A floating wind turbine platform stability control method and floating wind turbine platform

Abstract

A stability control method for a floating wind turbine platform and the floating wind turbine platform are disclosed. This control method dynamically adjusts the distribution of ballast water within each buoy of the wind turbine platform by transferring ballast water between multiple buoys based on environmental load information and the real-time attitude angle of the wind turbine platform. This alters the mass distribution characteristics of the wind turbine platform, thereby counteracting the unbalanced torque caused by external environmental loads. This invention can actively, dynamically, efficiently, and reliably control the stability of floating wind turbine platforms under complex sea conditions, overcoming the poor adaptability of existing passive anti-roll devices and the problems of complex structure, high energy consumption, insufficient reliability, or limited damping effect of existing active anti-roll devices. It belongs to the field of wind power generation technology.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, specifically to a method for stability control of a floating wind turbine platform and a floating wind turbine platform. Background Technology

[0002] With the global energy transition and the advancement of the "carbon neutrality" goal, offshore wind power, as an important component of clean energy, is rapidly developing from near-shore and shallow seas to deep-sea areas. In deep-sea regions, floating wind turbines have become the mainstream technology due to their strong adaptability to complex seabed topography and superior economic efficiency. However, floating wind turbines generate complex six-degree-of-freedom motion responses under the strong coupling of wind, waves, and currents, including pitch, roll, bow, sway, heave, and yaw. Among these, pitch and roll are the key attitude motions affecting platform stability. Severe swaying can lead to oscillation effects at the tower top and blade tips, generating large bending moments and shear forces at the tower base and blade roots, increasing fatigue loads on components and reducing the service life of the floating wind turbine. In extreme sea conditions, excessive pitch motion can cause the platform to take on water or even capsize, seriously threatening the self-sustaining safety of the floating wind turbine.

[0003] Therefore, effectively and dynamically adjusting the stability of floating wind turbine platforms is key to addressing the insufficient vibration resistance of existing platforms under the coupling of wind, waves, and currents. Currently, an important solution is to introduce various anti-sway devices that provide damping, suppressing the swaying amplitude of the floating wind turbine by adjusting the parameters of these devices, optimizing their arrangement, or applying active control.

[0004] For example, Chinese invention patent CN202510201360.6 (Floating Wind Power Anti-sway System and Control Method Based on Wind Turbine-Gyroscope Coupling) discloses a scheme to suppress platform sway by generating counter-torque through a dual-gyroscope device. This scheme utilizes the gyro effect to achieve active sway reduction, but it suffers from problems such as complex structure, high energy consumption, and slow response speed. Furthermore, the high-speed rotation of the gyroscopes leads to high mechanical wear and maintenance costs, making it difficult to adapt to the variable working conditions in deep-sea areas.

[0005] Chinese invention patent CN202411614424.7 (A Semi-Submersible Anti-Roll Floating Platform and Wind Turbine) enhances stability through a stepped pontoon structure and anti-roll layer design. This scheme passively enhances stability by optimizing the floating body structure (such as stepped pontoons and anti-roll layers). Its core drawback is the lack of initiative. Once the structural design is fixed, its anti-roll performance is also fixed, and it cannot be dynamically adjusted according to real-time wind and wave loads, thus limiting its ability to cope with extreme or sudden sea conditions.

[0006] Chinese invention patent CN202510983931.6 discloses an active anti-roll device for a floating platform. It uses sensors to detect the platform's angular velocity, a controller to calculate the tilt angle, and then controls the movement of a counterweight along a guide assembly circumferentially around the platform, dynamically adjusting the platform's center of gravity distribution to resist the swaying tendency. However, this solution relies on sophisticated mechanical transmission mechanisms (guide assemblies, lead screws, telescopic arms, etc.). In harsh marine corrosive environments, the reliability, wear resistance, and sealing of these moving parts face severe challenges, resulting in high maintenance costs. Furthermore, although the counterweight can move circumferentially, the anti-roll torque it generates is essentially achieved through adjusting the center of gravity position, which, compared to dampers based on hydrodynamic principles, results in a weaker damping effect in absorbing and dissipating wave energy.

[0007] In summary, existing active roll stabilization technologies often fall short in terms of reliability, response speed, maintenance costs, and adaptability to transient environmental changes. Therefore, floating wind turbine platforms urgently require an active stability control technology that balances rapid response, resistance to marine environments, efficient energy dissipation, and redundancy tolerance to achieve continuous, stable, and safe attitude stability control under complex and variable operating conditions. Summary of the Invention

[0008] In view of the technical problems existing in the prior art, the purpose of this invention is to provide a stability control method for floating wind turbine platforms, which can actively, dynamically, efficiently and reliably control the stability of floating wind turbine platforms under complex sea conditions.

[0009] Another object of the present invention is to provide a floating wind turbine platform.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: a stability control method for a floating wind turbine platform, which dynamically adjusts the distribution of ballast water in each pontoon of the wind turbine platform by transferring ballast water between multiple pontoons based on environmental load information and the real-time attitude angle of the wind turbine platform, thereby changing the mass distribution characteristics of the wind turbine platform and thus offsetting the unbalanced torque caused by external environmental load.

[0011] As a preferred option, the following steps are included:

[0012] A multi-source sensing system consisting of attitude sensors, wave sensors, wind lidar, and water level sensors on the pontoon is used to monitor the system status in real time. The system status includes the attitude angle of the wind turbine platform, environmental load, and ballast water level information.

[0013] The system state information is input into the pre-established dynamic prediction model of the floating wind turbine. The model predictive control algorithm is used as the attitude active adjustment strategy. Based on the current system state, the attitude angle change of the wind turbine platform in the next short period of time is predicted in each control cycle, and the optimal ballast water regulation control quantity is solved in the prediction time domain.

[0014] Model predictive control takes minimizing the absolute attitude angle deviation of the platform as the main optimization objective. At the same time, it introduces the system energy consumption caused by ballast water transfer and comprehensively considers physical constraints such as the response speed of water pumps and valves, the upper limit of flow rate, and the range of ballast water level. Under the premise of meeting the system's safe operation constraints, the optimal ballast water distribution and transfer strategy among each float is calculated.

[0015] By employing the aforementioned model-based predictive control strategy, the platform can anticipate attitude change trends under the influence of external wind, waves, and current disturbances, and generate compensating torque by adjusting the distribution of ballast water, thereby achieving forward-looking and stable control of the platform's attitude.

[0016] Based on the obtained optimal control quantity, the mass distribution of the wind turbine platform is changed by transferring ballast water between the floats through valves or water pumps, thereby generating a compensating torque to counteract the attitude deviation caused by environmental loads.

[0017] The execution system includes controllable valves, circulating water pumps and piping systems. It adopts a unidirectional circulating ballast water transfer structure and has redundant valve groups and actuator fault switching strategies to ensure that the system can still maintain continuous and stable control in the event of a partial failure.

[0018] The system monitors the changes in ballast water level and ballast water transmission status in real time, and continuously updates the control parameters based on the deviation between the feedback attitude angle and the target attitude angle. This forms an attitude control method that combines rolling prediction and closed-loop regulation, ensuring that the wind turbine platform remains within a safe attitude range.

[0019] As an alternative, the method also includes an attitude calibration step: the actual draft of the buoy is measured by a water level sensor under still water conditions, the actual tilt angle of the wind turbine platform relative to the still water reference plane is calculated by combining geometric parameters, and the attitude angle output by the attitude sensor under the same still water conditions is compared with the initial offset of the attitude sensor to obtain the initial offset of the attitude sensor, which is then stored in the attitude calibration parameter table to complete the attitude calibration; the real-time attitude angles subsequently monitored by the attitude sensor are corrected by the initial offset to obtain the absolute attitude angle.

[0020] As a preferred embodiment, the number of pontoons is three. The model predictive control algorithm takes minimizing the platform's pitch and roll angle fluctuations as its main optimization objective, while also considering system energy consumption, actuator response speed, and physical constraints of valves and pumps. The objective function is constructed in the prediction time domain as follows:

[0021] ;

[0022] in,

[0023] These are the pitch and roll angles of the wind turbine platform at the predicted time t+k, respectively.

[0024] denoted as target attitude angles, and denoted as target pitch angle and target roll angle, respectively;

[0025] This refers to the current change in the control quantity, which can be either a change in valve opening or a change in pump flow rate.

[0026] Fixed weighting coefficients;

[0027] The dynamic prediction model for floating wind turbines is obtained by linearizing the dynamic model of floating wind turbines built on OpenFAST. Its discrete state-space form is as follows:

[0028]

[0029]

[0030] in, The system state vector at the current moment includes the wind turbine platform attitude angle, angular velocity, and ballast water level in each pontoon, specifically expressed as:

[0031]

[0032] These are the real-time absolute roll and absolute pitch angles after attitude calibration correction, respectively. These are the angular velocities corresponding to the roll and pitch directions, respectively. These are the ballast water levels in the three floats;

[0033] This represents the system state vector at the next moment.

[0034] To control the input vector, the corresponding ballast water transfer control quantities between each float, including pump flow rate or valve opening, are specifically expressed as follows:

[0035]

[0036] This refers to the ballast water transfer flow rate from the first pontoon to the second pontoon. This refers to the ballast water transfer flow rate from the second to the third buoy. The ballast water transfer flow rate from the third pontoon to the first pontoon;

[0037] The system output vector corresponds to the predicted attitude angle, specifically expressed as:

[0038]

[0039] in, To predict the pitch angle, To predict the roll angle;

[0040] This is the system model matrix that is linearized from the dynamic model of the floating wind turbine built based on OpenFAST and solidified after calibration with actual wind turbine platform parameters.

[0041] As a preferred embodiment, the floating wind turbine dynamic model is constructed based on the platform's geometry, mass distribution, mooring parameters, and fluid force characteristics, and can describe the six-degree-of-freedom motion response of the floating platform under the coupled effects of wind, waves, and current.

[0042] A floating wind turbine platform for implementing the above-mentioned floating wind turbine platform stability control method includes a wind turbine main unit, a tower, and three pontoons. The three pontoons are arranged in an equilateral triangular array. The wind turbine main unit and the tower are fixed in the middle of the three pontoons. Each pontoon is equipped with a sway damping chamber, which is filled with ballast water. Hollow connecting pipes connect adjacent pontoons. Each connecting pipe is equipped with a water pump and a valve for internal adjustment of the ballast water level between the pontoons.

[0043] In summary, the present invention has the following advantages:

[0044] (1) Existing active roll reduction devices are mostly based on real-time attitude deviation feedback adjustment. Due to the inertia of fluid transmission and the physical characteristics of the actuator, the control often has obvious "time lag", which may even lead to system instability under short-period wave action. The core of this scheme is to build a closed-loop control system that integrates real-time environment and attitude perception, multibody dynamics model prediction, and active ballast water allocation. This technical approach is fundamentally different from existing passive roll reduction technology that mainly relies on structural optimization or mechanical active roll reduction schemes with different principles. This scheme makes dynamic decisions through model predictive control algorithms and actively drives the transfer of ballast water between pontoons to change the mass distribution of the platform and actively generate restoring torque. It belongs to a predictive intelligent closed-loop active control.

[0045] (2) The key components involved in this solution, such as the high-precision integrated navigation system, wind-measuring lidar, wave sensors, solenoid valves, and water pumps, are all mature products in the industrial or marine engineering fields, with reliable technology and easy procurement and integration. The constructed control system has clear input, processing, and output logic, and the algorithm can run stably on industrial-grade PLCs and dedicated controllers. This method directly addresses the practical engineering problem of the stability of floating wind turbines under complex sea conditions, effectively improving the platform's anti-sway performance, thereby ensuring the normal operation and power generation efficiency of the wind turbine generator set, and extending the structural lifespan. The redundant valves and fault-switching strategies included in the system design further ensure its long-term reliability in harsh marine environments. Therefore, this solution has clear engineering feasibility and significant practical application value. Attached Figure Description

[0046] Figure 1 This is an overall structural diagram of a floating wind turbine platform.

[0047] Figure 2 This is a schematic diagram of the structure of a floating platform.

[0048] Figure 3 This is a schematic diagram illustrating the internal adjustment of the ballast water level between the pontoons.

[0049] Figure 4 This is a schematic diagram of the anti-roll tank structure inside the buoy.

[0050] Figure 5 This is a schematic diagram of the baseline coordinate system and the dynamic coordinate system.

[0051] Figure 6 This is a control flowchart for the stability control method of a floating wind turbine platform.

[0052] Among them, 1 is the wind turbine main unit, 2 is the tower, 3 is the float, 4 is the connecting pipe, 5 is the mooring line, 6 is the wind measurement lidar, and 7 is the fixed pole.

[0053] 301 is a valve, 302 is a water pump, 303 is a rock damping chamber, and 304 is a water level sensor. 401 is an attitude sensor, and 402 is a wave sensor. Detailed Implementation

[0054] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0055] Example 1

[0056] like Figure 6As shown, a stability control method for a floating wind turbine platform dynamically adjusts the distribution of ballast water within each float by transferring ballast water between multiple floats based on environmental load information and the real-time attitude angle of the wind turbine platform. This alters the mass distribution characteristics of the wind turbine platform, thereby counteracting the unbalanced torque caused by external environmental loads. The control method comprises two stages:

[0057] 1. Attitude reference calibration stage (S01–S04)

[0058] 2. Closed-loop attitude control stage (S05–S11)

[0059] The attitude reference calibration stage must be conducted in a preset statically stable environment, which includes, but is not limited to, port waters, wharf basins, or deep-sea areas with calm sea conditions. The attitude reference calibration stage is used to establish the attitude offset parameters of the attitude sensor relative to the absolute static water reference surface; the closed-loop attitude control stage performs dynamic closed-loop adjustments based on the calibration results during platform operation.

[0060] S01, First, determine the calibration environment conditions: The control system determines whether the current sea state meets the calibration conditions of the attitude monitoring system based on data from the wave sensor and wind-measuring lidar. The calibration environment must meet the following requirements:

[0061] 1) Wind speed less than 2 m / s;

[0062] 2) The meaningful wave height is less than 0.2 m;

[0063] If all the above environmental conditions are met, proceed to the next calibration step; otherwise, postpone the attitude reference calibration.

[0064] S02, Static waterline measurement: The control system controls the water level sensor in each float to enter the calibration step. Each float has at least two liquid level measuring points. The control system automatically collects water level data from multiple points and takes the arithmetic mean of the liquid level measurements of the same float to obtain the static waterline height value of the three floats.

[0065] S03, Static Water Inclination Angle Calculation: Based on the obtained static waterline height value and the geometric dimensions of the wind turbine platform, calculate the static water inclination angle and roll angle φ relative to the static water reference plane. meas Pitch angle θ meas ;

[0066] S04, record the attitude angle φ output in this static state. base and θ base Based on this, a reference coordinate system for the attitude sensor relative to the absolutely still water surface is established. Figure 5 );

[0067] S05, Deviation Calculation and Storage: Calculate the difference between the still water tilt angle and the output attitude angle, Δφ0 = φmeas−φ base Δθ0=θ meas −θ base The difference is used as the initial bias of the attitude sensor and stored in the attitude calibration parameter table to complete the calibration process.

[0068] S06, After the wind turbine enters normal operating condition, the attitude angle output by the attitude sensor in real time is corrected by the initial offset to obtain the true absolute attitude angle:

[0069] φ(t)=φ sensor (t)+Δφ0;θ(t)=θ sensor (t)+Δθ0;

[0070] Where φ sensor (t), θ sensor (t) represents the attitude angle output in real time by the attitude sensor.

[0071] S07: Real-time Data Acquisition and Processing: After the wind turbine enters normal operating conditions, the control system collects environmental information and platform status data in real time through sensors and sensing modules. Specifically, this includes:

[0072] The corrected true absolute attitude angles φ(t) and θ(t) are obtained by high-precision attitude sensors, namely the real-time absolute roll angle and absolute pitch angle.

[0073] Wave direction, current velocity, and wave height are monitored using wave sensors installed on the pontoons;

[0074] The wind speed and direction in the upwind area are monitored by a wind-measuring lidar installed in the cabin.

[0075] The liquid level height of each buoy is monitored by the water level sensor built into the ballast tank;

[0076] All sensor data is preprocessed by the data acquisition module to remove invalid data and ensure data timing consistency before being transmitted to the main controller.

[0077] S08: Multibody dynamics model prediction: The main controller inputs the real-time collected absolute attitude angle, wind speed and direction, wave parameters and liquid level information into the pre-established floating wind turbine dynamics prediction model.

[0078] S09: Model Predictive Control (MPC) Algorithm Decision: The controller employs a model predictive control algorithm with the objective function of minimizing platform pitch and roll angle fluctuations, while also considering system energy consumption, actuator response speed, and physical constraints of valves and pumps. The controller constructs the following objective function in the predictive time domain:

[0079] ;

[0080] in,

[0081] These are the pitch and roll angles of the wind turbine platform at the predicted time t+k, respectively.

[0082] denoted as target attitude angles, and denoted as target pitch angle and target roll angle, respectively;

[0083] This refers to the current change in the control quantity, which can be either a change in valve opening or a change in pump flow rate.

[0084] Fixed weighting coefficients;

[0085] The dynamic prediction model for floating wind turbines is obtained by linearizing the dynamic model of floating wind turbines built on OpenFAST. Its discrete state-space form is as follows:

[0086]

[0087]

[0088] in, The system state vector at the current moment includes the wind turbine platform attitude angle, angular velocity, and ballast water level in each pontoon, specifically expressed as:

[0089]

[0090] These are the real-time absolute roll and absolute pitch angles after attitude calibration correction, respectively. These are the angular velocities corresponding to the roll and pitch directions, respectively. These are the ballast water levels in the three floats;

[0091] This represents the system state vector at the next moment.

[0092] To control the input vector, the corresponding ballast water transfer control quantities between each float, including pump flow rate or valve opening, are specifically expressed as follows:

[0093]

[0094] This refers to the ballast water transfer flow rate from the first pontoon to the second pontoon. This refers to the ballast water transfer flow rate from the second to the third buoy. The ballast water transfer flow rate from the third pontoon to the first pontoon;

[0095] The system output vector corresponds to the predicted attitude angle, specifically expressed as:

[0096]

[0097] in, To predict the pitch angle, To predict the roll angle;

[0098] This is the system model matrix that is linearized from the dynamic model of the floating wind turbine built based on OpenFAST and solidified after calibration with actual wind turbine platform parameters.

[0099] This floating wind turbine dynamic model (multibody dynamic model) is constructed based on the platform's geometry, mass distribution, mooring parameters, and fluid force characteristics. It can accurately describe the six-degree-of-freedom motion response of the floating platform under the coupled effects of wind, waves, and current. The controller uses the predictive model to predict the changes in the platform's pitch and roll angles within a short future timeframe (a 10-second rolling prediction window), generating an attitude response prediction sequence, which provides a basis for subsequent model predictive control optimization solutions.

[0100] S10: Control command execution and ballast water regulation: The control system uses a unidirectional loop method based on the control commands output by the MPC algorithm. Figure 3 The system regulates ballast water flow, forming a closed-loop link between the three pontoons. The control system adjusts the valve openings in the connecting pipelines via low-power solenoid valves, controlling the ballast water flow between the pontoons. This transfers water from one pontoon to the other, altering the platform's mass distribution and generating a counter-torque to counteract the unbalanced torque caused by external wind, waves, and currents. Furthermore, the system employs a redundant valve group design to automatically detect valve or pump failures. If a single actuator fails, a backup channel is immediately activated to ensure continuous control.

[0101] S11: Closed-loop feedback and dynamic balance recovery: Real-time monitoring of pipeline pressure and ballast levels in each float using pressure and water level sensors establishes a closed-loop feedback mechanism. The controller compares the current attitude angle (φ(t), θ(t)) with the target attitude angle (φ...). ref ,θ ref When the attitude angle remains within the allowable error range of the target angle for 5 seconds (i.e., |φ(t) - φ), ref |≤δ φ And |θ(t)-θ ref | ≤δ θWhen the attitude angle change rate is lower than the preset threshold, the platform is considered to have restored dynamic balance, the control system enters steady-state holding mode, and the valves and water pumps are temporarily shut off; otherwise, return to step S07 to continue collecting data and repeat the above control process until the attitude is stable.

[0102] Example 2

[0103] This embodiment discloses a floating wind turbine platform for implementing the control method of Embodiment 1.

[0104] like Figure 1 As shown, the floating wind turbine platform includes a wind turbine main unit, a tower, and three pontoons. The three pontoons are arranged in an equilateral triangular array, and the wind turbine main unit and the tower are fixed in the middle of the three pontoons.

[0105] like Figure 2 , Figure 3 , Figure 4 As shown, each pontoon is equipped with a roll damping chamber containing ballast water. Hollow connecting pipes connect adjacent pontoons, and each connecting pipe is equipped with a water pump and valves for internal adjustment of the ballast water level between pontoons.

[0106] The floating wind turbine platform also includes multiple anchor rods, which connect adjacent pontoons and between the pontoons and the tower. The pontoons, anchor rods, and other components constitute the floating platform, which is anchored to the seabed by a mooring system. The overturning moment generated by the weight of the tower and the wind turbine main unit is balanced by the gravitational moment of the ballast water within the three pontoons. A mooring line is installed below each pontoon to ensure the entire floating platform remains within a designated range.

[0107] The system consists of pumps, valves, connecting pipes, etc. It adopts a unidirectional circulation ballast water transfer structure and has redundant valve groups and actuator fault switching strategies to ensure that the system can still maintain continuous and stable control in the event of a partial failure.

[0108] The above embodiments are preferred embodiments of the invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A floating wind turbine platform stability control method, characterized by: Based on environmental load information and the real-time attitude angle of the wind turbine platform, the distribution of ballast water in each pontoon of the wind turbine platform is dynamically adjusted by transferring ballast water between multiple pontoons, thereby changing the mass distribution characteristics of the wind turbine platform and counteracting the unbalanced torque caused by external environmental loads.

2. The stability control method of a floating wind turbine platform according to claim 1, wherein: Includes the following steps, A multi-source sensing system consisting of attitude sensors, wave sensors, wind lidar, and water level sensors on the pontoon is used to monitor the system status in real time. The system status includes the attitude angle of the wind turbine platform, environmental load, and ballast water level information. The system state information is input into the pre-established dynamic prediction model of the floating wind turbine. The model predictive control algorithm is used as the attitude active adjustment strategy. Based on the current system state, the attitude angle change of the wind turbine platform in the next short period of time is predicted in each control cycle, and the optimal ballast water regulation control quantity is solved in the prediction time domain. Based on the obtained optimal control quantity, the mass distribution of the wind turbine platform is changed by transferring ballast water between the floats through valves or water pumps, thereby generating a compensating torque to counteract the attitude deviation caused by environmental loads. The system monitors the changes in ballast water level and ballast water transmission status in real time, and continuously updates the control parameters based on the deviation between the feedback attitude angle and the target attitude angle. This forms an attitude control method that combines rolling prediction and closed-loop regulation, ensuring that the wind turbine platform remains within a safe attitude range.

3. The method for controlling the stability of a floating wind turbine platform according to claim 2, characterized in that: It also includes the attitude calibration step: the actual draft of the float under still water conditions is measured by the water level sensor, the actual tilt angle of the wind turbine platform relative to the still water reference plane is calculated by combining the geometric parameters, and the attitude angle output by the attitude sensor under the same still water conditions is compared with the attitude angle to obtain the initial offset of the attitude sensor, which is stored in the attitude calibration parameter table to complete the attitude calibration. The real-time attitude angles subsequently detected by the attitude sensors are corrected using the initial offset to obtain the absolute attitude angles.

4. The method for controlling the stability of a floating wind turbine platform according to claim 2, characterized in that: The number of pontoons is three. The model predictive control algorithm takes minimizing the platform's pitch and roll angle fluctuations as its main optimization objective, while also considering system energy consumption, actuator response speed, and physical constraints of valves and pumps. The objective function is constructed in the prediction time domain as follows: ； in, These are the pitch and roll angles of the wind turbine platform at the predicted time t+k, respectively. denoted as target attitude angles, and denoted as target pitch angle and target roll angle, respectively; This refers to the current change in the control quantity, which can be either a change in valve opening or a change in pump flow rate. Fixed weighting coefficients; The dynamic prediction model for floating wind turbines is obtained by linearizing the dynamic model of floating wind turbines built on OpenFAST. Its discrete state-space form is as follows: ； ； in, The system state vector at the current moment includes the wind turbine platform attitude angle, angular velocity, and ballast water level in each pontoon, specifically expressed as: ； These are the real-time absolute roll and absolute pitch angles after attitude calibration correction, respectively. These are the angular velocities corresponding to the roll and pitch directions, respectively. These are the ballast water levels in the three floats; This represents the system state vector at the next moment. To control the input vector, the corresponding ballast water transfer control quantities between each float, including pump flow rate or valve opening, are specifically expressed as follows: ； This refers to the ballast water transfer flow rate from the first pontoon to the second pontoon. This refers to the ballast water transfer flow rate from the second to the third buoy. The ballast water transfer flow rate from the third pontoon to the first pontoon; The system output vector corresponds to the predicted attitude angle, specifically expressed as: ； in, To predict the pitch angle, To predict the roll angle; This is the system model matrix that is linearized from the dynamic model of the floating wind turbine built based on OpenFAST and solidified after calibration with actual wind turbine platform parameters.

5. The method for controlling the stability of a floating wind turbine platform according to claim 4, characterized in that: The floating wind turbine dynamic model is constructed based on the platform's geometry, mass distribution, mooring parameters, and fluid force characteristics, and can describe the six-degree-of-freedom motion response of the floating platform under the coupled effects of wind, waves, and current.

6. A floating wind turbine platform for implementing the stability control method for the floating wind turbine platform according to any one of claims 1 to 5, characterized in that: It includes a wind turbine main unit, a tower, and three pontoons. The three pontoons are arranged in an equilateral triangular array. The wind turbine main unit and the tower are fixed in the middle of the three pontoons. Each pontoon is equipped with a ballast tank, which is filled with ballast water. Hollow connecting pipes connect adjacent pontoons. Each connecting pipe is equipped with a water pump and valve for internal adjustment of the ballast water level between pontoons.

Images