Bucket type foundation air floatation dynamic deviation correction installation method and system

Abstract

The invention relates to the technical field of ocean engineering, and discloses a barrel type foundation air floatation dynamic deviation correction installation method and system.The method comprises the following steps that real-time attitude information of a barrel type foundation is obtained; predicting external hydrodynamic disturbance acting on the bucket foundation; generating a cooperative control instruction based on the real-time attitude information and the predicted external hydrodynamic disturbance; and according to the cooperative control instruction, an internal dynamic torque compensation system is driven to offset high-frequency dynamic inclination caused by external hydrodynamic disturbance, and a partition air pressure control system is driven to correct static or low-frequency inclination of the bucket foundation. The self-adaptive model predictive control algorithm is applied, changes of system kinetic parameters can be identified online, the control model is updated in real time, the control strategy can be kept optimal all the time, and the defect that in the prior art, a fixed parameter controller is adopted, and system reliability is high is overcome. And in the sinking process of the barrel foundation, the control performance is sharply reduced or even fails due to the change of working conditions.

Description

Technical Field

[0001] The present invention relates to the field of marine engineering technology, and in particular to a barrel-type foundation air flotation dynamic deviation correction installation method and system. Background Art

[0002] Precast concrete bucket foundations for water transport projects can weigh up to 5,000-5,500 tons, making them difficult to hoist and install. Air flotation transport, degassing, and positioning techniques are typically employed. Traditional bucket foundation installation methods rely on passive control strategies in air flotation environments, where tilt is detected and then leveled by inflating or deflating the bucket. Existing systems primarily adjust the air pressure within each chamber of the bucket to offset external hydrodynamic disturbances. This approach is often slow to respond and cannot react promptly to high-frequency disturbances, resulting in unstable foundation posture.

[0003] Existing technologies rely on a single sensor for attitude monitoring, which does present numerous challenges. For example, inertial measurement units are susceptible to drift and cannot consistently provide accurate, real-time attitude data. This clearly lacks reliability for precise operations requiring complex water flow.

[0004] Furthermore, existing inflation and deflating systems often lack flexibility. Static pressure settings make the system less adaptable to environmental changes. When sea conditions are severe or there are sudden dynamic disturbances, the pressure level cannot be adjusted in a timely manner, which directly leads to excessive tilt in the basic attitude and may even cause capsizing.

[0005] Looking at energy management, many systems waste kinetic energy generated during braking as heat during control. This reduces energy efficiency and impacts the economic viability of the system.

[0006] Furthermore, some control systems, such as basic PID control, employ simple algorithmic models that lack adaptability to complex environments. Consequently, faced with constantly changing external conditions and underlying states, the systems often become uncontrollable. While such systems may operate well under ideal conditions, their stability and reliability are significantly compromised in the harsh marine environment.

[0007] Therefore, existing technologies have many deficiencies in dynamic control, attitude tracking, and energy efficiency management. To overcome these limitations, current technical solutions urgently need a more flexible, efficient, and intelligent control system to improve the safety and stability of bucket foundations in complex marine environments. This is precisely the problem addressed by this invention. Summary of the Invention

[0008] In response to the shortcomings of the existing technology, the present invention provides a barrel-type foundation air flotation dynamic correction installation method and system, which solves the problems of delayed response of traditional control systems to high-frequency dynamic disturbances, insufficient attitude monitoring accuracy and low energy efficiency utilization in complex marine environments.

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: A barrel-type foundation air flotation dynamic deviation correction installation method, comprising the following steps: S1: Acquire real-time posture information of the bucket foundation; S2: Predict the external hydrodynamic disturbance acting on the bucket foundation; S3: generating a coordinated control instruction based on the real-time posture information and the predicted external hydrodynamic disturbance; S4: According to the collaborative control instruction, drive the internal dynamic torque compensation system to offset the high-frequency dynamic tilt caused by the external hydrodynamic disturbance, and drive the partitioned air pressure control system to correct the static or low-frequency tilt of the bucket foundation.

[0010] Preferably, the real-time posture information of the bucket foundation is obtained as follows: High-frequency attitude data is acquired through an inertial measurement unit, low-frequency attitude data is acquired through a global navigation satellite system, and a Kalman filter algorithm is used to fuse the high-frequency and low-frequency attitude data to obtain the real-time attitude information.

[0011] Preferably, the predicted external hydrodynamic disturbance acting on the bucket foundation is specifically: The upstream water flow information is measured in real time by an acoustic Doppler flow profiler arranged on the incoming flow surface of the bucket-type foundation, and a prediction model is established based on the water flow information to obtain the predicted external hydrodynamic disturbance.

[0012] Preferably, the generating of the collaborative control instruction is specifically: A model predictive control algorithm is used to take the real-time posture information and the predicted external hydrodynamic disturbance as input within a rolling finite time window, and the collaborative control instructions are generated by solving a constrained optimization problem.

[0013] Preferably, the system dynamics model used in the model predictive control algorithm can perform online parameter identification and adaptive update based on the error between the real-time posture information of the bucket foundation and the model prediction output.

[0014] Preferably, the internal dynamic torque compensation system includes a compensation mass block arranged inside the barrel-type foundation and a servo actuator driving the compensation mass block to move at high speed; The driving internal dynamic torque compensation system specifically includes: controlling the servo actuator to drive the compensation mass block to generate an inertia torque opposite to the high-frequency dynamic tilt direction.

[0015] Preferably, when the servo actuator performs deceleration braking on the compensating mass block, the kinetic energy is converted into electrical energy through a regenerative braking function and stored in an energy recovery unit.

[0016] Preferably, before obtaining the real-time posture information, the system is initialized, and the initialization is specifically as follows: The initial position of the compensation mass block in the internal dynamic moment compensation system is fine-tuned before installation to calibrate and compensate for the initial unbalanced center of mass of the bucket foundation.

[0017] Preferably, the real-time posture information, external hydrodynamic disturbance, and collaborative control instruction data are integrated to generate a digital twin file for evaluating installation quality.

[0018] A barrel-type foundation air flotation dynamic deviation correction installation system, comprising: An information acquisition module, used to obtain real-time posture information of the bucket foundation; a disturbance prediction module for predicting external hydrodynamic disturbances acting on the bucket foundation; a collaborative control module, configured to generate collaborative control instructions based on the real-time posture information and the predicted external hydrodynamic disturbance; An execution module is connected to the collaborative control module, and the execution module includes: an internal dynamic torque compensation system; a zoned air pressure control system; The execution module is used to drive the internal dynamic torque compensation system to offset high-frequency dynamic tilt and drive the partitioned air pressure control system to correct static or low-frequency tilt according to the coordinated control instruction.

[0019] The present invention provides a method and system for dynamic deviation correction installation of a barrel-type foundation with air flotation, which has the following beneficial effects.

[0020] 1. This invention uses an acoustic Doppler current profiler and a predictive model to achieve advanced prediction of external hydrodynamic disturbances. This feedforward control mechanism enables the system to respond before disturbances arrive. Existing technologies typically passively wait for tilt to occur before making corrections, resulting in a naturally delayed response and difficulty in effectively controlling attitude in complex water flows.

[0021] 2. The present invention innovatively sets up an internal dynamic torque compensation system and a partitioned air pressure control system. The two are uniformly scheduled by a collaborative control algorithm with clear division of labor. High-frequency dynamic tilt is instantaneously offset by the internal system, and low-frequency static tilt is stably corrected by the air pressure control system. Compared with the existing technology that only relies on a single air pressure control system for correction, the present invention solves its fundamental defects of slow response speed and inability to cope with high-frequency dynamic disturbances.

[0022] 3. The present invention applies an adaptive model predictive control algorithm, which can identify changes in system dynamic parameters online and update the control model in real time, so that the control strategy can always remain optimal. It overcomes the problem of the existing technology using fixed parameter controllers, which causes the control performance to drop sharply or even fail due to changes in working conditions during the sinking process of the bucket base.

[0023] 4. The present invention adopts a multi-sensor fusion solution based on Kalman filtering, which integrates the high dynamic characteristics of the inertial measurement unit with the long-term stability of the global navigation satellite system to obtain high-precision, drift-free real-time attitude information. This solution provides a reliable data basis for accurate decision-making of the entire control system, avoiding the problem of excessive final installation error caused by insufficient accuracy or drift of a single sensor in the existing technology.

[0024] 5. The present invention integrates an energy recovery unit into the internal dynamic torque compensation system. During the braking process, the servo actuator converts kinetic energy into electrical energy and stores it for use in the next acceleration. Compared with the prior art method of dissipating braking energy in the form of heat energy, the present invention significantly reduces the system energy consumption and peak power, and improves the system's operating endurance and economy. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 Schematic diagram of the method steps of the present invention; Figure 2 Schematic diagram of the overall system architecture of the present invention; Figure 3 Schematic diagram of the internal dynamic torque compensation system of the present invention; Figure 4 This is a schematic diagram of the zoned air pressure control system of the present invention; Figure 5 Schematic diagram of the control algorithm flow of the present invention; Figure 6 It is a schematic structural diagram of the compensation mass block of the present invention and a servo actuator that drives the compensation mass block to move at high speed. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the present specification. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0027] Please see the attached Figure 1-5 ,in Figure 4 This is just an illustration. There may be several actual chambers, and each chamber executes the same operating logic as chamber 1.

[0028] An embodiment of the present invention provides a barrel-type foundation air flotation dynamic deviation correction installation method, comprising the following steps: S1: Acquire real-time posture information of the bucket foundation; Specifically, the system's physical structure includes an information acquisition module, illustratively integrated on top of or within the bucket foundation. The information acquisition module includes an inertial measurement unit (IMU), a global navigation satellite system (GNSS) receiver, and a data fusion processor. Both the IMU and the GNSS receiver are communicatively connected to the data fusion processor.

[0029] For example, the inertial measurement unit (IMU) is preferably an industrial-grade fiber optic or micro-electromechanical system (MEMS) IMU, mounted near the geometric center of the bucket foundation's top head. This IMU measures and outputs raw data on the bucket foundation's three-axis angular velocity and acceleration at a high update rate. This high-frequency characteristic ensures that any instantaneous, high-frequency attitude changes caused by external disturbances such as water impact can be captured.

[0030] For example, the GNSS receiver preferably supports real-time dynamic differential (RTD) technology, with its antenna also mounted on top of the barrel foundation to ensure high-quality signal reception. This GNSS receiver can provide absolutely accurate three-dimensional position, velocity, and heading information at a relatively low update frequency. The attitude information it outputs is free of drift errors accumulated over time, resulting in high absolute accuracy.

[0031] The data fusion processor addresses the technical limitations of a single sensor: the drift error accumulated over time in the inertial measurement unit (IMU), while the global navigation satellite system (GNSS) update frequency is too low to meet real-time control requirements. To this end, the processor embeds and runs a data fusion algorithm to combine the strengths of both sensors.

[0032] In this embodiment, the data fusion algorithm uses an extended Kalman filter algorithm. This algorithm establishes the rocking motion of the bucket foundation as a nonlinear system model and achieves optimal fusion of high-frequency IMU data and low-frequency global navigation satellite system data through a continuous prediction-update cycle.

[0033] First, establish the state vector of the system , which contains the physical quantities that need to be estimated in real time. The state vector is exemplarily defined as: ; In the formula, They represent the attitude inclination angles of the bucket foundation around the X axis (roll) and Y axis (pitch), Respectively represent the angular velocity of the bucket base around the X-axis and Y-axis, Represents a transpose operation.

[0034] The execution flow of the extended Kalman filter algorithm includes the following steps: The first step is the prediction step. This step uses the optimal state estimate from the previous moment and the current IMU measurement to predict the current state. This step is performed continuously at a high frequency (e.g., 100 Hz) of the IMU.

[0035] The state prediction equation is: ; In the formula, Indicates at time The predicted value of the state, based on the time Information, Indicates at time The estimated value of the state, Indicates at time The control input or manipulated quantity, Represents a function.

[0036] The prediction covariance equation is: ; In the formula, Indicates the state at time The covariance matrix of Information, State transition matrix, Indicates at time The state covariance matrix of The uncertainty estimate of the state is based on the information of State transition matrix The transpose of Process noise covariance matrix.

[0037] The second step is the update step. This step is triggered when low-frequency measurement data is received from the GNSS. This high-precision absolute measurement value is used to correct the drift generated in the prediction step.

[0038] First calculate the Kalman gain : ; In the formula, Indicates at time The Kalman gain, Indicates the state at time The covariance matrix of Information, Indicates at time The observation matrix, Representation matrix The transpose of Indicates at time The observation noise covariance matrix, This is a summed matrix that represents the uncertainty in the measurement predictions.

[0039] The state estimate is then updated to obtain the time The updated state estimate vector : ; In the formula, Indicates at time The updated state estimate vector, Indicates at time The predicted value of the state, based on the time Information, Indicates at time The Kalman gain, Indicates at time The observed measurement value, represents the observation model, where The state estimation Mapped into a nonlinear function of the observation space, Represents the observation residual, that is, at time Observed measurements and predicted measurements The difference between is the observation function, which maps the predicted pose in the state vector to the measurement space.

[0040] Finally update the state estimation covariance matrix : ; In the formula, Indicates at time The state covariance matrix of Observation information, The identity matrix, Indicates at time The Kalman gain, The observation matrix, Indicates the state at time The covariance matrix of information.

[0041] By executing the above steps in a loop, the data fusion processor can output a time in real time. The updated state estimate vector The attitude information contained in this vector not only maintains the high update rate of the IMU, but also eliminates the accumulated drift through the periodic correction of the GNSS, thereby obtaining a high-precision, high-frequency, drift-free bucket-based real-time attitude information.

[0042] This method of obtaining the real-time posture information of the bucket foundation can provide stable and reliable data input for the subsequent model predictive control algorithm. It is a necessary prerequisite for achieving high performance of the entire closed-loop control system and has the technical effect of ensuring that the control system can accurately perceive the foundation posture.

[0043] S2: Predict the external hydrodynamic disturbance acting on the bucket foundation; Specifically, the purpose of this step is to enable the control system to have feedforward control capabilities and to actively respond to upcoming environmental disturbances, which is achieved through a disturbance prediction module.

[0044] The disturbance prediction module includes an acoustic Doppler current profiler and a prediction model processor. The acoustic Doppler current profiler is in communication with the prediction model processor to transmit measurement data to the processor for analysis and prediction.

[0045] For example, the acoustic Doppler current profiler is mounted on the outer wall of the bucket foundation's incoming flow surface. This mounting position enables it to continuously and contactlessly measure a certain distance ahead of it. More specifically, its output is a three-dimensional water velocity vector at multiple discrete depth units.

[0046] The prediction model processor receives and processes the velocity data measured by the acoustic Doppler current profiler and runs a prediction algorithm to output real-time, rolling predictions of future disturbances. This processing flow further includes three consecutive sub-steps: data preprocessing, hydrodynamic torque conversion, and time series prediction.

[0047] First, in the data preprocessing substep, the prediction model processor filters the raw velocity vector data received from the acoustic Doppler current profiler to remove measurement noise. For example, a median filter or Kalman filter can be used to improve the signal-to-noise ratio of the data. The filtered velocity vectors are then converted from the acoustic Doppler current profiler's coordinate system to the bucket foundation's body coordinate system to unify the reference for subsequent calculations.

[0048] Next, in the hydrodynamic torque conversion substep, the prediction model processor converts the preprocessed upstream flow velocity profile data into the total disturbance torque that will act on the bucket foundation. This conversion is based on fluid mechanics principles, such as the Morrison equation, which calculates the force exerted by the fluid on the slender rod.

[0049] The hydrodynamic force acting on a cylinder per unit length It can be expressed as a linear superposition of resistance and inertia. By dividing the submerged portion of the bucket foundation into multiple micro-segments along the depth direction and using the water velocity at the corresponding depth measured by the acoustic Doppler current profiler, the hydrodynamic force acting on each micro-segment is calculated. Finally, by integrating the hydrodynamic forces generated by all micro-segments along the entire submerged depth and multiplying them by the corresponding lever arm, the total external hydrodynamic disturbance torque acting on the center point of the bucket foundation's swing can be obtained. . This calculation process is repeatedly executed at a high frequency in the processor to form a time series of disturbance torque. Again, in the time series prediction sub-step, the prediction model processor analyzes and predicts the generated hydrodynamic torque time series. In this embodiment, it is preferred to use the autoregressive integrated moving average model for prediction because it can effectively capture the randomness and time correlation presented by waves and water flow. Before the model is applied, it is first necessary to identify the model to determine its optimal order. This process may include performing a stationarity test on historical data, such as using an augmented Dickey-Fuller test. Then, by analyzing the autocorrelation function and partial autocorrelation function graph of the sample data, the autoregressive order of the model is preliminarily determined. and the moving average order To obtain the optimal model, we can further use the Akaike Information Criterion or the Bayesian Information Criterion to evaluate and select the order combination that minimizes the criterion function value. .

[0050] A general ARIMA The mathematical expression of the model is: ; In the formula, Indicates time The predicted or estimated value of The original disturbance torque time series is After order difference at time The value of yes The predicted value of and are the autoregressive and moving average coefficients of the model, is the model order identified above, represents the difference order, is a constant term, is the white noise error term, Expressing the past time point The weighted sum of Expressing the past The weighted sum of the error (white noise) terms at each time point.

[0051] In actual operation, the model is predicted in a rolling manner. That is, in each control cycle, the model is updated using the latest disturbance torque observation value, and the disturbance torque sequence within a finite prediction time domain N in the future is predicted forward, which is recorded as ,in To adapt to the slowly changing sea conditions, the parameters of the model can also be set to be re-identified and updated periodically.

[0052] Finally, the predicted disturbance torque sequence is transmitted in real time to the subsequent collaborative control module, serving as a key input for feedforward control and rolling optimization. This implementation, through real-time measurement of upstream water flow and advanced prediction of disturbance torque, transforms the entire dynamic correction installation system from a traditional passive response to an active defense. This significantly enhances the system's ability to withstand sudden disturbances, effectively reduces attitude control deviations, and thus ensures the accuracy and safety of the installation process.

[0053] S3: generating a coordinated control instruction based on the real-time posture information and the predicted external hydrodynamic disturbance; Specifically, this step is performed by a collaborative control module, which serves as the decision-making core of the system and aims to generate an optimal control strategy that can simultaneously cope with high-frequency dynamic disturbances and low-frequency static tilt.

[0054] The collaborative control module, illustratively, is a high-performance industrial computing platform. This platform logically serves as the system's central processing unit (CPU). Its input is connected to the information acquisition module and the disturbance prediction module to receive real-time status and future disturbance information; its output is connected to the execution module to send control instructions.

[0055] The core algorithm of this module is preferably an adaptive model predictive control algorithm. This algorithm is adopted because its predictive control framework can naturally utilize the feedforward information provided by the disturbance prediction module, and its optimization-based approach can explicitly handle the various physical constraints in the system.

[0056] The implementation of the adaptive model predictive control algorithm first requires the establishment and maintenance of a mathematical model that accurately describes the dynamic behavior of the bucket foundation. This model is exemplarily represented in a discrete-time state space form: ; In the formula, Indicates at time The system state vector, Indicates at time The state transition matrix, Indicates at time The control input matrix, Indicates at time The current state vector, Indicates at time The control input, represents the disturbance influence matrix, which is used to describe the response of the system to external disturbances. Indicates at time The disturbance influence vector.

[0057] The cooperative control instruction vector It embodies the cooperative control concept of the present invention. It is a composite vector, which includes high-frequency control components for controlling the internal dynamic torque compensation system. , and low-frequency control components for controlling zone air pressure control systems .

[0058] In order to deal with the time-varying dynamic characteristics of the bucket foundation due to factors such as soil depth and soil quality changes during the sinking process, the collaborative control module further includes an online parameter identification unit. This unit ensures that the parameter matrix in the state space model and Capable of online adaptive updates.

[0059] The online parameter identification unit exemplarily uses a recursive least squares algorithm with a forgetting factor. The algorithm continuously compares the predicted state of the model with the actual state output by the information acquisition module and uses the prediction error between the two to recursively correct the model parameters.

[0060] The specific implementation of the recursive least squares algorithm is to first rewrite the state space equation into a linear regression form. Then, in each control cycle, the vector containing the model parameters is adjusted according to the newly acquired data. The core update law can be described by the following formula group: ; ; ; In the formula, Indicates at time The gain matrix, Indicates at time The error covariance matrix of Indicates at time Matrix The transpose of is a scalar constant, reflects the combined effect of measurement uncertainty and state uncertainty. Indicates at time The updated estimate of Indicates at time Estimates of states or parameters, Indicates at time The current observation or measurement data of Represented by the observation matrix For the previous estimate Convert the predicted value to represents the measurement residual, Indicates at time The updated error covariance matrix, Used to adjust the size of the covariance, Represents the previous covariance The updated result is subtracted by the gain matrix and the measurement matrix Uncertainty after the impact.

[0061] Then, in each control cycle, the collaborative control module constructs and solves a finite horizon rolling optimization problem based on the updated adaptive model. This problem aims to calculate the future prediction horizon. The objective function of this optimization problem is It can be expressed as: ; In the formula, Represents the objective function to be minimized , represents the objective function, represents the control input sequence, represents the prediction time domain, Indicates at time The state prediction, Indicates the reference state value, represents the weighted sum of squares of state errors, Indicates at time The control input, represents the weighted sum of squares of the control inputs.

[0062] The solution process of this optimization problem is subject to a series of physical constraints, which further include state constraints, input constraints, and input change rate constraints.

[0063] This constrained optimization problem is usually a quadratic programming problem. The collaborative control module integrates an efficient, real-time system-suitable quadratic programming numerical solver to quickly solve the problem online in each control cycle and obtain the optimal future control input sequence. .

[0064] Finally, according to the rolling horizon principle of model predictive control, the system does not execute the entire calculated optimal control sequence. Instead, it only extracts the first element of the sequence, i.e. , as the final collaborative control instruction generated at the current moment. The instruction vector is then sent to the execution module for decomposition and physical implementation.

[0065] This implementation utilizes adaptive model predictive control, enabling forward-looking decisions based on future predictions and online adaptation to system changes, while respecting physical constraints. This approach generates optimal and safe coordinated control instructions, achieving precise and efficient coordinated control of high- and low-frequency disturbances.

[0066] S4: according to the coordinated control instruction, driving the internal dynamic torque compensation system to offset the high-frequency dynamic tilt caused by the external hydrodynamic disturbance, and driving the zoned air pressure control system to correct the static or low-frequency tilt of the bucket foundation; Specifically, this step is completed by a physically integrated execution module, which is responsible for efficiently and accurately converting the abstract digital instructions generated by the collaborative control module into physical actions that can stabilize the bucket-type basic posture.

[0067] The execution module physically includes an internal dynamic torque compensation system and a partitioned air pressure control system. The input end of the execution module is connected to the output end of the cooperative control module to receive the cooperative control instructions generated in real time.

[0068] The execution process of this step first includes the collaborative control instruction The instruction is a composite vector, in which the high-frequency and low-frequency control tasks have been optimally allocated by the model predictive control algorithm. An instruction distribution unit, which can be a software function of the collaborative control module or a front-end controller of the execution module, decomposes the instruction vector into high-frequency control instructions. and low-frequency control instructions . Subsequently, the high frequency control instruction The internal dynamic torque compensation system is typically installed in a dry chamber on top of the barrel foundation. Its core consists of a compensation mass made of a high-density alloy and a servo actuator consisting of two orthogonally arranged, high-acceleration linear servo motors.

[0069] The high-frequency control instructions For example, it is a two-dimensional vector that specifies the target acceleration of the compensation mass block in the XY plane. After receiving the instruction, the controller of the servo actuator drives the compensation mass block to generate the corresponding acceleration through its internal high-speed closed-loop position and current controller. According to Newton's third law, the moving mass exerts an equal and opposite reaction force on its base, the bucket foundation. , is the mass of the compensation mass. This reaction force acts on the installation location of the internal dynamic moment compensation system, thereby generating a compensation moment on the rotation center of the bucket foundation: ; In the formula, Indicates time The compensation torque, It is the system from the barrel base rotation center to the internal dynamic torque compensation system The moment arm vector at the center, Indicates time reaction force.

[0070] The compensation torque The response speed can reach milliseconds, and it can accurately and instantaneously offset the high-frequency external disturbance torque predicted by the disturbance prediction module, thereby maintaining the dynamic stability of the foundation under high-frequency disturbances.

[0071] Furthermore, to improve the energy efficiency and sustainable working capability of the system, the internal dynamic torque compensation system also includes an energy recovery unit. This unit is exemplarily a supercapacitor module and is connected to the DC bus of the servo actuator through a bidirectional DC-DC converter.

[0072] When the servo actuator decelerates the compensating mass, its regenerative braking function is activated. The servo motor then operates in generator mode, efficiently converting the kinetic energy of the compensating mass into electrical energy, which is then used to charge the supercapacitor module via the bidirectional DC-DC converter.

[0073] When the servo actuator needs to accelerate the mass again, the bidirectional DC-DC converter draws stored energy from the supercapacitor module to supplement or replace the main power supply. This energy recycling significantly reduces the system's peak power demand and overall energy consumption.

[0074] At the same time, the low-frequency control instruction The air is sent to the zoned air pressure control system. This system consists of multiple independently sealed chambers separated by sealed partitions at the top of the barrel foundation. Each chamber is connected to an inverter-controlled air compressor and an electronically controlled exhaust valve. Each chamber is also equipped with a high-precision pressure sensor to achieve closed-loop control.

[0075] The low-frequency control instruction For example, a vector is represented by its elements, each representing the air pressure setpoint for each chamber. After receiving the corresponding air pressure setpoint, the inverter in each chamber compares it with the real-time feedback from the pressure sensor through its internal PID controller. When the setpoint exceeds the measured value, the compressor's operating frequency is automatically adjusted to inflate the air. When the setpoint falls below the measured value, the exhaust valve is opened to exhaust the air, thereby precisely controlling the air pressure in each individual chamber.

[0076] By generating differential air pressure between the chambers, the system applies a net corrective moment to the bucket foundation. This gentle but powerful moment is specifically designed to correct static or low-frequency overall tilt of the bucket foundation caused by factors such as uneven seabed geology or persistent, slowly changing ocean currents.

[0077] Through the above-mentioned high-frequency and low-frequency coordinated execution method, the present invention can achieve targeted and efficient control of tilt disturbances of different properties.

[0078] The internal dynamic torque compensation system acts as an agile dynamic stabilizer, while the zoned air pressure control system acts as a powerful attitude adjuster. This method offers fast dynamic response, stable static correction, and low energy consumption, fully ensuring the installation accuracy and operational safety of bucket foundations in complex marine environments.

[0079] A barrel-type foundation air flotation dynamic deviation correction installation system, comprising: An information acquisition module, used to obtain real-time posture information of the bucket foundation; a disturbance prediction module for predicting external hydrodynamic disturbances acting on the bucket foundation; a collaborative control module, configured to generate collaborative control instructions based on the real-time posture information and the predicted external hydrodynamic disturbance; An execution module is connected to the collaborative control module, and the execution module includes: an internal dynamic torque compensation system; a zoned air pressure control system; The execution module is used to drive the internal dynamic torque compensation system to offset high-frequency dynamic tilt and drive the partitioned air pressure control system to correct static or low-frequency tilt according to the coordinated control instruction.

[0080] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A barrel-type foundation air flotation dynamic deviation correction installation method, characterized in that: The following steps are involved: S1: Acquire real-time posture information of the bucket foundation; S2: Predict the external hydrodynamic disturbance acting on the bucket foundation; S3: generating a coordinated control instruction based on the real-time posture information and the predicted external hydrodynamic disturbance; S4: According to the collaborative control instruction, drive the internal dynamic torque compensation system to offset the high-frequency dynamic tilt caused by the external hydrodynamic disturbance, and drive the partitioned air pressure control system to correct the static or low-frequency tilt of the bucket foundation.

2. The barrel-type foundation air flotation dynamic deviation correction installation method according to claim 1 is characterized in that: The real-time posture information of the bucket foundation is obtained as follows: High-frequency attitude data is acquired through an inertial measurement unit, low-frequency attitude data is acquired through a global navigation satellite system, and a Kalman filter algorithm is used to fuse the high-frequency and low-frequency attitude data to obtain the real-time attitude information.

3. The method for dynamic deviation correction installation of a barrel-type foundation with air flotation according to claim 1 is characterized in that: The external hydrodynamic disturbance predicted to act on the bucket foundation is specifically: The upstream water flow information is measured in real time by an acoustic Doppler flow profiler arranged on the incoming flow surface of the bucket-type foundation, and a prediction model is established based on the water flow information to obtain the predicted external hydrodynamic disturbance.

4. The barrel-type foundation air flotation dynamic deviation correction installation method according to claim 1 is characterized in that: The generation of the collaborative control instruction is specifically as follows: A model predictive control algorithm is used to take the real-time posture information and the predicted external hydrodynamic disturbance as input within a rolling finite time window, and the collaborative control instructions are generated by solving a constrained optimization problem.

5. The method for dynamic deviation correction installation of a barrel-type foundation with air flotation according to claim 4 is characterized in that: The system dynamics model used in the model predictive control algorithm can perform online parameter identification and adaptive update based on the error between the real-time posture information of the bucket foundation and the model prediction output.

6. The barrel-type foundation air flotation dynamic deviation correction installation method according to claim 1 is characterized in that: The internal dynamic torque compensation system includes a compensation mass block arranged inside the barrel-type foundation and a servo actuator that drives the compensation mass block to move at high speed; The driving internal dynamic torque compensation system specifically includes: controlling the servo actuator to drive the compensation mass block to generate an inertia torque opposite to the high-frequency dynamic tilt direction.

7. The method for dynamic deviation correction installation of a barrel-type foundation with air flotation according to claim 6 is characterized in that: When the servo actuator performs deceleration braking on the compensating mass block, the kinetic energy is converted into electrical energy through a regenerative braking function and stored in an energy recovery unit.

8. The method for dynamic deviation correction installation of a barrel-type foundation with air flotation according to claim 1 is characterized in that: Before obtaining the real-time posture information, the system is initialized. The initialization is specifically as follows: The initial position of the compensation mass block in the internal dynamic moment compensation system is fine-tuned before installation to calibrate and compensate for the initial unbalanced center of mass of the bucket foundation.

9. The barrel-type foundation air flotation dynamic deviation correction installation method according to claim 1 is characterized in that: The real-time posture information, external hydrodynamic disturbances, and collaborative control command data are integrated to generate a digital twin file for evaluating installation quality.

10. A barrel-type foundation air flotation dynamic deviation correction installation system, applied to a barrel-type foundation air flotation dynamic deviation correction installation method according to any one of claims 1 to 9, characterized in that: include: An information acquisition module, used to obtain real-time posture information of the bucket foundation; a disturbance prediction module for predicting external hydrodynamic disturbances acting on the bucket foundation; a collaborative control module, configured to generate collaborative control instructions based on the real-time posture information and the predicted external hydrodynamic disturbance; An execution module is connected to the collaborative control module, and the execution module includes: an internal dynamic torque compensation system; a zoned air pressure control system; The execution module is used to drive the internal dynamic torque compensation system to offset high-frequency dynamic tilt and drive the partitioned air pressure control system to correct static or low-frequency tilt according to the coordinated control instruction.

Images