instantaneous impact of sudden changes in the strata on the pile. When the pile driving process is stable, this term value is close to zero, and the indicator is mainly determined by the horizontal offset trend; when a sudden change in the strata causes a rapid change in dip angle, this term value increases significantly, enabling the system to quickly identify sudden deviations.

[0028] It should be explained that the differential weighting coefficient is an empirical parameter set to adjust the contribution of the inclination rate of change to the recursive index. When the pile encounters a hard interlayer or local void, the inclination angle will change abruptly in a very short time. This coefficient is used to control the weight ratio of the discrete differential term in the total index. If the differential weighting coefficient is set too small, the inclination acceleration signal caused by the sudden change in strata will be excessively attenuated, and the system will not be able to capture the early instability characteristics in time, resulting in a lag in the correction command. If it is set too large, it will amplify the high-frequency noise of the sensor and the occasional mechanical vibration, causing the servo valve to oscillate frequently, thereby accelerating the wear of the hydraulic seals. The differential weighting coefficient is calibrated by combining the step response test of the hydraulic system on site and the noise spectrum analysis of the inclination sensor. Its value range is set to 0.5s to 2s. In this embodiment, it is set to 1.2s to achieve an engineering balance between dynamic disturbance capture and noise suppression. The specific value can also be adaptively adjusted by the implementers according to the sampling frequency and the frequency response characteristics of the hydraulic valve.

[0029] The above steps employ a sliding window accumulation structure, superimposing data from multiple consecutive periods within a preset time window. This mechanism follows the continuous variation of soil stress, effectively filtering out instantaneous interference from single mechanical vibrations or sensor noise, ensuring the final result accurately reflects the macroscopic development trend of the tilt. The resulting yaw angle recursive index comprehensively reflects the intensity and direction of the coupling effect between penetration settlement and attitude deflection over time. When the yaw angle recursive index approaches zero, it indicates that the pile is in a vertically stable penetration state, with positive and negative tilt angle disturbances canceling each other out and no continuous tilt accumulation. Subsequent steps maintain routine monitoring and rolling updates of benchmark data based on the yaw angle recursive index to prevent the hydraulic system from making ineffective adjustments due to oversensitivity. When the yaw angle recursive index is significantly greater than zero, it indicates that the pile body has experienced a continuous axial shift in the positive direction, and the cumulative drift trend has exceeded the engineering safety boundary. Subsequent steps directly extract the positive deviation polarity based on the yaw angle recursive index, triggering pressure reduction commands on the positive side leveling cylinder and pressure increase commands on the reverse side, forcing the pile body trajectory to return to positive. When the yaw angle recursive index is significantly less than zero, it indicates that the pile body has experienced a continuous axial shift in the negative direction, and the cumulative drift trend has exceeded the safety boundary in the reverse direction. Subsequent steps directly extract the negative deviation polarity based on the yaw angle recursive index, triggering pressure reduction commands on the reverse side leveling cylinder and pressure increase commands on the positive side, forcing the pile body trajectory to return to positive, thus providing a clear and directional calculation benchmark for subsequent hydraulic servo flow distribution.

[0030] It should be noted that the preset time window length is an empirical parameter set to determine the time span for accumulating historical deviation data. The evolution of pile driving attitude has temporal inertia, requiring multi-cycle accumulation to eliminate transient interference and construct a trend baseline. If the preset time window length is set too small, historical deviation information will be truncated too frequently, causing the indicator to degenerate into instantaneous value judgments, weakening trend prediction capabilities and easily triggering hydraulic overshoot. If it is set too large, old stratum state data will remain in the calculation window, causing indicator response lag and failing to reflect the true resistance distribution of the current soft-hard interface in a timely manner. The preset time window length is calibrated based on the physical time corresponding to the typical stratum layer thickness and average penetration rate, and its value range is set to 10-50 control cycles. In this embodiment, it is set to 20 control cycles to ensure a balance between deviation trend smoothness and real-time control.

[0031] S3: Based on the spatial gradient of pore water pressure and the projected penetration rate, a soil-water interaction compensation coefficient is constructed.

[0032] In one embodiment, the instantaneous disturbance intensity is obtained by multiplying the spatial gradient of pore water pressure and the projected penetration rate; a preset dynamic viscosity coefficient of the soil layer and a pressure dissipation characteristic constant are obtained; the product of the pressure dissipation characteristic constant and the cumulative pile driving time is calculated as the first product; the first product is negatively correlated using an exponential function to obtain the mapping result; the difference between constant 1 and the mapping result is calculated; the sum of the difference and the preset zero-prevention constant is calculated as the first sum; the product of the first sum and the preset dynamic viscosity coefficient of the soil layer is used as the second product; the ratio of the instantaneous disturbance intensity term to the second product is used as the soil-water interaction compensation coefficient.

[0033] Among them, the soil-water interaction compensation coefficient is used to characterize the relative proportion of the feedforward compensation flow required by the hydraulic system. The pore water pressure spatial gradient is used to characterize the proportion of the driving force of water pressure changes within the soil layer relative to the standard pressure threshold of the stratum. The projected penetration rate is used to characterize the scaling factor of the pile driving cylinder's advance speed after scaling down to the design penetration rate reference value. The soil dynamic viscosity coefficient is used to characterize the proportion of the flow resistance of saturated clay relative to the standard viscosity. The pressure dissipation characteristic constant is used to characterize the rate of decay of excess pore water pressure. The current cumulative pile driving time is used to characterize the duration of the penetration process. The preset zero-prevention constant is used to prevent the denominator calculation result from being zero. For example, the pressure dissipation characteristic constant is set to 0.2; the specific value can be set by those skilled in the art.

[0034] It should be explained that the dynamic viscosity coefficient of soil is a property parameter set to characterize the proportion of the flow resistance of saturated clay to the standard viscosity. If the dynamic viscosity coefficient of soil is set too small, the denominator will be too small overall, the calculated value of the compensation coefficient will be too large, the system output flow will be continuously too high, causing the hydraulic system to overload and the cylinder to move excessively, which will lead to pipeline pressure fluctuations and accelerated wear of seals. If it is set too large, the denominator will be too large overall, the calculated value of the compensation coefficient will be too small, the system output flow will be continuously too low, and it will be unable to effectively offset the attitude drift caused by pore pressure disturbance, which will lead to the accumulation of pile body deviation. The dynamic viscosity coefficient of soil is calibrated based on the ratio of the dynamic viscosity of soil measured by field geotechnical tests to the reference value of the dynamic viscosity of pure water at 20 degrees Celsius. Its value range is set to 0.5 to 2.0. In this embodiment, it is set to 1.0 to ensure that the flow compensation intensity matches the actual flow resistance of the soil. The specific value can also be adaptively adjusted by the implementers according to the soil moisture content, mineral composition and field rheological test data.

[0035] The core logic of constructing the soil-water interaction compensation coefficient lies in establishing a deep engineering correlation between the groundwater compression state and the changes in soil resistance. In constructing the numerator, the spatial gradient of pore water pressure is multiplied by the projected penetration rate. The physical meaning of this product is to characterize the instantaneous disturbance intensity of the penetration action on the pore water in the soil layer. Specifically, the larger the water pressure gradient and the faster the penetration rate, the more water is squeezed out per unit time, and the more intense the fluctuation of the soil layer's reaction force on the pile. In this case, the hydraulic system needs a higher compensation flow rate to maintain the stability of the pile's posture. In constructing the denominator, the sum of the soil layer's dynamic viscosity coefficient, the pressure dissipation dynamic evolution term, and the preset zero-constant is multiplied to simulate the process of the soil layer gradually recovering its bearing capacity after being compressed. Specifically, the soil layer's dynamic viscosity coefficient represents the soil layer's own flow resistance; the pressure dissipation dynamic evolution term is calculated by subtracting the result of a natural exponential function with the negative exponent of the product of the pressure dissipation characteristic constant and the current cumulative pile driving time from a constant of 1. This evolution term reflects the law of natural dissipation of pore water pressure over time. As the cumulative time of pile driving increases, the value of this evolution term gradually increases, resulting in a larger overall value of the denominator, which in turn causes the soil-water interaction compensation coefficient to gradually fall back to the baseline level from its initial high value.

[0036] The engineering intent behind employing a numerator-to-denominator ratio structure is to ensure that flow regulation strictly corresponds to the actual soil condition. In the initial stage of pile driving, pore water pressure is high and dissipates slowly, resulting in a large output of the soil-water interaction compensation coefficient. The system uses this to provide sufficient compensation flow to resist disturbances. As pore water pressure gradually dissipates and the soil stabilizes, the soil-water interaction compensation coefficient automatically decays, and the hydraulic flow returns to its normal distribution. Furthermore, the introduction of a preset zero-prevention constant ensures that the denominator is not zero during the first control cycle after power-on, thus avoiding calculation overflow errors. This overall calculation structure ensures a smooth transition of the compensation flow with the progress of soil consolidation, fundamentally reducing the risk of lateral deviation caused by pore pressure accumulation. The soil-water interaction compensation coefficient, as a global reference variable, directly participates in subsequent matrix superposition calculations, providing the hydraulic system with bottom-level flow support to resist ground disturbances.

[0037] The specific value of the soil-water interaction compensation coefficient directly reflects the relative proportion of the feedforward compensation flow required by the hydraulic system. When the soil-water interaction compensation coefficient approaches zero, it indicates that the soil layer is sufficiently compressed and the pore water pressure has basically dissipated. Subsequent control links maintain the baseline flow distribution based on this near-zero value, keeping the system in a stable operating state. When the soil-water interaction compensation coefficient is significantly greater than zero, it indicates that the soil layer resistance is rapidly decaying and there is a risk of softening and instability. Subsequent control links trigger a large-flow feedforward mechanism based on this significantly large value, forcibly increasing the hydraulic system output to maintain the stability of the pile posture. Finally, the hydraulic actuator dynamically adjusts the oil supply ratio of the dual-sided leveling cylinders according to the soil-water interaction compensation coefficient, and simultaneously records the working condition boundary parameters.

[0038] S4: Based on the transverse filter tilt angle data, longitudinal filter tilt angle data, longitudinal settlement increment and transverse displacement increment, construct the torsional decoupling skew correction factor.

[0039] In one embodiment, when driving piles in strata with alternating hard and soft surfaces or with inclined interlayers, the pile body is prone to rotation around its own axis. Existing methods typically adjust based on the tilt angle in only one direction, failing to distinguish whether the pile body is truly shifting to one side or merely changing sensor readings due to rotation, easily leading to erroneous adjustment commands. Engineering principles show that the actual lateral drift of the pile body and the spurious offset caused by rotation exhibit different patterns in the two directions. To address this, this embodiment constructs a torsional decoupling skew correction factor. By cross-comparing the tilt data and displacement increments in two orthogonal directions, the true offset trend is independently extracted from the composite motion. This logic effectively solves the misjudgment problem of single-plane adjustment under rotational conditions. Cross-calculation filters out interference signals caused by pile body rotation, ensuring that control commands only compensate for the actual lateral drift. The data from the two directions obtained in the preprocessing stage are aligned and calculated here, ensuring that the input features have a clear spatial correspondence.

[0040] The core of calculating the torsional decoupling skew correction factor lies in distinguishing between true offset and rotational disturbances through cross-comparison. First, a numerator is constructed: the product of the lateral filtered tilt angle data and the longitudinal settlement increment is calculated as the third product; the product of the longitudinal filtered tilt angle data and the lateral displacement increment is calculated as the fourth product. Then, the difference between the third and fourth products is calculated to obtain a cross-cancellation term. The engineering intent of this cross-cancellation term is to utilize the differences in tilt and displacement changes in two directions to mutually cancel each other out: when the pile body only rotates around its axis, the changes in lateral tilt and longitudinal displacement, and the changes in longitudinal tilt and lateral displacement, will show a synchronous proportional relationship. After cross-subtraction, the result approaches zero, thus automatically filtering out false signals caused by rotation. When the pile body experiences true lateral drift, the tilt and displacement changes in one direction will be significantly stronger than in the other direction. After cross-subtraction, a non-zero result is retained, directly reflecting the strength and direction of the pure offset trend. Secondly, a denominator term is constructed to smoothly scale the calculation results: the sum of the squares of the lateral and longitudinal filtered tilt angles and a preset zero-prevention constant is used as a second sum. The square root of this second sum is then calculated, and its product with the lower limit clamping projection penetration rate is used as a smooth scaling term. The square root term represents the current overall tilt level and is used to normalize the cross-operation results to a uniform order of magnitude. The lower limit clamping projection penetration rate is a proportionality coefficient obtained after lower limit protection processing of the projection penetration rate, used to prevent abnormal amplification of the calculation results due to the velocity approaching zero when the pile body is paused or penetration is extremely slow. A structure combining cross-subtraction and denominator scaling is adopted to ensure that the torsional decoupling skew correction factor automatically decays to zero when the pile body rotates, and responds linearly when actual skew occurs, enabling the multi-degree-of-freedom leveling mechanism to accurately identify the offset direction and perform targeted compensation. The torsional decoupling skew correction factor is directly embedded as an asymmetric adjustment weight into the main diagonal of the flow distribution matrix to achieve differentiated flow control of the dual-side cylinders.

[0041] The specific value of the torsional decoupling skew correction factor directly reflects the separation ratio between the actual spatial attitude deviation and rotational disturbance. When the torsional decoupling skew correction factor approaches zero, it indicates that the pile is in a pure axial penetration state or only has slight rotation, with no significant actual lateral drift. Subsequent stages maintain the single-plane conventional compensation strategy based on this near-zero value. When the torsional decoupling skew correction factor is significantly greater than zero, it indicates that the pile has undergone significant spatial deflection and the actual deviation trend continues to amplify, posing a risk of uneven stress on the joint. Subsequent stages trigger multi-degree-of-freedom synchronous correction and lock the pile splicing mechanism based on this significantly larger value. Finally, the servo valve group distributes the flow in each direction and performs spatial attitude reconstruction according to the torsional decoupling skew correction factor, thereby effectively blocking the coupling and mutual interference paths between multiple degrees of freedom.

[0042] S5: Construct a hydraulic servo flow feedforward distribution matrix based on the yaw angle recursive index, the soil-water interaction compensation coefficient, and the torsional decoupling skew correction factor.

[0043] In one embodiment, the core logic of constructing the hydraulic servo flow feedforward allocation matrix lies in combining and allocating the global reference flow rate with the skew differential flow rate. First, the product of the two-dimensional identity matrix and the water-soil interaction compensation coefficient is calculated as the global reference flow rate term. The engineering intent of this operation is to provide a unified basic oil supply ratio for the dual-sided leveling cylinders, ensuring that the hydraulic system can maintain steady-state pressure and synchronous propulsion when there is no significant skew.

[0044] Secondly, the yaw differential flow term is constructed. Specifically, a two-dimensional diagonal matrix is ​​constructed with the torsional decoupling yaw correction factor and its negative values ​​as the main diagonal elements. The yaw differential flow term is obtained by multiplying the two-dimensional diagonal matrix with the yaw angle recursive index. This aims to automatically allocate the flow difference between the two sides according to the yaw intensity and direction: when the yaw angle recursive index is positive, the system automatically converts the positive deviation into the flow ratio of depressurization on the positive side cylinder and pressurization on the reverse side, increasing the oil supply ratio of the left cylinder and decreasing the oil supply ratio of the right cylinder; when the yaw angle recursive index is negative, the oil supply ratio of the left and right sides is automatically reversed. This positive and negative symmetrical structure eliminates the need for additional direction judgment logic, enabling self-identification of yaw direction and self-generation of differential commands.

[0045] The final hydraulic servo flow feedforward allocation matrix is ​​obtained by superimposing the global baseline flow term and the skew differential flow term. The engineering intent of using a diagonal matrix superposition structure is to adhere to the principle of independent control of hydraulic channels, ensuring that the baseline flow and differential flow do not interfere with each other. After expanding the hydraulic servo flow feedforward allocation matrix, the main diagonal elements correspond to the target oil supply ratios of the left and right leveling cylinders, respectively. Specifically, the target oil supply ratio of the left leveling cylinder is the sum of the products of the water-soil interaction compensation coefficient, the yaw angle recursive index, and the torsional decoupling skew correction factor; the target oil supply ratio of the right leveling cylinder is the water-soil interaction compensation coefficient minus the product of the yaw angle recursive index and the torsional decoupling skew correction factor.

[0046] The values ​​of the hydraulic servo flow feedforward distribution matrix directly reflect the target oil supply ratio distribution of the dual-side leveling cylinders. When the absolute value of the difference between the main diagonal elements of the hydraulic servo flow feedforward distribution matrix is ​​smaller, it indicates that the forces on both sides of the strata are balanced and there is no significant tendency for deviation. Subsequent control mechanisms maintain symmetrical oil supply based on this difference, causing the left and right leveling cylinders to advance synchronously. When the absolute value of the difference between the main diagonal elements of the hydraulic servo flow feedforward distribution matrix is ​​larger, it indicates that there is significant asymmetric resistance in the strata and the deviation trend continues to amplify. Subsequent control mechanisms trigger the dual-side asymmetric high-pressure oil supply setting based on this difference, forcibly increasing the flow difference between the two cylinders for rapid correction. Finally, the proportional directional valve outputs the target opening command based on the calculation results of the hydraulic servo flow feedforward distribution matrix, driving the left and right leveling cylinders to perform differential extension and retraction, achieving precise reconstruction and dynamic compensation of the pile's spatial attitude.

[0047] S6: Generate leveling cylinder control commands based on the hydraulic servo flow feedforward distribution matrix, and perform dynamic compensation based on the yaw angle recursive index determination results of the state machine.

[0048] In one embodiment, a preset stabilization threshold and a preset warning threshold are set. When the absolute value of the yaw angle recursive index is less than the preset stabilization threshold for a first preset number of control cycles, a stable penetration state is determined, and synchronous oil supply to the dual-side leveling cylinders is maintained. When the absolute value of the yaw angle recursive index is not less than the preset stabilization threshold and not greater than the preset warning threshold for a first preset number of control cycles, a warning state is determined, the current reference is locked, and cylinder preheating is initiated. When the absolute value of the yaw angle recursive index is greater than the preset warning threshold for a second preset number of control cycles, a high-risk state is determined, and the main pump frequency reduction and forced reset commands are executed. Under different determination states, the hydraulic servo flow feedforward allocation matrix is ​​invoked. Based on the positive and negative symmetry structure of the main diagonal elements of the matrix, the deviation polarity of the yaw angle recursive index is converted into the differential flow ratio command of the dual-side leveling cylinders.

[0049] For example, the control system continuously monitors the magnitude and direction of the yaw angle recursive index with a fixed control cycle of one second. When the absolute value of the yaw angle recursive index is below the preset stability threshold of 0.3 for five consecutive control cycles, it is determined that the current state is stable. At this time, regular data updates are maintained, and the leveling cylinders on both sides are controlled to keep synchronous oil supply to ensure that the pile body penetrates vertically and stably. When the absolute value of the yaw angle recursive index is in the range of 0.3 to 0.6 for five consecutive control cycles, that is, not less than the preset stability threshold and not greater than the preset warning threshold, the system enters the warning state. In this state, the system locks the current reference data and starts the leveling cylinder preheating program to prepare for possible correction actions with hydraulic response. When the absolute value of the yaw angle recursive index stably exceeds the preset warning threshold of 0.6 for three consecutive control cycles, the system triggers a high-risk state and immediately executes the main pump frequency reduction and forced reset commands to prevent irreversible serious tilting or structural damage to the pile body.

[0050] Under the aforementioned different judgment states, the system directly calls the calculated hydraulic servo flow feedforward allocation matrix. Utilizing the positive and negative symmetry of the main diagonal elements of this matrix, the system can automatically convert the positive deviation of the yaw angle recursive index into the flow ratio of pressure reduction in the forward-side leveling cylinder and pressure increase in the reverse-side leveling cylinder, or the negative deviation into the flow ratio of pressure reduction in the reverse-side leveling cylinder and pressure increase in the forward-side leveling cylinder. After being scaled by a fixed coefficient, the differential flow ratio is directly output as the drive current command for the electro-hydraulic servo proportional valve. This command drives the left and right leveling cylinders to perform differential extension and retraction actions strictly according to the calculated flow difference, thereby ensuring a strict match between the correction force and the cumulative skew intensity. This mechanism enables smooth fine-tuning in the early stages of skew and quickly pulls the pile trajectory back to the design axis in critical conditions, effectively ensuring the safety and quality of pile driving operations.

[0051] In summary, the present invention effectively overcomes the problems of slow adjustment lag, poor robustness, and slow convergence and low accuracy caused by coupling interference in the prior art by using multi-source feature extraction and feedforward allocation matrix. It has significant engineering application value and substantial progress.

[0052] The present invention was compared and verified with the traditional PID control method based on instantaneous dip angle feedback. The results showed that: (1) When dealing with the tilt trend caused by sudden changes in the strata, the traditional method often triggers correction only after the dip angle deviation has accumulated to a significant degree due to the lack of feedforward prediction, which easily leads to overshoot and oscillation; while the present invention extracts the evolution trend of continuous settlement and dip angle coupling through the yaw angle recursive index, realizes feedforward prediction and early intervention, eliminates the lag of the control action from the root, and significantly improves the dynamic response bandwidth and real-time action of the system.

[0053] (2) In saturated soft soil strata with drastic changes in pore water pressure, traditional methods lack direct quantification of soil-water mechanical disturbances, making it difficult for control parameters to adapt to strata changes. However, this invention uses the soil-water interaction compensation coefficient to quantify the instantaneous disturbance intensity caused by pore pressure gradient and penetration rate in real time, and dynamically corrects the feedforward flow by combining soil viscosity and dissipation laws, so that the leveling strategy has the ability to adapt to the geological environment, effectively overcomes the sudden attitude deviation caused by pore pressure accumulation, and greatly improves the robustness of attitude maintenance.

[0054] (3) In complex working conditions with inclined interlayers or alternating soft and hard surfaces, the pile body is prone to rotation around the axis. Traditional methods will cause repeated pulling due to the cross-influence caused by torsion when each degree of freedom moves independently. However, this invention uses a torsion decoupling skew correction factor to cross-compare the inclination angle and displacement increment in the orthogonal direction, accurately separating the real lateral drift and rotation interference. From the control mechanism, it blocks the mechanical coupling and mutual interference path between multi-axis leveling actions, avoids actuator oscillation, significantly accelerates the overall convergence speed of multi-axis correction, and improves the verticality accuracy of the pile.

[0055] A dynamic compensation system for the driving posture of a static pressure pile driver includes a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a dynamic compensation method for the driving posture of a static pressure pile driver according to the first aspect of the present invention.

[0056] A dynamic compensation system for the driving posture of a static pressure pile driver also includes other components well known to those skilled in the art, such as a communication bus and a communication interface. Their settings and functions are known in the art and will not be described in detail here.

[0057] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for dynamic compensation of the pile driving posture of a static pressure pile driver, characterized in that, include: Multi-source sensor data were collected and preprocessed during the pile driving process to obtain a standardized data sequence. The standardized data sequence includes longitudinal settlement increment, longitudinal filtered tilt angle data, pore water pressure spatial gradient, projected penetration rate, lateral filtered tilt angle data, and lateral displacement increment. Based on longitudinal settlement increment and longitudinal filtered tilt angle data, a yaw tilt angle recursive index is constructed; Based on the spatial gradient of pore water pressure and the projected penetration rate, a soil-water interaction compensation coefficient is constructed. Based on the transverse filter tilt angle data, the longitudinal filter tilt angle data, the longitudinal settlement increment and the transverse displacement increment, a torsional decoupling skew correction factor is constructed. Based on the yaw angle recursive index, the soil-water interaction compensation coefficient, and the torsional decoupling skew correction factor, a hydraulic servo flow feedforward allocation matrix is ​​constructed. The leveling cylinder control command is generated based on the hydraulic servo flow feedforward distribution matrix, and dynamic compensation is performed based on the determination result of the yaw angle recursive index by the state machine. The constructed yaw angle recursive index includes: Take any control cycle as the target cycle and the previous control cycle adjacent to the target cycle as the control cycle. Calculate the product of the longitudinal settlement increment of the target cycle and the longitudinal filtered tilt angle data to obtain the horizontal offset trend. The ratio of the change in longitudinal filtered tilt angle data between the target period and the control period to the control period is calculated to obtain the tilt angle change rate. The horizontal offset trend and the rate of change of inclination are weighted and summed, and then accumulated over a preset time window to obtain the yaw inclination recursive index. The constructed soil-water interaction compensation coefficient includes: The instantaneous disturbance intensity is obtained by calculating the product of the spatial gradient of pore water pressure and the projected penetration rate. Obtain the preset dynamic viscosity coefficient of the soil layer and the pressure dissipation characteristic constant. Calculate the product of the pressure dissipation characteristic constant and the cumulative time of pile driving as the first product. Apply an exponential function to the first product to perform a negative correlation mapping to obtain the mapping result. Calculate the difference between constant 1 and the mapping result. Calculate the sum of the difference and the preset zero-prevention constant as the first sum. Multiply the first sum and the preset dynamic viscosity coefficient of the soil layer as the second product. The ratio of the instantaneous disturbance intensity term to the second product is used as the soil-water interaction compensation coefficient. The constructed torsional decoupling skew correction factor includes: The product of the lateral filtered tilt angle data and the longitudinal settlement increment is calculated as the third product, and the product of the longitudinal filtered tilt angle data and the lateral displacement increment is calculated as the fourth product. The difference between the third and fourth products is used to obtain the cross-cancellation term; The projection penetration rate is subjected to a lower limit protection process to obtain the lower limit clamped projection penetration rate. The sum of the squares of the transverse filter tilt angle data, the squares of the longitudinal filter tilt angle data, and the preset zero-prevention constant is used as the second sum. The square root of the second sum is calculated, and the product of the square root and the lower limit clamping projection penetration rate is used as the smooth scaling term. The ratio of the cross-cancellation term to the smooth scaling term is used as the torsional decoupling skew correction factor. The construction of the hydraulic servo flow feedforward allocation matrix includes: The product of the calculated two-dimensional identity matrix and the soil-water interaction compensation coefficient is used as the global baseline flow term; A two-dimensional diagonal matrix with the torsional decoupling skew correction factor and its negative values ​​as the main diagonal elements is constructed, and the two-dimensional diagonal matrix is ​​multiplied by the yaw angle recursive index to obtain the skew differential flow term. The global reference flow term and the skew differential flow term are superimposed to obtain the hydraulic servo flow feedforward allocation matrix; The dynamic compensation for the determination result of the yaw angle recursive index by combining the state machine includes: Set preset stability thresholds and preset warning thresholds; When the absolute value of the yaw angle recursive index is less than the preset stable threshold for a first preset number of consecutive control cycles, it is determined to be in a stable penetration state, and the synchronous oil supply of the dual-side leveling cylinders is maintained. When the absolute value of the yaw angle recursive index is not less than the preset stability threshold and not greater than the preset warning threshold for a first preset number of consecutive control cycles, it is determined to be in a warning state, the current benchmark is locked and the cylinder preheating is started. When the absolute value of the yaw angle recursive index is greater than the preset warning threshold for a second consecutive number of control cycles, it is determined to be a high-risk state, and the main pump frequency reduction and forced reset commands are executed. Under different judgment states, the hydraulic servo flow feedforward allocation matrix is ​​invoked. Based on the positive and negative symmetry structure of the main diagonal elements of the matrix, the deviation polarity of the yaw angle recursive index is converted into the differential flow ratio command of the dual-side leveling cylinders.

2. The method for dynamic compensation of pile driving posture of a static pressure pile driver according to claim 1, characterized in that, The process of collecting and preprocessing multi-source sensor data during pile driving includes: The original inclination angle sequence of the pile body in two orthogonal directions, the change in vertical distance at the pile top, the offset of horizontal light spot at the pile top, the original pore fluid pressure value at different soil depths, and the advance distance of the pile driving cylinder are obtained. The original tilt angle sequence and displacement data are filtered to obtain the filtered sequence; Using the control cycle as the time reference, the filtered sequence, the original pore fluid pressure value, and the advance distance of the pile driving cylinder are periodically truncated and arithmetically averaged to obtain the average attitude, average displacement, average pore pressure gradient, and average advance speed for any control cycle; where the average attitude includes the average lateral attitude and the average longitudinal attitude, and the average displacement includes the average lateral displacement and the average longitudinal displacement. The mean lateral attitude is normalized to obtain the lateral filtered tilt angle data, the mean longitudinal attitude is normalized to obtain the longitudinal filtered tilt angle data, the mean longitudinal displacement is normalized to obtain the longitudinal settlement increment, the mean lateral displacement is normalized to obtain the lateral displacement increment, the mean pore pressure gradient is normalized to obtain the pore water pressure spatial gradient, and the mean propulsion velocity is normalized to obtain the projected penetration rate.

3. A dynamic compensation system for the driving posture of a static pressure pile driver, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement a dynamic compensation method for the pile driving posture of a static pressure pile driver according to any one of claims 1-2.

Images