Hydraulic vibration hammer multi-machine linkage piling control method and piling system under complex working conditions

Abstract

The invention relates to the technical field of offshore pile foundation construction and intelligent control, and discloses a hydraulic vibratory hammer multi-machine linkage piling control method and piling system under complex working conditions, and the control method comprises the steps of zero-amplitude starting, piling equipment operation and data acquisition, pile sinking process control, shutdown control and the like. A composite algorithm combining a double-loop AI model and neural network adaptive PID is adopted to realize multi-hammer linkage high-precision phase synchronization; the piling system comprises a plurality of groups of hydraulic vibration hammers, a sensing unit, an electric control pressure relief type phase transient compensator, an edge calculation controller and a domestic core part group. The problems that the multi-hammer synchronization precision is low, the pile body perpendicularity is difficult to control, the equipment reliability is insufficient and the like under the complex offshore working condition are solved, the phase synchronization precision can be smaller than or equal to + / -0.6 degrees, the noise is smaller than or equal to 72 dB, and the overall performance of a piling system is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of offshore pile foundation construction and intelligent control, and particularly to a pile driving control method and a pile driving system for multi-machine linkage of hydraulic vibratory hammers under complex working conditions. Background Technique

[0002] A hydraulic vibratory hammer is a device that uses hydraulic drive to rotate an eccentric block at high speed to generate a periodic centrifugal force, and applies vibration to a pile body to drive or extract the pile. Because of its high power density, good controllability, and strong adaptability to complex geological conditions, it has been widely used in the fields of building pile foundation construction, bridge and wharf construction, offshore oil platforms, wind power foundation construction, and other large-scale civil engineering fields.

[0003] In the construction of large-diameter piles, super-long piles, or hard, dense, and uneven geological conditions such as offshore wind power pile foundation construction, the single-hammer excitation force is often insufficient to overcome the formation resistance. Therefore, a method of synchronously operating multiple groups of hydraulic vibratory hammers is often adopted. By vector superposition of the excitation forces of multiple hammers, the total excitation force is increased to improve the penetration efficiency. At present, there are mainly two types of implementation methods for multi-machine hydraulic vibratory hammer synchronization: Mechanical synchronization: The output shafts of multiple groups of vibratory hammers are mechanically coupled through a rigid coupling shaft, a gearbox, or bevel gears to force each hammer to rotate in the same phase. Its advantage is that the phase consistency is guaranteed by mechanical rigidity. Theoretically, the synchronization accuracy is high and the reliability is good. However, it has the disadvantages of complex structure, high installation requirements, inapplicability to distributed pile driving with large spans, and difficult maintenance; Simple flow synchronization: It belongs to open-loop control. Multiple hammers share the same hydraulic pump source and are supplied with equal flow through a flow dividing valve. Theoretically, each hammer can have the same frequency. However, factors such as load differences, changes in hydraulic oil viscosity, and motor characteristic deviations are not considered. In actual operation, phase drift is likely to occur, resulting in uneven vibration, reduced efficiency, and even structural damage.

[0004] With the development of large-scale projects such as offshore wind power and cross-sea bridges, ocean pile driving has become an important application scenario for hydraulic vibratory hammers. Ocean pile driving will face problems such as dynamic deviation of pile position and attitude caused by hull sway, sudden change of complex geological load, low equipment reliability in a highly corrosive environment, and insufficient flexibility of distributed pile driving. Higher requirements are put forward for the synchronization accuracy, position and attitude stability, and environmental adaptability of multi-machine linkage control. The existing mechanical synchronization and simple flow synchronization methods are difficult to meet the requirements in the ocean construction scenario. Moreover, in the existing technology, the automation degree of pile sinking and pulling control of construction machinery vibratory hammers is low, and the workload of workers is large. It is difficult to take into account aspects such as automation and intelligent control, precise control, structural strength, and safety protection. The equipment performance and operation convenience are poor and cannot meet the needs of modern engineering construction.

[0005] For example, CN118936943A discloses a multi-hammer linkage synchronization test system and evaluation method for hydraulic vibratory hammers. The core of this method is to collect vibration frequency, rotation frequency, and amplitude data, and calculate parameters such as eccentric moment and excitation force through curve analysis to evaluate whether the frequency and amplitude deviations meet ±5%. However, it cannot solve the problems of real-time high-precision control of multi-hammer linkage and dynamic correction under marine conditions. CN114457802A discloses a multi-group linkage hydraulic vibratory hammer group for stable pile clamping. It achieves mechanical synchronization through linkage shaft and angular gearbox and optimizes the pile clamping structure to improve stability. However, it also cannot solve the problems of phase drift and verticality deviation under dynamic marine conditions. CN119843658A discloses a pile driving construction control method for pipe piles. It achieves pile driving by obtaining the stratum thickness and natural frequency, adjusting the frequency of the vibratory hammer, and using satellite positioning to locate the position of the pipe pile. However, it cannot solve the problems of multi-hammer coordination and dynamic correction under complex marine conditions.

[0006] Therefore, existing technologies either focus only on testing and evaluating the synchronization of multiple hammers, without addressing real-time dynamic control and intelligent adaptive adjustment; or rely on mechanical structures for synchronization, which cannot adapt to complex dynamic conditions at sea; or only adjust the pile driving frequency, failing to solve the core problems of high-precision phase synchronization of multiple hammers and dynamic correction of pile verticality, and lack effective solutions for adapting to domestically produced core components and optimizing energy consumption and noise. Therefore, there is an urgent need for a multi-machine linkage control method and device that does not rely on rigid mechanical connections, can synchronize the phase difference of each hammer in real time, dynamically correct deviations, and is compatible with domestically produced components, to ensure construction safety, efficiency, and equipment reliability under complex marine conditions. Summary of the Invention

[0007] The purpose of this invention is to address the aforementioned shortcomings by providing a multi-machine linkage control method and system for pile driving under complex working conditions using hydraulic vibratory hammers. Through coordinated improvements in hardware and software, intelligent control is adopted, combined with improved equipment design, to enhance the intelligence level of the equipment. This reduces the workload of operators and effectively balances automation and intelligent control, precise control, structural strength, and safety protection, thereby improving equipment performance and ease of operation to meet the needs of modern engineering construction under complex working conditions such as offshore pile foundation construction.

[0008] This invention provides the following technical solution: A pile driving control method for multi-machine linkage of hydraulic vibratory hammers under complex working conditions, characterized by the following steps: S1, Zero Amplitude Start Set up a multi-machine linkage hydraulic vibratory hammer pile driving equipment and reset it. Before starting the equipment, adjust the eccentric angle of the paired eccentric blocks to zero through the eccentric block adjustment mechanism, start the drive motor, and make the eccentric blocks rotate to the rated speed without generating amplitude or excitation force. S2. Operation and Data Acquisition of Pile Driving Equipment Power on the piling equipment and collect the operating data of each hammer group in real time through the sensing unit of the piling equipment. The operating data includes the rotational speed of the eccentric block, phase angle, hydraulic motor working flow, eccentric vibrator working temperature, pile body position and inclination angle, and acceleration data. S3, Pile Driving Process Control Once all hammer groups are fully synchronized, a dual-loop AI model is used to fuse the collected operational data. Based on the on-site geological conditions and real-time pile driving resistance data, the angle of the eccentric block is dynamically adjusted to achieve the optimal value of the eccentric torque required for pile driving. During the pile driving process, the angle of the eccentric block, the bearing status, and the verticality of the pile body of each hammer group are continuously monitored through the sensing unit to avoid polarization phenomena. When any eccentric block is detected to have a speed deviation or angle deviation, it is automatically corrected by the AI ​​composite control algorithm to make the eccentric blocks of each hammer group reach the same phase angle, so that the phase synchronization accuracy of multiple hammer groups is ≤±0.6°; the AI ​​composite control algorithm is a composite algorithm that combines a dual-loop AI model with a neural network adaptive PID algorithm; S4, Shutdown Control When the pile reaches the design elevation, the phase angle of the eccentric block is adjusted to zero, the vibration disappears, and then the machine is stopped by synchronously reducing the speed of the drive motor.

[0009] A multi-machine linkage piling system for hydraulic vibratory hammers under complex working conditions, comprising: The piling unit includes multiple sets of hydraulic vibratory hammers arranged in parallel below the main body's upper hanger. The top is rigidly connected to the mounting flange at the bottom of the hanger by a set of high-strength bolts, and the bottom is clamped by a pile clamp to hold the steel pipe pile to be constructed. The main mechanical structure of each set of hydraulic vibratory hammers consists of an outer vibration damping frame, a vibration box, and a pile clamp. The vibration box is equipped with an eccentric vibrator driven by a hydraulic motor. The sensing unit includes a rotary encoder, a hydraulic flow sensor, a temperature sensor, an inclination sensor, and an accelerometer array, which are installed at corresponding positions on the piling unit to collect operating data. The AI ​​control unit, specifically an edge computing controller deploying the dual-loop AI model, includes a signal input module, a central controller, a control output module, a digital twin module, and a human-machine interaction module. The signal input module receives data collected by the sensing unit. The central controller incorporates the AI ​​composite control algorithm to process the data and output control commands. The control output module sends control commands to the synchronous motor and synchronous flow compensator. The digital twin module is used for offline training of the Transformer-TCN network. The human-machine interaction module is used for synchronous parameter configuration and visualization of multi-machine operating status. The hydraulic power unit includes a hydraulic vibratory hammer power station, a synchronous motor, and a synchronous flow compensator. The hydraulic vibratory hammer power station provides basic hydraulic and electrical energy to the system. The synchronous motor is integrated into the outlet oil circuit of the power station to achieve coarse synchronous control of the initial flow distribution of multiple hydraulic oil circuits. The synchronous flow compensator is arranged in parallel with the synchronous motor to provide hydraulic power for closed-loop fine adjustment and achieve micro-synchronous correction. The verticality monitoring and correction module is connected to the tilt sensor and AI control unit to receive pile position and tilt angle data in real time and output correction commands through the AI ​​control unit. The advantages of this invention include at least the following: 1. High-precision phase synchronization: By adopting a "dual-loop AI model + electronically controlled pressure relief compensator", the phase synchronization accuracy of multi-hammer linkage is ≤ ±0.6°, which is better than the ±0.8° of the existing technology, and solves the problem of phase drift and dynamic deviation under complex marine conditions.

[0010] 2. Strong adaptability to working conditions: It integrates technologies such as ground natural frequency feedforward and high damping mode switching, and is suitable for complex working conditions such as offshore construction of ultra-large diameter steel pipe piles with a diameter of more than 10m, wall thickness ≥25mm, wind speed ≥15m / s, and wave height ≥1.5m.

[0011] 3. High localization rate: The localization rate of core components is ≥90%, and domestic AI chips are used to reduce equipment costs and dependence on foreign countries, which meets the requirements of domestic self-reliance and control.

[0012] 4. Excellent performance indicators: noise ≤72dB, energy consumption is reduced by more than 15% compared with imported equipment, peak tensile stress at the pile head ≤2.1MPa, avoiding polarization phenomenon and pile damage, and the overall performance reaches the international advanced level.

[0013] 5. High level of intelligence: The system adopts federated learning deployment, with each construction vessel only uploading gradients, ensuring continuous model evolution in low-bandwidth offshore scenarios and enabling visualization of the construction process, equipment health management, and intelligent scheduling. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the overall structure of the piling system according to an embodiment of the present invention; Figure 2 This is a schematic diagram showing the composition and connection relationship of the piling system according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating the piling control method according to an embodiment of the present invention; Figure 4 This is a three-dimensional structural schematic diagram of a set of vibratory hammers in an embodiment of the present invention; Figure 5 This is a schematic diagram of the connection and control structure of multiple sets of vibratory hammers in an embodiment of the present invention.

[0015] In the picture: 1. External damping frame of hydraulic vibratory hammer; 2. Vibration box; 3. Hydraulic motor; 4. Hydraulic flow sensor; 5. Eccentric exciter; 6. Rotary encoder; 7. Pile clamp; 8. Temperature sensor; 9. Main body upper hanging frame; 10. Steel pipe pile to be constructed; 11. Single set of hydraulic vibratory hammer; 12. Synchronous motor; 13. Hydraulic vibratory hammer power station; 14. Synchronous flow compensator; 15. Edge computing controller; 16. Accelerometer array (accelerometer sensor matrix). Detailed Implementation

[0016] The embodiments of the present invention will be described in detail below.

[0017] Basic Implementation See Figures 1-5 The piling control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions provided in this embodiment includes the following steps: S1, Zero Amplitude Start Set up a multi-machine linkage hydraulic vibratory hammer pile driving equipment and reset it. Before starting the equipment, adjust the eccentric angle of the paired eccentric blocks to zero through the eccentric block adjustment mechanism, start the drive motor, and make the eccentric blocks rotate to the rated speed without generating amplitude or excitation force. S2. Operation and Data Acquisition of Pile Driving Equipment Power on the piling equipment and collect the operating data of each hammer group in real time through the sensing unit of the piling equipment. The operating data includes the rotational speed of the eccentric block, phase angle, hydraulic motor working flow, eccentric vibrator working temperature, pile body position and inclination angle, and acceleration data. S3, Pile Driving Process Control Once all hammer groups are fully synchronized, a dual-loop AI model is used to fuse the collected operational data. Based on the on-site geological conditions and real-time pile driving resistance data, the angle of the eccentric block is dynamically adjusted to achieve the optimal value of the eccentric torque required for pile driving. During the pile driving process, the angle of the eccentric block, the bearing status, and the verticality of the pile body of each hammer group are continuously monitored through the sensing unit to avoid polarization phenomena. When any eccentric block is detected to have a speed deviation or angle deviation, it is automatically corrected by the AI ​​composite control algorithm to make the eccentric blocks of each hammer group reach the same phase angle, so that the phase synchronization accuracy of multiple hammer groups is ≤±0.6°; the AI ​​composite control algorithm is a composite algorithm that combines a dual-loop AI model with a neural network adaptive PID algorithm; S4, Shutdown Control When the pile reaches the design elevation, the phase angle of the eccentric block is adjusted to zero, the vibration disappears, and then the machine is stopped by synchronously reducing the speed of the drive motor.

[0018] The dual-loop AI model comprises: an outer loop, a prediction model based on digital twins, which uses on-site geology, pile type, and historical construction data as inputs, and trains an offline Transformer-TCN hybrid network to predict the optimal phase difference setpoint Δθ 0–5 seconds in advance; and an inner loop, an online reinforcement learning fine-tuning model (PPO reinforcement learning network model), which uses the proximal policy optimization (PPO) algorithm in the neighborhood of the predicted value Δθ, with phase synchronization error e_θ, excitation force fluctuation rate e_F, and energy consumption increment ΔP as the immediate reward function, and outputs the hydraulic pump displacement correction amount Δu in real time.

[0019] In the dual-loop AI model, the Transformer-TCN hybrid network consists of two Transformer encoder layers (each with four attention heads) and two TCN layers (inflation coefficient 2), with ReLU as the activation function; the discount factor γ=0.99, learning rate η=0.003, and batch size 64 for the PPO algorithm; the immediate reward function R = −(w1·|e_θ| + w2·|e_F| + w3·|ΔP|), where w1:w2:w3=3:2:1; The neural network adaptive PID algorithm has the following input layers: total error E and error rate of change EC = dE / dt; two hidden layers with 10 nodes each; ReLU activation function; and PID parameter adjustments ΔKp, ΔKi, and ΔKd. The PID parameter update formula is as follows: Kp(t) = Kp(t-1) + η × ΔKp Ki (t) = Ki (t-1) + η × ΔKi Kd(t) = Kd(t-1) + η × ΔKd Where η is the learning rate, which ranges from 0.01 to 0.05 and is determined through experimental calibration.

[0020] In step S3, the calculation of the total error E includes: The master-slave flow difference eq=Qref-Qi, the master-slave phase difference eθ=θref -θi, the master-slave temperature difference eT=Tref-Ti, and the pile inclination angle difference eα=αref-αi are dynamically optimized using a reinforcement learning algorithm. The formula for calculating the total error E is: E (i)=wQ×|eq|+wθ×|eθ|+wT×|eT|+wα×|eα| Where Qref is the reference host traffic, Qi is the i-th slave traffic; θref is the reference host phase angle, θi is the i-th slave phase angle; Tref is the reference host temperature, Ti is the i-th slave temperature; αref is the target inclination angle of the pile, and αi is the real-time inclination angle of the pile.

[0021] Step S3 further includes the following steps: S3-1 Training Data Augmentation: Using vibration acceleration-time curves as additional state variables, a 5-dimensional observation vector O={e_θ, e_F, ΔP, a_rms, ψ} is constructed, where a_rms is the root mean square of acceleration and ψ is the vibration reduction coefficient, thereby improving the model's sensitivity to polarization-pile head tear risk; S3-2 Federated Learning Deployment: Each construction vessel only uploads gradients, and the central server aggregates and updates them, ensuring continuous model evolution in low-bandwidth maritime scenarios and meeting the requirements of domestic independent control (domestic production rate of core hydraulic pumps and valves ≥90%). S3-3 Compensates for the lag in elastic deformation of mechanical components: An electrically controlled pressure relief phase transient compensator is added to the mechanical synchronization of the angular gearbox-universal coupling. When the phase difference between adjacent hammers is detected to be >1°, the controller commands the high-speed switching valve to instantaneously depressurize the low-pressure chamber of the corresponding motor for 20–40ms, generating a braking torque of −50~−150N·m, which pulls back the phase within 0.3s. S3-4 Formation natural frequency feedforward: The digital twin model performs modal matching between the formation natural frequency f_s and the system natural frequency f_0. If |f_s−f_0|<2Hz, the excitation force is automatically reduced by 10% to avoid resonance causing pile floating or boom fatigue, and the noise is ≤72dB. S3-5 High Damping Mode Switching: If e_θ is still > 0.6° after 3 consecutive PPO corrections, switch to high damping mode, simultaneously reduce the rotational speed of all hammers by 8%, and increase the included angle of the eccentric blocks by 3° to ensure that the peak tensile stress at the pile head is ≤ 2.1MPa; S3-6 Instruction Output and Execution For multiple slave computing UIs, the instruction output module converts the UIs into electrical signal instructions, which are then transmitted to the synchronization flow compensator through the control output module to dynamically adjust the hydraulic pump displacement, ultimately achieving multi-machine synchronization.

[0022] In step S3, for ship sway compensation under marine conditions, the tilt angle data of the pile body is collected in real time by tilt angle sensor. When the tilt angle deviation exceeds 0.3°, the AI ​​composite control algorithm prioritizes adjusting the phase angle and excitation force of the corresponding side hammer group to generate a reverse correction torque so that the pile body returns to a vertical state.

[0023] In step S3, the eccentric block angle adjustment strategy is dynamically set based on the formation resistance level. When the formation resistance is less than 5000kN, the eccentric block angle adjustment range is 0°-30°; when the formation resistance is 5000kN-10000kN, the adjustment range is 30°-60°; when the formation resistance is greater than 10000kN, the adjustment range is 60°-90°, and each adjustment increment does not exceed 5°.

[0024] Step S3 also includes an energy consumption optimization step: the AI ​​composite control algorithm dynamically adjusts the hydraulic pump displacement and drive motor speed based on real-time load data and hydraulic system efficiency model, so that the system works in the optimal efficiency range; it also includes a noise control step: by monitoring the operating noise data of the vibratory hammer, when the noise exceeds 72 decibels, the rotation frequency of the eccentric block and the hydraulic system pressure are adjusted, and combined with the vibration reduction structure design, the noise is controlled below 75 decibels.

[0025] The sensing unit includes a rotary encoder, a hydraulic flow sensor, a temperature sensor, an accelerometer array, and a tilt sensor. The rotary encoder is installed at the end of the shaft of the eccentric vibrator, the hydraulic flow sensor is installed in the oil outlet of the hydraulic motor, the temperature sensor is installed on the outer end cover of the eccentric vibrator, the tilt sensor is installed on the top of the pile, and the accelerometer array is installed at the vibrator, bearing, and pile cap. The data fusion processing adopts the Kalman filter algorithm to filter, reduce noise, and complement the raw signals collected by each sensor.

[0026] A multi-machine linkage piling system for hydraulic vibratory hammers under complex working conditions, used to implement the aforementioned piling control method, includes: The piling unit includes multiple sets of hydraulic vibratory hammers arranged in parallel below the main body's upper hanger. The top is rigidly connected to the mounting flange at the bottom of the hanger by a set of high-strength bolts, and the bottom is clamped by a pile clamp to hold the steel pipe pile to be constructed. The main mechanical structure of each set of hydraulic vibratory hammers consists of an outer vibration damping frame, a vibration box, and a pile clamp. The vibration box is equipped with an eccentric vibrator driven by a hydraulic motor. The AI ​​control unit, specifically an edge computing controller deploying the dual-loop AI model, includes a signal input module, a central controller, a control output module, a digital twin module, sensing units, and a human-machine interaction module. The signal input module receives data collected by the sensing units. The central controller incorporates the AI ​​composite control algorithm to process the data and output control commands. The control output module sends control commands to the synchronous motor and synchronous flow compensator. The digital twin module is used for offline training of the Transformer-TCN network. The sensing units, including a rotary encoder, hydraulic flow sensor, temperature sensor, tilt sensor, and accelerometer array, are installed at corresponding positions on the piling unit to collect operational data. The human-machine interaction module is used for synchronous parameter configuration and visualization of the multi-machine operating status. The hydraulic power unit includes a hydraulic vibratory hammer power station, a synchronous motor, and a synchronous flow compensator. The hydraulic vibratory hammer power station provides basic hydraulic and electrical energy to the system. The synchronous motor is integrated into the outlet oil circuit of the power station to achieve coarse synchronous control of the initial flow distribution of multiple hydraulic oil circuits. The synchronous flow compensator is arranged in parallel with the synchronous motor to provide hydraulic power for closed-loop fine adjustment and achieve micro-synchronous correction. The verticality monitoring and correction unit is connected to the tilt sensor and AI control unit to receive pile position and tilt angle data in real time and output correction commands through the AI ​​control unit. It includes an electrically controlled pressure relief phase transient compensator, which is a combination of a high-speed switching valve and a low-pressure accumulator with a response time of ≤20ms, an accumulator volume of 5L, and a working pressure of 10–15MPa.

[0027] The domestically produced core components, including hydraulic pumps, hydraulic valves, sensors, and actuators, have a domestic production rate of over 90% and are adapted to the control strategies of the AI ​​control unit and the working requirements of the hydraulic system.

[0028] The rotary encoder is installed at the end of the shaft of the eccentric vibrator, the hydraulic flow sensor is installed in the oil outlet of the hydraulic motor, the temperature sensor is installed on the outer end cover of the eccentric vibrator, the tilt sensor is installed on the top of the pile and the outside of the vibration box, and the accelerometer array is installed at the vibrator, bearing and pile cap. The acquisition frequency of each sensor is not less than 1000Hz.

[0029] The central controller of the AI ​​control unit adopts an ARM Cortex-A9 quad-core processor with a main frequency of no less than 1.2GHz, and has 8GB of built-in storage space for storing operating data and algorithm models. It supports remote data transmission and firmware upgrades. The signal input module supports analog, digital and communication interface inputs, and the control output module supports PWM signals and analog outputs to adapt to the control requirements of the swashplate actuator and drive motor.

[0030] In the hydraulic power unit, the synchronous motor adopts a synchronous motor with the same displacement, and the flow distribution error does not exceed ±2%; the synchronous flow compensator adopts a small displacement hydraulic pump with a displacement adjustment range of 0-500mL / r and a response time of no more than 20ms, ensuring the speed and accuracy of phase adjustment.

[0031] Among the domestically produced core components, the hydraulic pump is an axial piston pump with a rated working pressure of not less than 31.5 MPa and a volumetric efficiency of not less than 95%; the hydraulic valve is an electro-hydraulic proportional valve with a control accuracy of not less than ±0.5% FS; and the sensor has anti-corrosion and anti-vibration properties and is suitable for high humidity and high salt spray environments at sea.

[0032] The piling system also includes a human-machine interface module with a touch screen that can display the rotation speed, phase angle, excitation force, energy consumption, noise, and pile verticality data of each hammer group in real time. It supports manual parameter setting and control mode switching (automatic / manual).

[0033] The piling system also includes a remote data transmission module that uses 5G or satellite communication technology to upload construction data to a cloud platform in real time, enabling visualization of the construction process, equipment health management, and intelligent scheduling.

[0034] The main body's upper hanger integrates a standard hoisting interface, which is rigidly connected to external lifting equipment. The spacing of multiple sets of hydraulic vibratory hammers can be adjusted according to the diameter of the steel pipe pile, adapting to the construction needs of ultra-large steel pipe piles with a diameter of 10m or more.

[0035] The complex working conditions refer to offshore ultra-large diameter steel pipe piles of more than 10 m, with a wall thickness of ≥25 mm, wind speed ≤15 m / s, and wave height ≤1.5 m.

[0036] The piling system supports a ring arrangement of 8 or more hammers, with a single hammer excitation force ≥2000 kN and a total excitation force ≥15000 kN.

[0037] Example 1 This invention provides a pile driving control method for multi-machine linkage of hydraulic vibratory hammers under complex working conditions, based on the basic embodiment. It is applied to offshore wind power foundation construction (steel pipe pile diameter 12m, wall thickness 30mm, wind speed 18m / s, wave height 1.8m). The specific steps are as follows: S1. Zero Amplitude Start-up: Set up and reset the multi-machine linkage hydraulic vibratory hammer pile driving equipment. Adjust the eccentric angle of all pairs of eccentric blocks to zero through the eccentric block adjustment mechanism. Start the drive motor (hydraulic motor) of each hammer group and control the speed of the drive motor to gradually increase to the rated speed of 1500r / min. At this time, the eccentric blocks rotate but do not generate amplitude or excitation force.

[0038] S2. Operation and Data Acquisition of Pile Driving Equipment: Power on the pile driving equipment and collect real-time operating data of each hammer group through a rotary encoder (resolution 0.01°), flow sensor (measurement range 0-200L / min, accuracy ±0.5%FS), and accelerometer array (sampling rate 1kHz), including the rotational speed and phase angle of the eccentric block, the working flow of the hydraulic motor, the working temperature of the eccentric vibrator, and the pile body posture and inclination angle and acceleration data.

[0039] S3. Pile driving process control: Once all hammer groups are fully synchronized, a dual-loop AI model is used to fuse the collected operational data. Based on the on-site geological conditions (such as the natural frequency of the stratum f_s=12Hz) and real-time pile driving resistance data, the angle of the eccentric block is dynamically adjusted so that the eccentric moment reaches the optimal value required for pile driving. During the pile driving process, the eccentric block angle, bearing status and pile verticality of each hammer group are continuously monitored through the sensing unit to avoid polarization phenomenon. When any eccentric block is detected to have a speed deviation or angle deviation, it is automatically corrected through AI composite control algorithm to make the eccentric blocks of each hammer group reach the same phase angle, so that the phase synchronization accuracy of multiple hammer groups is ≤±0.6°. The outer ring Transformer-TCN network of the dual-loop AI model predicts the optimal phase difference setting value Δθ=0.2° within the next 3 seconds based on historical construction data (geological parameters, pile type, pile driving resistance). The inner-loop PPO algorithm uses the phase synchronization error e_θ=θref-θi, the excitation force fluctuation rate e_F=|Fref-Fi| / Fref, and the energy consumption increment ΔP=P-Pref as the instantaneous reward function, and outputs a hydraulic pump displacement correction amount Δu=±5mL / r to control the phase synchronization accuracy of each hammer within ±0.5°. Training data augmentation introduces vibration acceleration-time curves to construct a 5-dimensional observation vector O={e_θ, e_F, ΔP,a_rms, ψ}, thereby improving the model's sensitivity to the risk of "polarization-pile head tearing". Federated learning deployment: Each construction vessel only uploads gradients, while the central server of the AI ​​control unit aggregates and updates the model, ensuring continuous model evolution in low-bandwidth maritime scenarios; When the phase difference between adjacent hammers is detected to be 1.2°, the electronically controlled pressure relief phase transient compensator commands the high-speed switching valve to instantaneously depressurize the corresponding motor low-pressure chamber for 30ms, generating a braking torque of −100N·m, and pulling the phase back to ±0.5° within 0.25s. The digital twin model performs modal matching between the natural frequency of the stratum f_s=12Hz and the natural frequency of the system f_0=11Hz. |f_s−f_0|=1Hz<2Hz, automatically reducing the excitation force by 10% to avoid resonance causing pile floating or crane fatigue, with noise ≤72dB. S4. Shutdown control: When the pile reaches the design elevation (-60m), the phase angle of the eccentric block is adjusted to zero, the vibration disappears, and then the drive motor (hydraulic motor) speed is reduced to 300r / min to achieve shutdown.

[0040] Example 2 This embodiment, based on the basic embodiment, provides a multi-machine linkage piling system for hydraulic vibratory hammers under complex working conditions, which includes: The piling unit consists of 8 sets (units) of hydraulic vibratory hammers. Figure 5 Four sets are shown in the middle, evenly arranged side by side along the bottom of the main body upper hanger, with a spacing of 1.5m. The top is rigidly connected to the installation flange at the bottom of the hanger by M30 high-strength bolts, and the bottom is clamped by a pile clamp to clamp the steel pipe pile (12m in diameter and 30mm in wall thickness) to be constructed. Each set of hydraulic vibratory hammers has an external vibration damping frame connected to the vibration box using rubber vibration damping pads. The vibration box is equipped with two symmetrically arranged hydraulic motors that drive the eccentric vibrator to rotate. The eccentric vibrator contains three sets of paired eccentric block assemblies.

[0041] A rotary encoder (resolution 0.01°) is installed at the end of the shaft of the eccentric vibrator, a flow sensor (measurement range 0-200L / min, accuracy ±0.5%FS) is installed in the oil outlet of the hydraulic motor, and an accelerometer array (sampling rate 1kHz) is installed at the vibrator, bearing, and pile cap.

[0042] The electrically controlled pressure relief phase transient compensator is a combination of a high-speed switching valve (response time 15ms) and a low-pressure accumulator (capacity 5L, working pressure 12MPa), installed in the low-pressure oil circuit of the hydraulic motor.

[0043] The edge computing controller uses the domestic Huawei Ascend 310 AI chip with a main frequency of 1.2GHz, 8GB of built-in storage space, and deploys a dual-loop AI model and digital twin module. It supports 8 or more hammers in a ring, triangle, or straight arrangement, with a single hammer excitation force of 2500kN and a total excitation force of 20000kN.

[0044] The verticality monitoring and correction module is connected to the tilt sensor (measurement range ±5°, accuracy ±0.01°) and the AI ​​control unit to receive pile position and tilt angle data in real time and output correction commands through the AI ​​control unit.

[0045] In this embodiment, the domestically produced core components used include: a domestically produced axial piston pump (model CY14-1B, rated pressure 31.5MPa, volumetric efficiency 96%), a domestically produced electro-hydraulic proportional valve (model 4WREE6, control accuracy ±0.3%FS), and domestically produced corrosion-resistant sensors with an IP67 protection rating, suitable for marine salt spray environments.

[0046] The piling system is connected to a 2000t crane vessel via a gantry crane and has been tested to meet the DNV-GL classification society's offshore operation certification requirements.

[0047] The engineering prototype of this piling system has undergone confidential testing in offshore wind power foundation construction. It can achieve multi-hammer linkage phase synchronization accuracy of ±0.5°, noise ≤70dB, energy consumption is reduced by more than 10% compared with imported similar equipment, and the peak tensile stress of the pile head is 2.0MPa. The overall performance meets the requirements of modern intelligent pile foundation construction.

[0048] The structure of the single-unit hydraulic vibratory hammer used in this embodiment is as follows: Figure 4 As shown, the main mechanical structure consists of an outer vibration damping frame 1, a vibration box 2, and a pile clamp 7. The top of the outer vibration damping frame 1 of the hydraulic vibratory hammer is equipped with a lifting connection structure. The hook of the lifting equipment engages with this lifting connection structure, thereby enabling vertical lifting of the entire machine. The lower part of the outer vibration damping frame 1 is mechanically connected to the vibration box 2 in a vertically corresponding manner through a vibration damping connection structure. The internal power mechanism of the vibration box 2 consists of a hydraulic motor 3 and its driven eccentric vibrator 5. The hydraulic motor 3 is symmetrically arranged left and right, driving the eccentric vibrator 5 to rotate. The eccentric vibrator 5 includes at least two sets of eccentric block assemblies, utilizing the vector synthesis effect of the centrifugal force of the eccentric blocks to output a periodic excitation force distributed along the axis towards the pile body. The pile clamp 7 is installed at the bottom output end of the vibration box 2.

[0049] The vibration chamber 2 is equipped with a rotary encoder 6, a hydraulic flow sensor 4, and a temperature sensor 8 for signal acquisition and closed-loop control. The hydraulic flow sensor 4 is installed in the oil outlet circuit of the hydraulic motor 3 to collect real-time operating flow data. The rotary encoder 6 is installed at the end of the shaft of the eccentric vibrator 5 to detect the phase angle and rotational speed parameters of the eccentric vibrator 5. The temperature sensor 8 is installed on the external end cover of the eccentric vibrator 5 to detect the operating temperature of the eccentric vibrator.

[0050] The main connection and control structure of the multi-hammer linkage is as follows: Figure 5As shown. In this embodiment, the multi-hammer linkage piling system achieves overall suspension and positioning through the main upper hanger 9. This hanger is set at the top of the piling system and integrates standard lifting interfaces (such as lifting lugs and pin connection structures) for rigid connection with external lifting equipment (such as crawler cranes). Multiple sets of hydraulic vibratory hammers 11 are arranged side by side (in a straight line) below the main upper hanger 9, and their tops are rigidly connected to the mounting flange at the bottom of the hanger through high-strength bolt sets, forming an integrated suspension structure. Each set of hydraulic vibratory hammers 11 is clamped by a pile clamp 7 for the steel pipe pile 10 to be constructed.

[0051] The hydraulic vibratory hammer power station 13 is connected to the hydraulic vibratory hammer 11 via oil and electrical circuits, providing basic hydraulic and electrical energy to the system. To improve the synchronization of multi-hammer coordinated operation, a synchronous motor 12 is integrated into the outlet oil circuit of the power station to perform preliminary flow distribution on multiple hydraulic oil circuits, achieving "coarse synchronization" control. The synchronous flow compensator 14 is arranged in parallel with the synchronous motor 12, providing hydraulic power for closed-loop fine-tuning and performing "micro-synchronization" correction on the phase difference and excitation force of the vibratory hammer.

[0052] The controller 15 is deployed on the side of the power station or in an independent control station, and includes components such as a signal input module, a central controller, a control output module, and a human-machine interface. The controller 15 collects data from the rotary encoder 6, hydraulic flow sensor 4, temperature sensor, and accelerometer array 16 in real time, and outputs control commands to the synchronous flow compensator 14 through the software system to achieve precise flow matching of the excitation system and intelligent control of equipment status.

[0053] The signal input module has analog, digital, and communication input functions, and is used to receive rotary encoder signals, hydraulic flow sensor signals, and temperature sensor signals from the hydraulic vibratory hammer unit.

[0054] The central controller processes the received signals and executes a multi-machine synchronous control algorithm. The control output module has a signal output function, used to send control commands to the synchronous motor and the synchronous flow compensator.

[0055] The human-machine interface module is a display and input / output device used to visualize the configuration of synchronous parameters and the operating status of multiple machines. The controller's software system includes a drive interface, data processing, vibratory hammer phase processing, an AI composite control algorithm for synchronous control, and command output.

[0056] Example 3 This embodiment, based on Embodiments 1 and 2, provides a multi-machine linkage hydraulic vibratory hammer piling system for nearshore shallow sea construction in soft soil strata (stratum resistance < 5000kN) under complex working conditions. The difference lies in: This embodiment is applied to the auxiliary pile construction of a near-shore shallow sea cross-sea bridge (soft soil layer, soil resistance 3800kN), and is suitable for small-diameter piles and the requirements of ring-shaped hammer. The specific pile driving control steps are as follows: Zero amplitude start-up: Set up 6 sets of hydraulic vibratory hammers in a ring (hammer diameter 8m, spacing 1.3m) and reset them. Adjust the eccentricity angle of the paired eccentric blocks to 0° through the eccentric block adjustment mechanism, start the drive motor, and gradually increase to the rated speed of 1200r / min. No amplitude or excitation force is generated.

[0057] Operation and data acquisition of piling equipment: After the equipment is powered on, the sensing unit (rotary encoder resolution 0.01°, flow sensor measurement range 0-180L / min, accuracy ±0.5% FS, accelerometer array sampling rate 1kHz) collects data in real time, including eccentric block rotation speed 1200r / min, phase angle, hydraulic motor working flow rate 150L / min, eccentric vibrator working temperature 45℃, pile body posture inclination angle 0.4°, and acceleration data.

[0058] Pile driving process control: After all hammer groups are fully synchronized, the dual-loop AI model integrates the data. Due to the ground resistance of 3800kN < 5000kN, the angle of the eccentric block is dynamically adjusted to 15° (adjustment range 0°-30°, single adjustment range 3°) to optimize the eccentric torque. Continuous monitoring is carried out during pile driving to avoid polarization.

[0059] The outer loop of the dual-loop AI model, Transformer-TCN network, is based on historical data of soft soil strata and predicts the optimal phase difference Δθ=0.15° 2 seconds in advance. The inner loop PPO algorithm uses phase synchronization error e_θ, excitation force fluctuation rate e_F, and energy consumption increment ΔP as reward functions, and outputs hydraulic pump displacement correction Δu=±3mL / r, with phase synchronization accuracy controlled within ±0.4°.

[0060] Training data augmentation constructs a 5-dimensional observation vector O={e_θ, e_F, ΔP, a_rms, ψ}, improving sensitivity to the risk of "pile tilting" in soft soil pile driving; in federated learning deployment, the construction vessel only uploads gradients, while the central server aggregates and updates them, adapting to near-shore low-bandwidth scenarios.

[0061] The swaying of the ship caused the pile tilt angle to deviate by 0.4° (more than 0.3°). The AI ​​composite control algorithm prioritized adjusting the phase angle and excitation force of the two sets of hammers on the corresponding side of the ring-shaped hammer, generating a reverse correction torque of 0.8kN·m, which restored the pile to verticality within 0.5s.

[0062] Energy consumption optimization: The AI ​​composite control algorithm adjusts the hydraulic pump displacement to 145L / min and the drive motor speed to 1150r / min based on the real-time load (pile driving resistance 3800kN) and the hydraulic system efficiency model. The system works in the optimal efficiency range, and energy consumption is reduced by 18% compared with conventional control.

[0063] Shutdown control: When the pile reaches the design elevation (-40m), adjust the phase angle of the eccentric block to 0°. After the vibration disappears, synchronously reduce the speed of the drive motor to 300r / min to achieve shutdown.

[0064] The piling system corresponding to the control method in this embodiment includes: six sets of hydraulic vibratory hammers arranged in a ring, with an integrated ring mounting flange on the main body and adjustable spacing; an additional tilt sensor is installed on the outside of the vibratory box in the sensing unit (measurement range ±5°, accuracy ±0.01°); the domestically produced core components include a domestically produced axial piston pump (model CY14-1C, rated pressure 31.5MPa, volumetric efficiency 95.5%), and domestically produced electro-hydraulic proportional valves (model 4WREE8, control accuracy ±0.4% FS); the AI ​​control unit supports the phase synchronization algorithm for the ring-shaped hammer arrangement, and the human-machine interaction module displays the phase distribution heat map of each hammer group in real time. Testing showed that the system noise was stable at 68dB, and the peak tensile stress at the pile head was 1.8MPa, fully adaptable to the near-shore construction requirements in soft soil strata.

[0065] Example 4 This embodiment, based on the aforementioned embodiments, provides a multi-machine linkage hydraulic vibratory hammer piling system for a cross-sea bridge construction scenario in medium-hardness strata (5000kN≤strata resistance≤10000kN) under complex working conditions. The difference lies in: This embodiment is applied to the construction of main pile foundations for a cross-sea bridge (medium-hard stratum, stratum resistance 7500kN), and is suitable for medium-diameter piles and triangular hammer placement requirements. Its control method includes the following steps: Zero amplitude start-up: Set up 9 sets of hydraulic vibratory hammers in a triangular arrangement (side length 10m) and reset them. Adjust the eccentricity angle of the eccentric block to 0°, start the drive motor, and increase the speed to the rated speed of 1400r / min. There is no amplitude or excitation force.

[0066] Pile driving equipment operation and data acquisition: The sensing unit collects data in real time, including eccentric block rotation speed (1400 r / min), phase angle, hydraulic motor working flow rate (180 L / min), eccentric vibrator working temperature (52℃), pile body position and tilt angle (0.2°), and acceleration data; data fusion adopts Kalman filtering algorithm to filter out wave interference signals.

[0067] Pile driving process control: After each hammer group is synchronized, the dual-loop AI model integrates the data. Due to the ground resistance of 7500kN (5000-10000kN), the angle of the eccentric block is adjusted to 45° (adjustment range 30°-60°, single adjustment range 4°); continuous monitoring is carried out to avoid polarization.

[0068] The dual-loop AI model predicts the optimal phase difference Δθ = 0.25° 3 seconds in advance for the outer loop; the inner loop PPO algorithm outputs a hydraulic pump displacement correction Δu = ±4mL / r, with a phase synchronization accuracy of ±0.5°.

[0069] The digital twin model detects the formation's natural frequency f_s=15Hz, the system's natural frequency f_0=14Hz, and |f_s−f_0|=1Hz<2Hz. The excitation force is automatically reduced by 10% to avoid resonance, and the noise is controlled at 70dB.

[0070] When the noise level was detected to rise to 73dB (exceeding 72dB), the AI ​​composite control algorithm adjusted the rotation frequency of the eccentric block to 23Hz and the hydraulic system pressure to 28MPa. Combined with the vibration reduction structure design, the noise was reduced to 71dB within 300ms.

[0071] In the triangular cloth hammer, the phase difference of one set of hammers deviates by 0.8°. The electronically controlled pressure relief phase transient compensator commands the high-speed switching valve to depressurize the low-pressure chamber of the motor for 25ms, generating a braking torque of −80N・m, and pulling the phase back within 0.2s.

[0072] Shutdown control: When the pile reaches the design elevation (-55m), the phase angle of the eccentric block returns to zero, and the speed is simultaneously reduced to 300r / min before stopping.

[0073] The piling system in this embodiment includes: nine sets of hydraulic vibratory hammers arranged in a triangular pattern, with a triangular positioning flange at the bottom of the main frame; an accelerometer array newly installed on the side of the pile cap in the sensing unit to enhance the monitoring of lateral vibration of the pile; a synchronous motor in the hydraulic power unit with a three-outlet design of the same displacement (flow distribution error ±1.8%), a synchronous flow compensator with a displacement adjustment range of 0-550mL / r, and a response time of 18ms; and an AI control unit with a built-in triangular hammer phase synchronization calibration algorithm, supporting the switching function of the three-vertex hammer group as the reference host. The system has a 92% localization rate of core components, a total excitation force of 18000kN, and is suitable for the construction of steel pipe piles with a diameter of 10m and a wall thickness of 28mm, meeting the high-strength construction requirements of the main pile foundation of the cross-sea bridge.

[0074] Example 5 This embodiment, based on the aforementioned embodiments, provides a multi-machine linkage hydraulic vibratory hammer piling system for deep-sea extreme working conditions in hard strata (stratum resistance > 10000kN) under complex working conditions. The difference lies in: This embodiment is applied to the construction of deep-sea wind power foundations (hard strata, stratum resistance 12000kN), and is suitable for ultra-large diameter piles and extreme marine conditions (wind speed 15m / s, wave height 1.5m). The specific steps are as follows: Zero amplitude start-up: Set 12 sets of hydraulic vibratory hammers in a ring arrangement (hammer diameter 15m) and reset them. Adjust the eccentricity angle of the eccentric block to 0°, start the drive motor, and increase the speed to the rated speed of 1600r / min. There is no amplitude or excitation force.

[0075] Pile driving equipment operation and data acquisition: The sensing unit (acquisition frequency 1200Hz) collects data in real time, including eccentric block rotation speed 1600r / min, phase angle, hydraulic motor working flow 200L / min, eccentric vibrator working temperature 58℃, pile body posture tilt angle 0.5° and acceleration data; the tilt sensor monitors the pile body tilt angle change caused by the hull sway in real time.

[0076] Pile driving process control: After each hammer group is synchronized, the dual-loop AI model integrates the data. Due to the ground resistance of 12000kN>10000kN, the angle of the eccentric block is adjusted to 75° (adjustment range 60°-90°, single adjustment range 5°); continuous monitoring is carried out to avoid polarization.

[0077] The dual-loop AI model predicts the optimal phase difference Δθ = 0.3° 4 seconds in advance for the outer loop; the inner loop PPO algorithm outputs a hydraulic pump displacement correction Δu = ±6mL / r, with an initial phase synchronization accuracy of 0.7°.

[0078] After three consecutive PPO corrections, e_θ still = 0.7° > 0.6°. The system automatically switches to high-damping mode, simultaneously reducing the rotational speed of all hammers by 8% (to 1472 r / min), increasing the included angle of the eccentric blocks by 3° (to 78°), and controlling the peak tensile stress at the pile head at 2.0 MPa.

[0079] The swaying of the ship caused the pile to deviate by 0.6° in tilt angle. The AI ​​composite control algorithm adjusted the phase angle and excitation force of the four symmetrical sets of hammers in the ring-shaped hammer, generating a reverse correction torque of 1.2 kN·m, and restored the pile to verticality within 0.4 s.

[0080] A phase difference of 1.3° between adjacent hammers was detected. The electronically controlled pressure relief phase transient compensator depressurized for 35ms, generating a braking torque of −130N・m, and pulled the phase back to ±0.5° within 0.28s.

[0081] Energy consumption optimization: Adjusting the hydraulic pump displacement to 190L / min and the drive motor speed to 1550r / min reduces system energy consumption by 20% compared to imported equipment.

[0082] Shutdown control: When the pile reaches the design elevation (-70m), the phase angle of the eccentric block returns to zero, and the speed is simultaneously reduced to 300r / min before stopping.

[0083] The piling system corresponding to this embodiment includes: 12 sets of hydraulic vibratory hammers arranged in a ring; the main upper frame is made of high-strength alloy steel and integrates a wind and wave stabilizing structure; the electrically controlled pressure relief phase transient compensator adopts a dual high-speed switching valve redundant design (response time 15ms), the accumulator volume is 5L, and the working pressure is 15MPa; the AI ​​control unit has a built-in extreme working condition adaptive algorithm and supports automatic parameter calibration in high damping mode; the sensing unit has an enhanced design for resistance to salt spray and strong vibration (protection level IP68); the core components have a localization rate of 93%, the single hammer excitation force is 2200kN, the total excitation force is 26400kN, and it is suitable for ultra-large steel pipe piles with a diameter of 15m and a wall thickness of 35mm, meeting the construction needs of extreme working conditions in hard strata in deep sea.

[0084] The above embodiments of the invention focus on real-time dynamic control of multi-hammer linkage. They combine AI prediction with reinforcement learning, employing a dual-loop AI model and an electrically controlled pressure relief compensator to address phase deviations between multiple hammers, achieving a phase synchronization accuracy of ≤±0.6° for multi-hammer linkage. Furthermore, they introduce technologies such as formation natural frequency feedforward and high-damping mode switching to adapt to complex offshore conditions. This invention uses a soft synchronization method with a dual-loop AI model and an electrically controlled pressure relief compensator, eliminating the need for a rigid mechanical linkage structure, allowing for flexible hammer placement and adapting to distributed hammer placement requirements. This invention integrates an AI control unit and a sensing unit for real-time monitoring and dynamic adjustment of multiple parameters, including phase, flow rate, temperature, and tilt angle, achieving high-precision synchronous control under complex conditions. Simultaneously, it optimizes energy consumption, noise, and domestic compatibility, significantly improving synchronization accuracy, verticality control, and energy consumption. It is widely applicable to multi-machine hydraulic vibratory hammer systems driven by hydraulic motors, especially since the hydraulic pump displacement can be changed by adjusting the voltage, thereby controlling the coordinated operation of the vibratory hammers.

[0085] This invention enables coordinated control of any number of hydraulic vibratory hammers forming a pile driving system without the need for additional rigid mechanical linkage control structures. The invention employs a soft synchronization method of "automatic coarse synchronization of synchronous motors + electronic micro synchronization of synchronous flow compensators," eliminating the need for any rigid linkage structures between the vibratory hammers. The synchronous operation of two or more sets of vibratory hammers can be achieved solely through the control and power systems, significantly improving the flexibility of hammer placement and adaptability to various working conditions.

[0086] The control method and piling system provided by this invention facilitate intelligent expansion and digital management. They adopt a distributed and collaborative control architecture, which can integrate remote diagnostics, data storage and analysis functions, providing a technical foundation for subsequent visualization of the construction process, equipment health management and intelligent scheduling. This invention also significantly reduces the risk of damage and maintenance costs of hydraulic vibratory hammers, simplifies the overall layout, and reduces the frequency of lubrication, maintenance and replacement, thus reducing maintenance manpower and spare parts consumption, and lowering long-term operating costs.

[0087] The key feature of the embodiments of this invention is the integration of a multi-hammer linkage mechanical-hydraulic-electric control three-synchronization system, combined with an intelligent adaptive control algorithm, to achieve domestic adaptation of core components. This solves bottleneck problems such as multi-hammer phase synchronization accuracy control, energy loss suppression under high-frequency operating conditions, and improved reliability of core hydraulic components. The engineering prototype of this invention can achieve multi-hammer linkage phase synchronization accuracy within ±0.8°, excitation force exceeding 15000kN, and is suitable for ultra-large steel pipe piles with a diameter of over 10m. The domestic production rate of core hydraulic pumps and valves reaches 90%, energy consumption is reduced by 15% compared to imported equipment, and noise is controlled below 75 decibels. The overall performance is significantly improved, reaching the international advanced level.

[0088] The technical solutions provided in the above embodiments of the present invention are system integration and innovative design that combine multiple technical fields such as intelligent control, process control, mechanical design, and electronic engineering. The prototype of the present invention has undergone actual testing in a confidential site. By continuously collecting on-site feedback data and iteratively upgrading its AI composite control algorithm, and by optimizing the structural design multiple times, the overall performance of the equipment and user satisfaction have been improved. This allows it to better balance automation and intelligent control, simultaneously improving control accuracy, structural strength, and safety, and meeting the needs of modern engineering construction under complex working conditions in various scenarios of offshore pile foundation construction.

[0089] The above are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention; any substitutions and improvements made without departing from the concept of the present invention shall fall within the scope of protection of the present invention.

Claims

1. A pile driving control method for multi-machine linkage of hydraulic vibratory hammers under complex working conditions, characterized in that, Includes the following steps: S1, Zero Amplitude Start Set up a multi-machine linkage hydraulic vibratory hammer pile driving equipment and reset it. Before starting the equipment, adjust the eccentricity angle of the paired eccentric blocks to zero through the eccentric block adjustment mechanism, start the drive motor, and make the eccentric blocks rotate to the rated speed without generating amplitude or excitation force. S2. Operation and Data Acquisition of Pile Driving Equipment Power on the piling equipment and collect the operating data of each hammer group in real time through the sensing unit of the piling equipment. The operating data includes the rotational speed of the eccentric block, phase angle, hydraulic motor working flow, eccentric vibrator working temperature, pile body position and inclination angle, and acceleration data. S3, Pile Driving Process Control Once all hammer groups are fully synchronized, a dual-loop AI model is used to fuse the collected operational data. Based on the on-site geological conditions and real-time pile driving resistance data, the angle of the eccentric block is dynamically adjusted to achieve the optimal value of the eccentric torque required for pile driving. During the pile driving process, the angle of the eccentric block, the bearing status, and the verticality of the pile body of each hammer group are continuously monitored through the sensing unit to avoid polarization phenomena. When any eccentric block is detected to have a speed deviation or angle deviation, it is automatically corrected by the AI ​​composite control algorithm to make the eccentric blocks of each hammer group reach the same phase angle, so that the phase synchronization accuracy of multiple hammer groups is ≤±0.6°; the AI ​​composite control algorithm is a composite algorithm that combines a dual-loop AI model with a neural network adaptive PID algorithm; S4, Shutdown Control When the pile reaches the design elevation, the phase angle of the eccentric block is adjusted to zero, the vibration disappears, and then the machine is stopped by synchronously reducing the speed of the drive motor.

2. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions as described in claim 1, characterized in that, The dual-loop AI model comprises: an outer loop, a prediction model based on digital twins, which uses on-site geology, pile type, and historical construction data as inputs, and trains an offline Transformer-TCN hybrid network to predict the optimal phase difference setpoint Δθ 0–5 seconds in advance; and an inner loop, an online reinforcement learning fine-tuning model that uses a near-end policy optimization (PPO) algorithm in the neighborhood of the predicted value Δθ, with phase synchronization error e_θ, excitation force fluctuation rate e_F, and energy consumption increment ΔP as the immediate reward function, and outputs the hydraulic pump displacement correction amount Δu in real time.

3. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions as described in claim 2, characterized in that, In the dual-loop AI model, the Transformer-TCN hybrid network consists of two Transformer encoder layers (each with four attention heads) and two TCN layers (inflation coefficient 2), with ReLU as the activation function; the discount factor γ of the PPO algorithm is 0.99, the learning rate η is 0.003, and the batch size is 64. The neural network adaptive PID algorithm has the following input layers: total error E and error rate of change EC = dE / dt; two hidden layers with 10 nodes each; ReLU activation function; and PID parameter adjustment values ​​at the output layer. ΔKp, ΔKi, Δ Kd The PID parameter update formula is: Kp(t) = Kp(t-1) + η × ΔKp Ki (t) = Ki (t-1) + η × ΔKi Kd(t) = Kd(t-1) + η × ΔKd Where η is the learning rate, which ranges from 0.01 to 0.05 and is determined through experimental calibration.

4. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions as described in claim 3, characterized in that, In step S3, the calculation of the total error E includes: Master-slave traffic difference eq=Q ref -Q i 、 Master-slave phase difference eθ=θ ref -θ i Master-slave temperature difference eT=T ref -T i and pile inclination angle difference eα=α ref -α i ， The error weight coefficients are dynamically optimized using reinforcement learning algorithms. wQ, wθ, wT, wα The formula for calculating the total error E is: E (i)=wQ×|eq|+wθ×|eθ|+wT×|eT|+wα×|eα| in, Q ref For reference host traffic, Q i For the first i Slave data traffic; θ ref For reference host phase angle, θ i For the first i Phase angle of slave unit; T ref For reference host temperature, T i For the first i Slave temperature; α ref The target inclination angle of the pile body. α i This represents the real-time inclination angle of the pile.

5. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions as described in claim 4, characterized in that, Step S3 further includes the following steps: S3-1 Training Data Augmentation: Constructing a 5-dimensional observation vector using vibration acceleration-time curves as additional state variables. O= {e_θ, e_F, ΔP, a_rms, ψ} ,in a_rms The root mean square of the acceleration, ψ To reduce the vibration coefficient and improve the model's sensitivity to the risk of polarization-pile head tearing; S3-2 Federated Learning Deployment: Each construction vessel only uploads gradients, while the central server aggregates and updates them, ensuring continuous model evolution in low-bandwidth maritime scenarios. S3-3 Compensates for the lag in elastic deformation of mechanical components: An electrically controlled pressure relief phase transient compensator is added to the mechanical synchronization of the angular gearbox-universal coupling. When the phase difference between adjacent hammers is detected to be >1°, the controller commands the high-speed switching valve to instantaneously depressurize the low-pressure chamber of the corresponding motor for 20–40ms, generating a braking torque of -50~-150N·m, and pulling back the phase within 0.3s. S3-4 Formation natural frequency feedforward: The digital twin model performs modal matching between the formation natural frequency f_s and the system natural frequency f_0. If |f_s−f_0|<2Hz, the excitation force is automatically reduced by 10% to avoid resonance causing pile floating or boom fatigue, and the noise is ≤72dB. S3-5 High Damping Mode Switching: If after 3 consecutive PPO corrections e_θ If the angle is still >0.6°, switch to high damping mode, simultaneously reduce the rotational speed of all hammers by 8%, and increase the included angle of the eccentric blocks by 3° to ensure that the peak tensile stress at the pile head is ≤2.1MPa; S3-6 Instruction Output and Execution: Calculation for Multiple Slave Units u i The instruction output module will UI The signal is converted into an electrical signal command, which is transmitted to the synchronous flow compensator through the control output module to dynamically adjust the hydraulic pump displacement and ultimately achieve precise phase synchronization of multiple machines.

6. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions as described in claim 1, characterized in that, In step S3, for hull sway compensation under marine conditions, the tilt angle data of the pile body is collected in real time by tilt sensors. When the tilt angle deviation exceeds 0.3°, the AI ​​composite control algorithm prioritizes adjusting the phase angle and excitation force of the corresponding side hammer group to generate a reverse correction torque, so that the pile body returns to a vertical state. In step S3, the eccentric block angle adjustment strategy is dynamically set based on the stratum resistance level. When the stratum resistance is less than 5000kN, the eccentric block angle adjustment range is 0°-30°; when the stratum resistance is 5000kN-10000kN, the adjustment range is 30°-60°; when the stratum resistance is greater than 10000kN, the adjustment range is 60°-90°, and each adjustment increment does not exceed 5°.

7. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions according to claim 1, characterized in that, Step S3 also includes an energy consumption optimization step: the AI ​​composite control algorithm dynamically adjusts the hydraulic pump displacement and drive motor speed based on real-time load data and hydraulic system efficiency model, so that the system works in the optimal efficiency range. Including noise control steps: by monitoring the operating noise data of the vibratory hammer, when the noise exceeds 72 decibels, the rotation frequency of the eccentric block and the pressure of the hydraulic system are adjusted, and combined with the vibration reduction structure design, the noise is controlled below 75 decibels.

8. The pile driving control method for multi-machine linkage of hydraulic vibratory hammer under complex working conditions as described in claim 1, characterized in that, The sensing unit includes a rotary encoder, a hydraulic flow sensor, a temperature sensor, an accelerometer array, and a tilt sensor. The rotary encoder is installed at the end of the shaft of the eccentric vibrator, the hydraulic flow sensor is installed in the oil outlet of the hydraulic motor, the temperature sensor is installed on the outer end cover of the eccentric vibrator, the accelerometer array is installed at the vibrator, bearing, and pile cap, and the tilt sensor is installed at the top of the pile. The data fusion processing adopts the Kalman filter algorithm to filter, reduce noise, and complement the raw signals collected by each sensor.

9. A multi-machine linkage piling system for hydraulic vibratory hammers under complex working conditions, characterized in that, It includes: The piling unit includes multiple sets of hydraulic vibratory hammers arranged in parallel below the main body's upper hanger. The top is rigidly connected to the mounting flange at the bottom of the hanger by a set of high-strength bolts, and the bottom is clamped by a pile clamp to hold the steel pipe pile to be constructed. The main mechanical structure of each set of hydraulic vibratory hammers consists of an outer vibration damping frame, a vibration box, and a pile clamp. The vibration box is equipped with an eccentric vibrator driven by a hydraulic motor. The sensing unit includes a rotary encoder, a hydraulic flow sensor, a temperature sensor, an inclination sensor, and an accelerometer array, which are installed at corresponding positions on the piling unit to collect operating data. The AI ​​control unit, specifically an edge computing controller deploying the dual-loop AI model, includes a signal input module, a central controller, a control output module, a digital twin module, a sensing unit, and a human-machine interaction module. The signal input module receives data collected by the sensing unit. The central controller incorporates the AI ​​composite control algorithm to process the data and output control commands. The control output module sends control commands to the synchronous motor and synchronous flow compensator. The digital twin module is used for offline training of the Transformer-TCN network. The human-machine interaction module is used for synchronous parameter configuration and visualization of multi-machine operating status. The hydraulic power unit includes a hydraulic vibratory hammer power station, a synchronous motor, and a synchronous flow compensator; The hydraulic vibratory hammer power station provides basic hydraulic and electrical energy to the system. The synchronous motor is integrated into the outlet oil circuit of the power station to achieve coarse synchronous control of the initial flow distribution of multiple hydraulic oil circuits. The synchronous flow compensator is arranged in parallel with the synchronous motor to provide hydraulic power for closed-loop fine adjustment and achieve micro-synchronous correction. The verticality monitoring and correction module unit is connected to the tilt sensor and AI control unit to receive pile position and tilt angle data in real time and output correction commands through the AI ​​control unit. It includes a synchronous motor, a synchronous flow compensator, and an electrically controlled pressure relief phase transient compensator. The electrically controlled pressure relief phase transient compensator is a combination of a high-speed switching valve and a low-pressure accumulator, with a response time ≤20ms, an accumulator volume of 5L, and an operating pressure of 10–15MPa.

10. The multi-machine linkage piling system for hydraulic vibratory hammers under complex working conditions according to claim 9, characterized in that, The rotary encoder is installed at the end of the shaft of the eccentric vibrator, the hydraulic flow sensor is installed in the oil outlet of the hydraulic motor, the temperature sensor is installed on the outer end cover of the eccentric vibrator, the tilt sensor is installed on the top of the pile and the outside of the vibration box, and the accelerometer array is installed at the vibrator, bearing and pile cap. The acquisition frequency of each sensor is not less than 1000Hz.

Images