Adjustable floating wind power foundation based on reinforcement learning and modular construction method

Abstract

The invention provides an adjustable floating wind power foundation based on reinforcement learning and a modular construction method, and belongs to the technical field of offshore wind power. Comprising a fan, and the bottom of the fan is fixed to a tower drum; the bottom of the tower drum is fixedly connected with the floating base; the floating base comprises a first buoy, a second buoy and a third buoy which are arranged in a triangular shape; wherein the bottom of the tower drum is fixedly connected with the first buoy; the first buoy, the second buoy and the third buoy are fixedly connected through supporting steel pipes; the tower drum is fixedly connected with the second buoy and the third buoy through inclined supports; the buoy is further connected with an anchor chain, and the other end of the anchor chain is connected with the suction barrel and anchored to the seabed through the suction barrel. By adopting a coupling connection mode of the inclined support, the tower drum and the floating drum, the storm resistance of the whole structure is improved. A steel-concrete composite structure is adopted, the steel consumption is reduced, and the material cost is reduced. A modular construction method is adopted, the offshore construction time is greatly shortened, and the influence of weather and sea conditions on engineering is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of offshore wind power, and in particular relates to an adjustable floating wind power foundation and a modular construction method based on reinforcement learning. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Against the backdrop of global energy transformation and sustainable development, the demand for clean energy is becoming increasingly urgent. Offshore wind power, with its abundant wind energy resources and minimal impact on the ecological environment, is gradually becoming a core development direction in the energy sector. With the vigorous development of the offshore wind power industry, the requirements for wind power foundations are becoming increasingly stringent. Traditional fixed offshore wind power foundations can still play a certain role in shallow waters, but when faced with deep waters, their construction difficulty increases exponentially and costs also increase sharply, which to a large extent restricts the expansion of offshore wind power to deeper and wider waters. Therefore, floating wind power foundations came into being and became a research hotspot and application trend in the current offshore wind power field.

[0004] However, while floating wind turbine foundations offer new opportunities for deepwater wind power development, they also face numerous significant challenges. Floating wind turbines in deepwater environments are susceptible to swaying due to wind and wave forces, impacting power generation efficiency and equipment lifespan. The increased size of wind turbines creates significant loads at the base of the tower, which existing tower sizes cannot withstand. Furthermore, the sheer volume of steel used in construction contributes to high construction costs. Summary of the Invention

[0005] In order to overcome the deficiencies of the above-mentioned prior art, the present invention provides an adjustable floating wind power foundation and modular construction method based on reinforcement learning. It adopts a unique design combining steel and concrete, with steel as the main tension member and concrete as the main compression member. It effectively solves the problem of large steel consumption and high cost of existing floating wind power foundations, and enhances stability in deep water areas. In addition, the wind turbine tower is connected to the buoy through an oblique support, which increases stability while reducing the size of the tower and the bottom foundation steel pipe support, further reducing the amount of steel used. A pendulum damper is added to the tower. By introducing a deep reinforcement learning adaptive control algorithm, it can automatically adjust the size and direction of the damping force according to the real-time swing state of the tower, achieving more accurate and effective anti-sway control. By combining with a modular construction method, construction efficiency can be greatly improved and costs can be reduced.

[0006] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: A first aspect of the present invention provides an adjustable floating wind power foundation based on reinforcement learning; An adjustable floating wind power foundation based on reinforcement learning, comprising: A wind turbine, wherein the bottom of the wind turbine is fixed to the tower, and a controller is installed inside the wind turbine; the bottom of the tower is fixedly connected to the floating base; a pendulum damper is provided inside the tower, and the pendulum damper includes a top suspension cross brace, a top suspension spring, a lateral limit spring, and a mass block; The floating base includes a first pontoon, a second pontoon, and a third pontoon arranged in a triangle; the bottom of the tower is fixedly connected to the first pontoon; the first pontoon, the second pontoon, and the third pontoon are fixedly connected via supporting steel pipes; the tower is fixedly connected to the second pontoon and the third pontoon via oblique supports; The first buoy, the second buoy and the third buoy are also connected to anchor chains, the other ends of the anchor chains are connected to the suction buckets and are anchored to the seabed through the suction buckets.

[0007] As a further technical solution, the interior of the buoy is a cavity, and the buoy arms are composed of a buoy outer steel plate, a buoy concrete interlayer and a buoy inner steel plate from the outside to the inside.

[0008] As a further technical solution, the tower is composed of an outer tower steel plate, a tower concrete interlayer and an inner tower steel plate from the outside to the inside.

[0009] As a further technical solution, the bottom of the buoy is further connected by connecting ribs and heave plates.

[0010] As a further technical solution, the outer steel plate and the inner steel plate of the buoy are fixed by a supporting frame.

[0011] As a further technical solution, the interior of the buoy adopts a compartment sealing design.

[0012] As a further technical solution, the adjustable floating wind turbine based on reinforcement learning is also fixed with a variety of sensors, which are connected to the internal controller of the wind turbine, including six-degree-of-freedom motion sensors, stress sensors, displacement sensors, wind speed sensors, and wind direction sensors, which are used to obtain marine environment parameter data and motion parameter data.

[0013] As a further technical solution, a pneumatic adjustment device is also provided on the oblique support and is connected to the internal controller of the fan.

[0014] As a further technical solution, the supporting steel pipe includes a first supporting steel pipe, a second supporting steel pipe, a third supporting steel pipe and a fourth supporting steel pipe; the first supporting steel pipe is fixed above any two side walls of the pontoons; the second supporting steel pipe is arranged parallel to the first supporting steel pipe and fixed below any two side walls of the pontoons; the third supporting steel pipe and the fourth supporting steel pipe are fixedly connected between the first supporting steel pipe and the second supporting steel pipe, and the third supporting steel pipe, the fourth supporting steel pipe and the second supporting steel pipe form a triangular shape.

[0015] A second aspect of the present invention provides a modular construction method for an adjustable floating wind power foundation based on reinforcement learning.

[0016] A modular construction method for an adjustable floating wind turbine foundation based on reinforcement learning, including: a prefabrication stage, an offshore assembly stage, and a commissioning and testing stage: S1, prefabrication stage; the prefabrication stage includes prefabrication of tower modules, buoy modules, heave plate modules and wind turbine modules in the factory; commissioning of electrical equipment, testing equipment and communication equipment; S2, offshore assembly phase: using large offshore vessels to transport the prefabricated tower modules, buoy modules, heave plate modules, and wind turbine modules to designated sea areas for connection. After assembly, electrical equipment, testing equipment, and communication equipment are installed and further debugged. The buoys and anchor chains are connected, the suction bucket is transported to the designated location, and inserted into the seabed to complete the installation of the mooring system. S3, debugging and testing phase; using deep reinforcement learning control system to optimize the parameters of the diagonal support and damper; testing the structural stability, power generation performance and data transmission function.

[0017] One or more of the above technical solutions have the following beneficial effects: (1) The present invention uses a coupling connection method between the inclined support and the tower and pontoon to effectively disperse the bending moment and horizontal force on the tower, thereby improving the overall structure's ability to resist wind and waves. The steel-concrete composite structure (steel plate sandwiched with concrete) is designed to fully utilize the tensile properties of steel and the compressive properties of concrete, reduce the amount of steel used in the pontoon and tower, and significantly reduce material costs. The introduction of the inclined support optimizes the tower's force pattern, allowing the tower and bottom support structure to be reduced in size, further reducing the amount of steel used.

[0018] (2) A modular construction method is adopted. Key components such as buoys, towers, and diagonal supports are prefabricated in the factory and only need to be quickly assembled on site, which greatly shortens the offshore construction time and reduces the impact of weather and sea conditions on the project.

[0019] (3) The built-in pendulum damper in the tower is combined with a reinforcement learning control algorithm, which can dynamically adjust the damping force according to the real-time swing state, reduce the swing amplitude of the tower, and significantly improve the power generation efficiency of the wind turbine.

[0020] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0022] Figure 1 Schematic diagram of the floating wind power foundation structure in the first embodiment.

[0023] Figure 2 This is a top view of the tower structure in the first embodiment.

[0024] FIG3( a ) is a front view of the buoy structure in the first embodiment; FIG3( b ) is a top view of the buoy structure in the first embodiment.

[0025] FIG4( a ) is a top view of the pendulum damper in the first embodiment; FIG4( b ) is a front view of the pendulum damper in the first embodiment.

[0026] Figure 5 This is a diagram showing the verification results of the damper adjustment process in the first embodiment.

[0027] Figure 6 This is a flow chart of the modular construction method in the second embodiment.

[0028] In the figure, 1 is a wind turbine, 2 is a tower, 3 is a pontoon, 3-1 is a first pontoon, 3-2 is a second pontoon, 3-3 is a third pontoon, 4 is a supporting steel pipe, 4-1 is a first supporting steel pipe, 4-2 is a second supporting steel pipe, 4-3 is a third supporting steel pipe, 4-4 is a fourth supporting steel pipe, 5 is a connecting rib, 6 is a heave plate, 7 is a silo, 8 is a diagonal support, 9 is an anchor chain, 10 is a suction bucket, 11 is a pontoon outer steel plate, 12 is a pontoon concrete interlayer, 13 is a pontoon inner steel plate, 14 is a tower outer steel plate, 15 is a tower concrete interlayer, 16 is a tower inner steel plate, 17 is a cross brace, 18 is a top suspension spring, 19 is a mass block, and 20 is a lateral limit spring. DETAILED DESCRIPTION It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0029] It should be noted that the terms used herein are for describing particular embodiments only and are not intended to limit the exemplary embodiments according to the present invention.

[0030] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0031] Example 1 This embodiment discloses an adjustable floating wind power foundation based on reinforcement learning; like Figure 1 As shown, an adjustable floating wind power foundation based on reinforcement learning includes: Wind turbine 1, the bottom of the wind turbine 1 is fixed to the tower 2; the bottom of the tower 2 is fixedly connected to the floating base; Figure 2 The tower 2 is composed of a tower outer steel plate 14, a tower concrete interlayer 15 and a tower inner steel plate 16 from the outside to the inside.

[0032] The floating base includes a first pontoon 3-1, a second pontoon 3-2 and a third pontoon 3-3 arranged in a triangle; the traditional floating foundation is generally made of pure steel, which consumes a large amount of steel. Figure 3(a)-Figure 3(b) The pontoon 3 utilizes a composite structure of steel plates and concrete, employing a compartmentalized design. From the outside in, the pontoon consists of the outer steel plate 11, the concrete interlayer 12, and the inner steel plate 13. The outer and inner steel plates are secured by a support frame, and concrete is poured into the area enclosed by the steel plates and the support frame. This significantly reduces steel usage. The steel plates serve as the primary tensile component, while the concrete serves as the primary compressive component, bearing loads from above and below, as well as wind, wave, and current loads.

[0033] The buoy 3 is cylindrical in shape and features a compartmentalized, sealed design. This design utilizes prefabricated and spliced ​​assemblies for efficient assembly. The interior is divided into several independent silos 7, each of which is a fully sealed unit, enhancing the buoy's resistance to sinking and facilitating the installation and maintenance of internal equipment. Each sealed silo is a separate prefabricated component with a modular design. Prefabricated components are precisely joined together using pre-defined joints and secured with high-strength connectors (such as welding or bolts). Locating pins and grooves are embedded in the joints to ensure precise alignment during assembly. Sealing materials (such as rubber gaskets or epoxy resin) are used to prevent leakage. This structural design creates a tighter connection between steel and concrete, effectively improving the overall structural performance of the buoy. The steel plates and buoy 3 are tightly nested, forming a more stable integrated structure. This allows for better coordination when subjected to loads such as wind, waves, and currents, reducing structural deformation and stress concentration, and enhancing the stability and reliability of the buoy 3 in complex marine environments. The buoy 3 is also connected to an anchor chain 9 , the other end of which is connected to a suction bucket 10 , and is anchored to the seabed through the suction bucket 10 .

[0034] The bottom of the buoy 3 is also connected to the heave plate 6 via connecting ribs 5. The heave plate 6 is made of steel and features a retractable or foldable structure for easy transportation. It is connected to the buoy 6 via prefabricated connecting ribs 5. During installation, the steel connecting ribs 5 are welded or bolted to the buoy 3, ensuring a tight connection between the heave plate 6 and the buoy 3, effectively maximizing the heave plate's ability to suppress the heave motion of the buoy 3.

[0035] Furthermore, the bottom of the tower 2 is fixedly connected to the first pontoon 3-1, and the first pontoon 3-1, the second pontoon 3-2 and the third pontoon 3-3 are fixedly connected via a support steel pipe 4; wherein the support steel pipe 4 includes a first support steel pipe 4-1, a second support steel pipe 4-2, a third support steel pipe 4-3 and a fourth support steel pipe 4-4; the first support steel pipe 4-1 is fixed above the side walls of any two pontoons 3; the second support steel pipe 4-2 is arranged parallel to the first support steel pipe 4-1 and fixed below the side walls of any two pontoons 3; the third support steel pipe 4-3 and the fourth support steel pipe 4-4 are fixedly connected between the first support steel pipe 4-1 and the second support steel pipe 4-2, and the third support steel pipe 4-3, the fourth support steel pipe 4-4 and the second support steel pipe 4-2 form a triangle shape. The support steel pipes can further improve the stability of the device.

[0036] Furthermore, the device is equipped with various sensors (not shown), including six-degree-of-freedom motion sensors, stress sensors, displacement sensors, wind speed sensors, and wind direction sensors. The six-degree-of-freedom motion sensors monitor roll, pitch, and heave motion. Stress sensors and displacement sensors are installed at key locations, such as the tower, buoys, and diagonal supports, to detect forces and displacements in the device, facilitating the determination of its stability. Wind speed and direction sensors are mounted on the top of the tower to ensure accurate measurement of wind speed and direction. These sensors are connected to the wind turbine's internal controller to acquire marine environmental and motion parameter data. This allows the controller to adaptively adjust tension and damping forces based on these parameters. Tower 2 is also fixedly connected to the second buoy 3-2 and the third buoy 3-3 via diagonal supports 8. The connections between diagonal supports 8, buoys 3, and tower 8 feature hydraulic or pneumatic adjustment devices. These devices, connected to the wind turbine's internal controller and controlled using a deep learning algorithm, enable real-time adjustment of forces between the diagonal supports and buoys, improving the device's stability.

[0037] Combine Figure 4(a)-Figure 4(b)A pendulum damper is also installed above the tower 2. It consists of a top suspension cross brace 17, a top suspension spring 18, a mass 19, and lateral restraint springs 20. The spring and mass 19 are combined and installed within the tower 2, taking up no additional space. The damping is adjustable, adapting to varying wind and wave conditions by adjusting the spring stiffness and mass. The low cost, simple components, and easy manufacturing and installation reduce the turbine's operating costs. Furthermore, the spring and damper are connected to a controller. The damper's mass 19 serves as a vibration inertia unit, suspended by the top suspension spring 18. Lateral restraint springs 20 are connected on either side. These lateral restraint springs 20 are made of shape memory alloy (SMA) and feature a built-in graphene heating film, creating a three-dimensionally adjustable stiffness system (with fixed vertical stiffness and dynamically adjustable lateral stiffness via SMA phase transition). Control is achieved using a deep learning reinforcement algorithm to reduce the turbine's swing amplitude.

[0038] A deep reinforcement learning control system is used to optimize pendulum damper parameters; structural stability, power generation performance, and data transmission functions are tested.

[0039] The system comprises an environmental perception module, a dynamic model module, a reinforcement learning decision module, and an execution control module. The environmental perception module collects the floating body's six-degree-of-freedom motion parameters (position, attitude angle) and ocean environmental parameters (wind speed, wave height, and wave period) in real time. The dynamic model module establishes differential equations for the floating wind turbine system based on Newtonian mechanics and calculates the impact of damper changes on system stability. The reinforcement learning decision module employs a proximal policy optimization (PPO) algorithm to train an intelligent agent to learn the optimal damper adjustment strategy by constructing a multidimensional state space encompassing the floating body's state, environmental parameters, and current tension, as well as a continuous action space for damper adjustment. The execution control module adjusts the damper device in real time based on the decision output. By designing a multi-objective reward function that balances displacement minimization, attitude stabilization, and energy consumption control, this invention achieves adaptive dynamic adjustment of floating wind turbine structural parameters in complex ocean environments.

[0040] Furthermore, in the process of adjusting the damper parameters, by defining a dynamic model of the floating wind turbine and setting parameters such as mass, spring stiffness, and damping coefficient, the displacement and speed changes of the wind turbine under the action of wave force are simulated based on Newton's second law. A reward function is designed with the goal of minimizing displacement and speed. When the simulation duration reaches 3600 seconds, it is determined that one round has ended.

[0041] Specifically: The floating wind turbine system can be simplified as a mass-spring-damper system, whose equation of motion is based on Newton's second law:

[0042] Where: m is the equivalent mass of the floating wind turbine, k is the spring stiffness coefficient, c is the damping coefficient, x is the displacement, is the wave force, modeled as a simple harmonic function .

[0043] The reward function is designed to minimize displacement and velocity, as shown below:

[0044] in, is the reward function; In discrete time steps Next, use the Euler integration method to update the system state:

[0045] Furthermore, the Proximal Policy Optimization (PPO) algorithm is used to optimize the strategy using a truncated objective function. By building a strategy network, a two-layer fully connected neural network is used to output the spring stiffness and damping coefficient using a normal distribution to adjust the mean and standard deviation of the action. Specifically: The PPO algorithm uses a truncated objective function to optimize the policy:

[0046] in: To trim the loss function, it is used to limit the amplitude of the strategy update to avoid the large difference between the new and old strategies causing unstable training; is the probability ratio of the new and old strategies; is the advantage estimate; is the discounted return, ; is the clipping parameter, usually set to 0.2.

[0047] The advantage function is estimated using the generalized advantage estimator (GAE):

[0048] in, is a generalized advantage estimator used to approximate the advantage function of an action. is a discount factor used to attenuate the importance of future rewards. is a GAE hyperparameter that controls how much the advantage estimate weighs the multi-step TD error. is a non-negative integer index, is the time difference error, .

[0049] Policy Network Parameterized as a Gaussian distribution:

[0050] in, and are the mean and standard deviation of the output of the neural network. represents Gaussian distribution;

[0051] in, It is a hyperbolic tangent function that compresses the output value to the interval [-1,1] to limit the range of the mean; 、 、 are the weight matrices of the first layer, the second mean layer, and the second standard deviation layer respectively; 、 、 are the bias vectors of the corresponding layers respectively; is the state vector.

[0052] Further construction includes a multi-dimensional state space containing the floating body state, environmental parameters and current tension, as well as a continuous action space for the tension adjustment of each inclined support. In the state space, the state vector contains the physical state and time information of the system:

[0053] in: is the current displacement; is the current speed; t is the current time.

[0054] In the action space, the action is the adjustment of the stiffness and damping parameters:

[0055] in, is the control action vector at time t, which represents the damper parameter adjustment instruction output by the reinforcement learning algorithm; is the adjustment amount of the damper spring stiffness at time t; is the adjustment amount of the damper damping coefficient at time t.

[0056] The actual parameters are updated as follows:

[0057]

[0058] In the training phase, 50 training rounds are set (for demonstration purposes only, but tens of thousands of training rounds are actually required). In each round, the agent continuously interacts with the environment, stores experience, and updates parameters. Finally, the training reward curve, displacement changes under the optimal strategy, and the dynamic adjustment process of spring stiffness and damping coefficient are recorded and visualized, verifying the effectiveness and feasibility of the reinforcement learning algorithm in dynamically adjusting the structural parameters of floating wind turbines and improving stability. Figure 5 shown.

[0059] Example 2 This embodiment discloses a modular construction method for an adjustable floating wind power foundation based on reinforcement learning; like Figure 6 As shown, a modular construction method for an adjustable floating wind turbine foundation based on reinforcement learning includes: S1, prefabrication stage; the prefabrication stage includes prefabrication of tower modules, buoy modules, heave plate modules and wind turbine modules in the factory; commissioning of electrical equipment, testing equipment and communication equipment; S2, offshore assembly phase: using large offshore vessels to transport the prefabricated tower modules, buoy modules, heave plate modules, and wind turbine modules to designated sea areas for connection. After assembly, electrical equipment, testing equipment, and communication equipment are installed and further debugged. The buoys and anchor chains are connected, the suction bucket is transported to the designated location, and inserted into the seabed to complete the installation of the mooring system. S3, debugging and testing phase; using deep reinforcement learning control system to optimize the parameters of the diagonal support and damper; testing the structural stability, power generation performance and data transmission function.

[0060] Specifically, in step S1, the tower is manufactured in sections at a prefabrication plant on land according to the design size and strength requirements. Anchor installation reservations are made at specific locations on the tower, and the pre-installation and strength testing of the anchors are completed in advance to ensure that the anchors can withstand the tension of the diagonal supports. The annular steel frame and the longitudinal steel frame of the pontoon are welded and assembled in the prefabrication workshop, and reinforcement plates are set at the connection nodes of the annular steel frame and the longitudinal steel frame. Concrete is poured in an integrated manner between the two layers of steel plates. In the prefabrication plant, the diagonal supports and heave plates are manufactured in the factory according to the design length and tensile and compressive requirements, and high-strength connectors and adjustment devices are equipped. The assembly and commissioning of various components of the wind turbine are completed, including blades, hubs, nacelles, etc., to ensure that the overall performance of the wind turbine meets the design standards. At the same time, prepare the connection components between the wind turbine and the top of the tower, such as high-strength bolts, shock-absorbing pads, etc.

[0061] For electrical equipment modules, the assembly and commissioning of electrical equipment such as cables, transformers, and switchgear are completed in the factory to ensure good equipment performance. Cables are cut to the designed length in advance, and the ends are processed to facilitate subsequent installation and connection. For monitoring equipment modules, various sensors such as stress sensors, displacement sensors, wind speed sensors, and wind direction sensors are prefabricated, calibrated, and tested. The sensors are integrated with the supporting data acquisition devices and transmission lines to form a complete monitoring equipment module. For communication equipment modules, the assembly and commissioning of communication equipment such as communication base stations, antennas, and signal transmission cables are completed in the factory to ensure that the communication equipment can send and receive signals normally. The communication equipment is waterproof and moisture-proof to adapt to the harsh marine environment.

[0062] In step S2, the prefabricated buoys are transported to a designated sea area dock using a large offshore vessel. They are then connected using welding or high-strength bolts. After the connection is complete, the connection points are locally reinforced with stiffening ribs and reinforcement bars, and anti-corrosion treatment is performed.

[0063] After the pontoon is assembled, the prefabricated heave plates are hoisted to the corresponding positions on the pontoons. The heave plate's steel connecting ribs are welded or connected to the pontoon's steel frame using high-strength bolts according to design requirements. The connections are rigorously inspected for strength and accuracy. The prefabricated tower is hoisted in sections onto the pontoon columns for precise docking and assembly, ensuring that the tower's verticality meets strict standards.

[0064] After the tower, pontoons, and heave plates are installed, securely connect one end of the diagonal support to the hole on the tower using a high-strength connector. Install the other end to the connection structure at the top of the pontoon. Adjustment devices such as turnbuckles are then used to precisely adjust the tension of the diagonal support according to actual conditions to ensure balanced force across the entire structure. After the tower is securely installed and the diagonal support is adjusted into place, hoist the prefabricated wind turbine to the top of the tower. Connect the wind turbine to the tower using pre-prepared connection components. During the connection process, tighten the high-strength bolts strictly according to the torque requirements and check the installation of the shock-absorbing pads to ensure a secure connection between the wind turbine and the tower and effective shock absorption.

[0065] Cables are laid and electrical equipment is installed along the preset cable channels inside the tower and on the pontoon. After installation, electrical performance tests are carried out to check whether parameters such as line conductivity and insulation resistance meet the requirements.

[0066] Install relevant monitoring equipment such as stress sensors and displacement sensors on key locations such as the tower, buoys, and diagonal supports, and secure them with pre-buried mounting brackets. Connect the sensor's transmission line to the data acquisition device, and then connect the data acquisition device to the communication equipment to achieve real-time transmission of monitoring data. Install communication equipment such as communication base stations and antennas at the top of the tower or in appropriate locations on the buoy to ensure that the signal coverage meets the requirements. Lay signal transmission cables and connect the communication equipment to the data transmission interfaces of wind turbines, monitoring equipment, etc. to achieve remote data transmission and monitoring. After installation, perform a communication test to check whether indicators such as signal strength and transmission rate are normal.

[0067] Finally, the anchor chain is connected to the mooring anchor chain installation hole on the buoy steel frame. The suction bucket is transported to the designated location and penetrated into the seabed using specialized equipment. The other end of the anchor chain is then connected to the suction bucket to complete the mooring system installation.

[0068] In step S3, after all installations are completed, the entire inclined support coupled steel-concrete composite floating wind turbine foundation is fully debugged and tested. The firmness of each connection is checked, including the connection between the wind turbine and the tower, the tower and the pontoon, the inclined support and the tower and the pontoon, the heave plate and the pontoon, and the steel and concrete. The tension of the inclined support is tested to see if it meets the design requirements. The stability of the structure in simulated marine environments such as wind, waves, and currents is evaluated. The wind turbine is fully debugged, including blade rotation tests and power generation performance tests. The electrical equipment is fully tested to check whether power transmission and equipment operation are normal. The monitoring equipment is verified to ensure that the monitoring data is accurate and reliable. The communication quality of the communication equipment is tested to ensure stable data transmission. Ensure that the design requirements and safety standards are met.

[0069] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. An adjustable floating wind power foundation based on reinforcement learning, characterized in that: include: A wind turbine, wherein the bottom of the wind turbine is fixed to the tower, and a controller is installed inside the wind turbine; the bottom of the tower is fixedly connected to the floating base; a pendulum damper is provided inside the tower, and the pendulum damper includes a top suspension cross brace, a top suspension spring, a lateral limit spring, and a mass block; The floating base includes a first pontoon, a second pontoon, and a third pontoon arranged in a triangle; the bottom of the tower is fixedly connected to the first pontoon; the first pontoon, the second pontoon, and the third pontoon are fixedly connected via supporting steel pipes; the tower is fixedly connected to the second pontoon and the third pontoon via oblique supports; The first buoy, the second buoy and the third buoy are also connected to anchor chains, the other ends of the anchor chains are connected to the suction buckets and are anchored to the seabed through the suction buckets.

2. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The tower comprises, from outside to inside, a tower outer steel plate, a tower concrete interlayer and a tower inner steel plate.

3. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The interior of the buoy is a cavity, and the buoy arm is composed of a buoy outer steel plate, a buoy concrete interlayer and a buoy inner steel plate from the outside to the inside.

4. The adjustable floating wind power foundation based on reinforcement learning according to claim 3, characterized in that: The outer steel plate of the buoy and the inner steel plate of the buoy are fixed by a supporting frame.

5. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The bottom of the buoy is also connected by connecting ribs and heave plates.

6. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The interior of the buoy adopts a compartment sealing design.

7. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The adjustable floating wind turbine based on reinforcement learning is also fixed with a variety of sensors, which are connected to the internal controller of the wind turbine, including a six-degree-of-freedom motion sensor, a stress sensor, a displacement sensor, a wind speed sensor, and a wind direction sensor, which are used to obtain marine environment parameter data and motion parameter data.

8. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The oblique support is also provided with a pneumatic adjustment device, which is connected to the internal controller of the fan.

9. The adjustable floating wind power foundation based on reinforcement learning according to claim 1, characterized in that: The supporting steel pipe includes a first supporting steel pipe, a second supporting steel pipe, a third supporting steel pipe and a fourth supporting steel pipe; the first supporting steel pipe is fixed above any two side walls of the buoys; the second supporting steel pipe is arranged parallel to the first supporting steel pipe and is fixed below any two side walls of the buoys; the third supporting steel pipe and the fourth supporting steel pipe are fixedly connected between the first supporting steel pipe and the second supporting steel pipe, and the third supporting steel pipe, the fourth supporting steel pipe and the second supporting steel pipe form a triangular shape.

10. A modular construction method for an adjustable floating wind power foundation based on reinforcement learning according to any one of claims 1 to 9, characterized in that: include: Prefabrication stage, offshore assembly stage, commissioning and testing stage: S1, prefabrication stage; the prefabrication stage includes prefabrication of tower modules, buoy modules, heave plate modules and wind turbine modules in the factory; commissioning of electrical equipment, testing equipment and communication equipment; S2, offshore assembly phase: using large offshore vessels to transport the prefabricated tower modules, buoy modules, heave plate modules, and wind turbine modules to designated sea areas for connection. After assembly, electrical equipment, testing equipment, and communication equipment are installed and further debugged. The buoys and anchor chains are connected, the suction bucket is transported to the designated location, and inserted into the seabed to complete the installation of the mooring system. S3, debugging and testing phase; using deep reinforcement learning control system to optimize the parameters of the diagonal support and damper; testing the structural stability, power generation performance and data transmission function.

Images