Method and system for vibration reduction control of multi-body dynamic response of floating wind power system

Abstract

The present application relates to the field of wind power generation technology, and discloses a floating wind power system multi-body dynamic response damping control method and system, comprising: obtaining floating foundation pitch and flexible tower displacement data, establishing a global time domain control equation containing multiple types of loads; combined with the equation, executing two-stage cascading filtering, outputting positive damping variable pitch control instructions and hardware actuation parameters; adjusting the pitch angle by using the variable pitch mechanism, and simultaneously applying a reverse inertia counterforce to the tower top through the internal mass block; collecting dynamic response data, constructing a comprehensive objective function based on motion standard deviation and fatigue load, and running an adaptive optimization algorithm to dynamically correct parameters and close-loop iteration. The present application effectively suppresses the low-frequency pitch motion of the floating foundation through the cooperative control of software variable pitch and hardware actuation, reduces the equivalent fatigue load of the tower, and improves the adaptive damping performance and operation stability of the system in complex sea conditions through close-loop iteration.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, specifically to a multi-body dynamic response vibration reduction control method and system for floating wind power systems. Background Technology

[0002] Floating wind turbines, while operating at sea, are simultaneously subjected to coupled excitation from unstable wind and wave loads. Lacking rigid foundation support, floating wind power systems often exhibit large-amplitude, low-frequency rigid body motion of the floating foundation and high-frequency elastic vibration of the flexible tower. This multibody dynamic coupling response not only affects the power output quality of the turbine but also increases the stress on critical load-bearing components.

[0003] Under operating conditions above rated wind speed, existing conventional constant-power pitch control strategies are prone to inducing low-frequency resonance in floating wind power systems. When the floating foundation pitches backward with the waves, the relative inflow wind speed at the rotor plane decreases, causing a drop in generator speed. At this time, in order to restore the speed, the conventional pitch controller will actively reduce the pitch angle. This action increases the aerodynamic thrust on the rotor, further exacerbating the backward pitching amplitude of the platform. This physical process creates a negative aerodynamic damping effect, which can easily lead to system motion divergence.

[0004] Furthermore, existing vibration reduction solutions for multibody systems are often limited to single-dimensional interventions, such as modifying the propeller control logic only at the software level or adding structural dampers only at the hardware level, lacking a coordinated mechanism that integrates control commands with hardware inertial actuation. Moreover, the parameters of existing filtering algorithms and the physical parameters of structural dampers are mostly static preset values, unable to be dynamically and adaptively updated in the face of complex and changing actual sea conditions. The system struggles to adjust according to real-time instantaneous motion amplitude and cumulative fatigue damage, thus failing to achieve an effective balance between suppressing violent platform motion and extending the fatigue life of critical components. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a multi-body dynamic response vibration reduction and control method and system for floating wind power systems. This method solves the problem of system instability and shortened structural life caused by the coupling of severe swaying of the floating foundation and vibration fatigue of the flexible tower in complex wind and wave environments.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of this invention provides a multi-body dynamic response vibration reduction and control method for a floating wind power system, comprising the following steps: The instantaneous pitch angular velocity and pitch angular displacement of the floating foundation, as well as the dynamic displacement velocity and displacement of the top of the flexible tower, are obtained. The collected data are converted into dynamic displacement, velocity and acceleration vectors of the multibody flexible structure, and global time-domain control equations containing various loads and control forces are established. Receive pitch error input sequence or platform motion feedback signal, and perform two-level cascaded filtering in combination with the global time domain control equation to calculate and output positive damping pitch control command and hardware actuation parameters; According to the positive damping pitch control command, the pitch angle of the wind turbine blades is adjusted by the pitch drive mechanism, and at the same time, according to the hardware actuation parameters, the relative displacement is generated by the internal mass block, and an inertial counterforce in the opposite direction is applied to the top of the flexible tower. The system periodically collects and executes the pitch angle of the wind turbine blades and the multibody dynamic response data after applying the inertial counterforce action. It constructs a comprehensive objective function that includes the standard deviation of the rigid body motion response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower. It runs an adaptive boundary optimization algorithm to dynamically correct the control parameters and physical parameters in the hardware actuation parameters used for the second-level cascaded filtering, and re-inputs the corrected parameters for closed-loop iteration.

[0007] The core control principle of this invention lies in the decoupling and integration of active aerodynamic control and passive structural vibration reduction in the time domain. This is achieved through a two-stage cascaded filtering process. First, a low-pass filter is used to attenuate the high-frequency noise generated by the measurement. Second, a narrow-band notch filter is used to suppress the low-frequency negative damping component in the pitch control loop, so that the pitch action produces a positive damping adjustment effect to suppress the low-frequency rigid body oscillation of the platform caused by aerodynamic loads. At the same time, a mass block installed on the top of the flexible tower is used to generate relative displacement, and the resulting inertial impact force dissipates the elastic vibration energy of the tower.

[0008] Finally, by establishing a dual-objective optimization mechanism of motion response standard deviation and equivalent fatigue load, the mutual constraint between motion suppression and fatigue load during the control parameter adjustment process is overcome, so that the parameter optimization process converges to the minimum point of the comprehensive objective function, and closed-loop adaptive control of multibody dynamic response is realized.

[0009] Preferably, the step of converting the acquired data into dynamic displacement, velocity, and acceleration vectors of the multibody flexible structure includes: Based on the spatial geometric topological relationship between the components, a coordinate system transformation matrix is ​​established. The instantaneous pitch angular velocity, pitch angular displacement, and dynamic displacement velocity and displacement of the top of the flexible tower in the local coordinate system are uniformly mapped to a global reference coordinate system defined at the still water surface, with the initial overall center of gravity of the floating wind power system as the global coordinate origin.

[0010] Preferably, the step of establishing global time-domain control equations that include various loads and control forces includes: The generalized time-varying mass matrix, time-varying damping matrix, and stiffness matrix of the floating wind power system are multiplied by the dynamic acceleration vector, dynamic velocity vector, and dynamic displacement vector, respectively, and a balance relationship is established. The various loads include the calculated aerodynamic load vector, wave hydrodynamic load vector, and mooring nonlinear constraint reaction force vector; The physical parameters of mass, damping, and stiffness corresponding to the hardware actuation parameters are directly coupled to the time-varying damping matrix and stiffness matrix of the floating wind power system or included in the control force.

[0011] Preferably, in the step of performing two-stage cascaded filtering, the first-stage filtering step includes: The adjustment coefficient is determined by combining the discrete sampling period of the control loop with the set cutoff frequency. The current pitch error input sampling value and the previous filter history output value are input into the time-domain differential control equation of the first-order discrete low-pass filter for processing, and the corresponding reference control sequence is output. The cutoff frequency is set within a frequency band that is lower than the high-frequency measurement noise frequency of the sensor and higher than the basic response frequency of the pitch control loop.

[0012] Preferably, in the step of performing two-level cascaded filtering, the two-level filtering steps include: The internal control parameters that satisfy the natural logarithmic relationship and the notch bandwidth of the notch filter are set. Combined with the discrete sampling period and the notch center angular frequency of the floating foundation, the time-domain difference formula of the discrete narrowband notch filter is run using the current and historical input reference control sequence sample values ​​and the historical calculation values ​​of the output command to calculate and generate the positive damping pitch control command.

[0013] Preferably, the step of adjusting the pitch angle of the wind turbine blades using the pitch drive mechanism according to the positive damping pitch control command includes: The bottom-level pitch servo mechanism interprets the positive damping pitch control command as the target pitch angle to drive the blade deflection. When it detects that the floating foundation is tilting forward at a low frequency, it drives the blade to deflect in the feathering direction and reduces the lift coefficient of the blade airfoil section and the aerodynamic thrust acting on the wind turbine.

[0014] Preferably, the step of applying an inertial counterforce in the opposite direction to the top of the flexible tower by generating a relative displacement through an internal mass block according to the hardware actuation parameters includes: The dynamic inertial mass of the internal mass block is set according to the equivalent modal mass of the wind turbine tower, and the tuning stiffness and damping coefficient in the hardware actuation parameters are matched in combination with the first natural frequency of the flexible tower. When external excitation induces vibration, the internal mass block generates a relative displacement opposite to the direction of movement of the tower top based on the tuned stiffness and damping. The inertial force of the internal mass block and the frictional force of the damper element are combined to form the inertial counterforce.

[0015] Preferably, the step of constructing a comprehensive objective function that includes the standard deviation of the rigid body motion response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower includes: The standard deviation of the rigid body motion response is calculated using the time-domain sampling sequence within the specified evaluation window period. The equivalent fatigue load is calculated using the time-domain bending moment sequence collected by the strain sensor installed at the bottom section of the tower, through the rainflow counting method and linear damage theory. By introducing a dimensional normalization factor and a physical weighting coefficient, the extracted standard deviation of the rigid body motion response and the equivalent fatigue load are respectively compared with the benchmark value when vibration reduction control is not enabled. A dual-objective collaborative optimization mechanism is established, and a corresponding comprehensive objective function is constructed.

[0016] Preferably, the step of dynamically correcting the control parameters and physical parameters in the hardware actuation parameters using an adaptive boundary optimization algorithm includes: Run the built-in multi-objective particle swarm optimization algorithm or grid step optimization algorithm, and use the minimum value of the comprehensive objective function as the convergence criterion. Within the preset parameter space, the center angular frequency and notch bandwidth of the second-order discrete notch filter, as well as the tuning stiffness and damping coefficient in the physical parameters, are updated by linear step scan.

[0017] A second aspect of the present invention provides a multi-body dynamic response vibration reduction and control system for a floating wind power system, comprising: The sensing module is used to acquire the instantaneous pitch angular velocity and pitch angular displacement of the floating foundation, as well as the dynamic displacement velocity and displacement of the top of the flexible tower. It converts the acquired data into dynamic displacement, velocity and acceleration vectors of the multi-body flexible structure and establishes global time-domain control equations that include various loads and control forces. The calculation module is used to receive the pitch error input sequence or platform motion feedback signal, and perform two-level cascaded filtering in combination with the global time domain control equation to calculate and output the positive damping pitch control command and hardware actuation parameters. The execution module is used to adjust the pitch angle of the wind turbine blades using the pitch drive mechanism according to the positive damping pitch control command, and at the same time generate relative displacement through the internal mass block according to the hardware actuation parameters, and apply an inertial counterforce in the opposite direction to the top of the flexible tower. The optimization module is used to periodically collect multibody dynamic response data after the adjustment of the wind turbine blade pitch angle and the application of the inertial counterforce action. It constructs a comprehensive objective function that includes the standard deviation of the rigid body motion response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower. It runs an adaptive boundary optimization algorithm to dynamically correct the control parameters and physical parameters in the hardware actuation parameters used for the second-level cascaded filtering. The corrected parameters are then re-input into the calculation module and the execution module for closed-loop iteration.

[0018] This invention provides a multi-body dynamic response vibration reduction and control method and system for floating wind power systems. It has the following beneficial effects: 1. This invention constructs a two-stage cascaded filtering algorithm that includes a first-order discrete low-pass filter and a second-order discrete notch filter. Combined with the discrete sampling period of the system, it accurately locks the platform's rigid body resonant frequency, effectively removing the speed fluctuation component induced by the platform's low-frequency pitch motion. It cuts off the low-frequency negative feedback loop of traditional pitch control on platform motion, transforms the traditional pitch strategy into a positive damping pitch control loop, eliminates the aerodynamic negative damping effect when operating in the super-rated wind speed area, and enhances the instantaneous limit stability of the system under wind and wave excitation.

[0019] 2. This invention integrates the aerodynamic decoupling of positive damping variable pitch control with the physical hardware actuation of the tuned mass damper, achieving complementary advantages between software algorithms and hardware devices. At the rotor level, the system actively releases the aerodynamic load acting on the rotor by adjusting the pitch angle, avoiding the tower from being forced to generate excessive bending moment. At the structural level, the physical inertial displacement of the internal mass block is used to apply a reaction inertial counterforce in the opposite direction to the top of the flexible tower, passively absorbing and dissipating the vibration energy of the tower at the first elastic modal frequency, thus completing the load reduction of the overall dynamic response.

[0020] 3. This invention calculates the standard deviation of the rigid body response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower by collecting time-domain multibody dynamic response data. It constructs a dual-objective collaborative optimization comprehensive objective function with dimensionless normalization factor, and runs an adaptive boundary optimization algorithm in a normalized manner during continuous operation cycles. This breaks through the limitations of traditional emergency control, and can dynamically correct and update the control parameters of the filter and the physical parameters of the damper online. It finds the optimal balance between suppressing large-scale platform sway and extending the long-term fatigue life of structural components, thus ensuring the long-term safe operation of the unit. Attached Figure Description

[0021] Figure 1 This is a structural block diagram of a multi-body dynamic response vibration reduction control system for a floating wind power system provided in an embodiment of the present invention; Figure 2This is a flowchart of a multi-body dynamic response vibration reduction control method for a floating wind power system provided in an embodiment of the present invention; Figure 3 This is a bar chart comparing performance before and after multi-index optimization provided by an embodiment of the present invention; Figure 4 This is a time-domain dynamic response comparison curve of the pitch angle of a floating foundation provided by an embodiment of the present invention; Figure 5 This is a convergence trend curve of the comprehensive objective function with the number of iteration steps provided by an embodiment of the present invention. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] See attached document Figure 1 This invention provides a multi-body dynamic response vibration reduction and control system for a floating wind power system, comprising: The sensing module is used to collect dynamic rigid body motion response parameters of the floating foundation, constraint reaction force data of the anchor chain, and elastic deformation rate, displacement and local dynamic time-domain load of the multi-body structure. The computing module is used to receive the digital electrical signals transmitted by the sensing module and to process the time-domain signals and calculate the vibration reduction control commands in real time through the internally cascaded filtering algorithm. The execution module is used to receive vibration reduction control commands output by the calculation module and adjust the pitch angle of the wind turbine blades. At the same time, it absorbs the vibration energy of the multi-body structure through the physical inertial displacement of the internal mass block. The optimization module is used to continuously and periodically collect multibody dynamic response data after the vibration reduction control action is executed, run an adaptive boundary optimization algorithm, dynamically adjust the center angular frequency and notch bandwidth of the second-order notch filter in the calculation module, and the tuning stiffness and damping coefficient of the hardware execution unit in the execution module.

[0024] See attached document Figure 2 This invention provides a multi-body dynamic response vibration reduction control method for floating wind power systems, comprising the following steps: The instantaneous pitch angular velocity and pitch displacement of the floating foundation, as well as the dynamic displacement velocity and displacement of the top of the flexible tower, are obtained. The collected data are converted into dynamic displacement, velocity and acceleration vectors of the multibody flexible structure, and a global time-domain control equation containing various loads and control forces is established. The various loads include aerodynamic loads, wave hydrodynamic loads and mooring nonlinear constraint reaction forces. The system receives pitch error input sequences or platform motion feedback signals, and performs two-stage cascaded filtering in conjunction with the established global time-domain control equations. The first-stage filter processes the received sequence or signal using a first-order discrete low-pass filter equation and outputs a smoothed reference control sequence. The second-stage filter inputs the reference control sequence into a second-order discrete notch filter. Combining the discrete sampling period of the control system with the notch center angular frequency and notch bandwidth of the floating foundation in the runtime difference equation, the system calculates and outputs positive damping pitch control commands and hardware actuation parameters. The wind turbine blades are adjusted by using the pitch control command and hardware actuation parameters to perform actions based on the output positive damping pitch control command and hardware actuation parameters. At the same time, relative displacement is generated through the internal mass block to apply an inertial counterforce in the opposite direction to the top of the flexible tower. During continuous operation, multibody dynamic response data after the execution of actions is collected periodically and routinely. The standard deviation of the rigid body response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower are calculated. Based on this, a comprehensive objective function including the standard deviation of motion response and equivalent fatigue load is constructed. An adaptive boundary optimization algorithm is run to solve for the optimal control and physical parameters that minimize the comprehensive objective function. The notch center angular frequency, notch bandwidth, tuning stiffness or damping coefficient are dynamically corrected. The corrected parameters are then re-input into the two-level cascaded filtering and hardware execution unit for closed-loop iteration.

[0025] The aforementioned multibody dynamic response vibration reduction control system for floating wind power systems corresponds to the aforementioned multibody dynamic response vibration reduction control method for floating wind power systems.

[0026] Specifically, the aforementioned sensing module performs the steps of acquiring data and establishing global time-domain control equations in the aforementioned method; the aforementioned calculation module performs the steps of performing two-stage cascaded filtering and outputting parameters in the aforementioned method; the aforementioned execution module performs the steps of adjusting the pitch angle and applying inertial counterforce in the aforementioned method; and the aforementioned optimization module performs the steps of calculating response data and correcting parameters for iteration in the aforementioned method.

[0027] In the multibody dynamic response vibration reduction control method for a floating wind power system provided in this embodiment, the steps of obtaining the instantaneous state of the floating foundation and establishing the global time-domain control equation are specifically implemented in the following way.

[0028] This embodiment performs synchronous time-domain sampling using sensors installed on the multi-body structure of the floating wind power system. Specifically, an inertial measurement unit and a global positioning system installed on the main structure of the floating foundation are used to acquire the instantaneous pitch angular velocity and pitch angular displacement of the floating foundation under wave excitation according to a set sampling period. An accelerometer installed on the top of the flexible tower acquires the dynamic acceleration signal of the flexible tower top, and the underlying processor performs integration algorithm and high-pass filtering reconstruction processing in the time domain to eliminate integration drift, thereby converting it into the dynamic displacement velocity and displacement of the flexible tower top.

[0029] To ensure strict alignment of the output data from each spatial position sensor on the time axis, each sensor is connected to the underlying processor via an industrial fieldbus. Hardware sampling is triggered by a unified global clock signal, outputting digital electrical signals with a fixed time step. The sampling period is determined according to the Nyquist sampling theorem, and the sampling frequency must be at least twice the high-frequency boundary of the rigid body resonance of the floating foundation and the first-order elastic mode frequency of the tower. In this embodiment, the sampling frequency is set between 20Hz and 50Hz to cover the target control frequency band of the multibody dynamic response.

[0030] After acquiring local measurement data, the measurement signals collected in the local coordinate system need to be converted into multibody dynamic state variables in the global reference coordinate system. Since the sensors are distributed in different spatial locations such as the floating foundation and the flexible tower, the output raw data resides in their respective independent local coordinate systems. In this embodiment, the underlying processor establishes a coordinate system transformation matrix based on the spatial geometric topological relationships between the various components, uniformly mapping the instantaneous pitch angular velocity, pitch angular displacement, and dynamic displacement velocity and displacement at the top of the tower to the global reference coordinate system defined at the still water surface. The coordinate system transformation matrix is ​​constructed using Euler angles or direction cosine matrices, with the initial overall center of gravity of the system at the still water surface as the origin of the global coordinate system. The transformed data is uniformly encapsulated as a dynamic displacement vector of the multibody flexible structure. Dynamic velocity vector and dynamic acceleration vector The dimension of the above vector depends on the number of degrees of freedom retained by the floating wind power system, covering the six-degree-of-freedom rigid body motion of the floating foundation and the elastic modal coordinates of the flexible tower.

[0031] During system operation, external environmental excitation loads and internal nonlinear constraint forces are acquired synchronously. Tension sensors installed at the anchor chain guide hole section read the anchor chain constraint reaction force data. Combined with the geometric relationship of the mooring system layout, the mooring nonlinear constraint reaction force vector is calculated by solving the catenary equation or referring to a preset stiffness matrix table. Local dynamic time-domain loads are acquired using strain gauges at the blade root. Combined with the current wind speed and blade pitch angle, the aerodynamic load vector acting on the entire rotor is calculated. Based on the real-time wave elevation and the wetted surface area of ​​the floating foundation, the wave hydrodynamic load vector is calculated.

[0032] After obtaining the motion state vector and external loads, a global time-domain control equation for the multibody flexible structure is established. The obtained motion state vector and various external loads are integrated into a unified mathematical framework to form the physical receptor boundary model for subsequent filtering and vibration reduction control solutions. The specific expression of the global time-domain control equation constructed in this embodiment is as follows: ; In the formula, Represents the generalized time-varying mass matrix of the system; This represents the time-varying damping matrix of the system; Represents the stiffness matrix of the system; , , These represent the dynamic displacement vector, dynamic velocity vector (i.e., the first derivative of displacement with respect to time), and dynamic acceleration vector (i.e., the second derivative of displacement with respect to time) of the multibody flexible structure, respectively. The variable is time; the right side of the equation represents the generalized external load term, where... For aerodynamic load vector, The wave hydrodynamic load vector. The nonlinear constraint reaction force vector of the mooring. A vector of active or passive control force applied by a physical actuator; The system control input is specifically the pitch control command of the wind turbine; while the physical parameters of the hardware execution unit (tuned mass damper), such as mass, damping, and stiffness, are directly coupled to the system's generalized time-varying damping and stiffness matrices, or are incorporated into the active / passive control force vector by calculating the counterforce generated by the damper. middle.

[0033] Because floating wind turbine systems lack rigid foundation support, the pitch displacement of the floating foundation directly alters the relative inflow velocity at the rotor plane, thereby causing aerodynamic loads. The platform experiences fluctuations; simultaneously, changes in aerodynamic loads act as excitation forces on the foundation, exacerbating platform displacement. The purpose of establishing the global time-domain control equations is to transform the aforementioned coupled physical processes of aerodynamics, hydrodynamics, and elasticity into a controlled mathematical model. In the global time-domain control equations, the values ​​of the elements of the mass matrix and damping matrix are updated with the coupling of large-scale rigid body displacement and local elastic deformation in the structural space, exhibiting time-varying and nonlinear characteristics. The global time-domain control equations constructed in this embodiment incorporate external wind and wave excitation, platform rigid body motion, tower elastic deformation, and mooring constraints into the same matrix system, providing a data computation basis for subsequent extraction of resonance energy in specific low-frequency bands.

[0034] The steps for executing a two-stage cascaded filter and outputting a positive damping pitch control command are implemented in the following manner.

[0035] During the operation of floating wind turbines, the main frequency of conventional pitch control is typically above 0.3Hz. However, due to the introduction of rigid body motion in floating wind turbine systems, the main frequency of multi-body low-frequency resonance in pitch and roll is usually below 0.1Hz. When operating in wind speeds exceeding the rated range, the feedback loop of traditional pitch strategies responds to low-frequency motion and actively increases the pitch angle when the platform tilts forward, resulting in reduced aerodynamic thrust and a negative aerodynamic damping effect. From a physical aerodynamic response perspective, when the floating foundation pitches backward with the waves, the relative wind speed of the rotor plane to the inflow wind speed decreases, causing a drop in generator speed. Traditional constant power pitch controllers reduce the pitch angle to restore speed, which increases the aerodynamic thrust of the rotor, further exacerbating the backward pitch amplitude of the platform and causing system divergence or low-frequency resonance. In this embodiment, to cut off the low-frequency positive feedback loop, the control system receives the pitch error input sequence or platform motion feedback signal and calls the built-in digital filtering algorithm to perform two-stage cascaded filtering.

[0036] The control system uses a first-order discrete low-pass filter equation to process the received sequence or signal, smoothing high-frequency measurement noise and outputting a smoothed reference control sequence. In this embodiment, the time-domain differential control equation of the first-order discrete low-pass filter is expressed as: ; In the formula, Input the sampled value for the pitch error at the current moment; This is the first-order filter output value at the current moment; This is the historical output value of the first-order filter from the previous time step; This is the adjustment coefficient. In this embodiment, the adjustment coefficient... Combined with the discrete sampling period of the control system With the set cutoff frequency Calculation determined the cutoff frequency. The value is lower than the high-frequency measurement noise frequency of the sensor and higher than the basic response frequency of the pitch control loop. It is set between 0.1Hz and 0.3Hz to ensure that high-frequency interference signals are filtered out without affecting the basic adjustment function of the pitch system.

[0037] After the smoothed reference control sequence is output through primary filtering, the control system uses this smoothed reference control sequence as input and cascades it into a second-order discrete notch filter. For floating wind turbines, the platform's pitch or roll natural frequencies are low, and low-frequency resonance signals can pass through a low-pass filter. Therefore, this embodiment uses a discrete narrowband notch filter to filter low-frequency resonance signals and avoid causing system resonance. Combined with the underlying control logic, the discrete narrowband notch filter in this embodiment is expressed by the following time-domain difference formula: ; In the formula, The notch center angular frequency of the platform's low-frequency rigid body resonance; The discrete sampling period of the control system; The reference control sequence sample value is the input to the discrete narrowband notch filter at the current moment; and These are the historical sampled values ​​of the reference control sequence input at the previous time step and the time step before that, respectively; This is the output command value after filtering and calculation at the current moment; and These are the historical calculated values ​​of the output commands from the previous time step and the time step before that, respectively. These are the internal control parameters of the notch filter.

[0038] Internal control parameters in time-domain difference formula Notch bandwidth of notch filter It satisfies the following natural logarithm relation: ; During parameter tuning, to ensure the numerical stability of the discrete digital filter in the time domain, internal control parameters are used. The value range is limited to the interval greater than 0 and less than 1. Notch bandwidth. The size of the bandwidth determines the frequency filtering range. Choosing too large a bandwidth value will cause phase lag in the pitch response, while choosing too small a value will fail to cover platform frequency drift caused by sea state changes. In this embodiment, considering the floating body's motion characteristics, the notch bandwidth... The internal control parameters are configured between 0.02 rad / s and 0.08 rad / s, and thus obtained by inverse solving. .

[0039] By setting the notch bandwidth Inverse solution to obtain internal control parameters And combined with the notch center angular frequency of the floating foundation A discrete narrowband notch filter is used to filter low-frequency components of the platform's pitch in the input signal by the pitch controller. After removing the speed fluctuation components induced by the platform's low-frequency pitch motion, the final output control command eliminates the reverse adjustment information that leads to thrust degradation. The control system encapsulates the calculated positive damping pitch control command with the hardware actuation parameters and outputs it, cutting off the negative feedback of pitch control to platform motion and converting the traditional pitch control loop into a positive damping pitch control loop. The calculated positive damping pitch control command is then transmitted to the underlying hardware execution node, providing data support for physical-level energy reduction actions.

[0040] In the multi-body dynamic response vibration reduction control method for floating wind power systems provided in this embodiment, the steps regarding the physical execution of wind turbine aerodynamic decoupling and tower inertial counterbalancing are specifically implemented in the following ways.

[0041] After receiving the positive damping pitch control command output by the control system, the underlying pitch servo mechanism parses the command data and drives the blades to perform pitch maneuvers. In this embodiment, the pitch servo system receives the positive damping pitch control command as the target pitch angle and changes the deflection angle of the wind turbine blades around the pitch axis through a hydraulic or electric drive mechanism. The deflection of the pitch angle directly changes the aerodynamic angle of attack of the wind turbine blade section, cutting off the low-frequency positive feedback loop that causes a sudden increase in platform forward tilt thrust. From an aerodynamic perspective, when the platform tilts forward at low frequencies, the positive damping pitch control command drives the blades to deflect in the feathering direction. The lift coefficient of the blade airfoil section decreases as the angle of attack decreases, thereby actively releasing the aerodynamic thrust acting on the wind turbine and preventing the tower from being forced to generate an excessive forward bending moment. Through the above actions, aerodynamic decoupling of the wind turbine is achieved.

[0042] While the pitch system performs aerodynamic decoupling of the wind turbine, the hardware structure dynamic vibration reduction system simultaneously performs inertial counterbalancing of the tower. Under the coupled excitation of wave loads and wind loads, the flexible tower will generate structural bending and fatigue stress. In this embodiment, the hardware is equipped with a tuned mass damper or a tuned liquid column damper installed on the top of the flexible tower or inside the nacelle. Taking the tuned mass damper as an example, the elastic coupling vibration control equation of the tuned mass damper and the wind turbine multibody system is characterized as follows: ; Based on the principles of conservation of momentum and action-reaction, the tuned mass damper applies a reaction inertial counterforce to the top of the flexible tower. Expressed as: ; In the formula, , , These represent the dynamic inertial mass, damping coefficient, and tuning stiffness of the tuned mass damper, respectively. , and These represent the displacement, velocity, and acceleration of the internal mass block of the tuned mass damper relative to the ground. and These represent the elastic displacement and velocity of the multibody flexible degrees of freedom at the top of the wind turbine tower in the time domain; The reaction inertial counterforce applied to the top of the tower by the tuned mass damper.

[0043] In terms of hardware parameter configuration, dynamic inertial mass The value is determined based on the modal mass distribution of the entire wind turbine. In this embodiment, the dynamic inertial mass... The stiffness is set between 1% and 5% of the equivalent modal mass of the wind turbine tower to avoid adding extra top weight that could cause structural overload. (Tune stiffness) With damping coefficient The first-order natural frequency of the flexible tower is used for calculation and matching to ensure that the natural frequency of the damper is aligned with the main resonant frequency of the tower. The corresponding optimal tuning damping ratio is configured between 0.05 and 0.15.

[0044] The calculated reaction inertial impulse force The control force vector is directly coupled into the previously established global time-domain control equations. From a mechanical dynamics perspective, when external wind and wave excitation causes the tower to vibrate forward and backward, due to the hysteresis effect of the tuned stiffness and damping, the mass block inside the tuned mass damper will generate a relative displacement opposite to the direction of the tower top's movement. During this relative motion, the inertial force generated by the mass block and the frictional force of the damper elements form a shear force opposite to the tower's velocity. This is the reaction inertial counterforce. The system passively absorbs and dissipates first-order vibration energy in the forward and backward directions of the tower. This embodiment integrates the aerodynamic decoupling of the control signal of the pitch system with the inertial counterweight of the hardware vibration damper to achieve structural load reduction and attenuation of the overall dynamic response of the floating wind turbine system.

[0045] In the multibody dynamic response vibration reduction control method for floating wind power systems provided in this embodiment, the steps of multibody dynamic response evaluation and closed-loop optimization iteration are specifically implemented in the following way.

[0046] To achieve closed-loop optimization iteration of the control parameters of the second-order discrete notch filter and the physical parameters of the tuned mass damper, a multi-body response evaluation index is introduced into the control system to establish a dual-objective collaborative optimization mechanism. The control parameters include the notch filter center angular frequency. and notch bandwidth Physical parameters include the damping coefficient of the tuned mass damper. With tuning stiffness .

[0047] Analyzing from the perspective of multibody dynamics collaborative control principles, the standard deviation of the platform's rigid body oscillation motion response reflects the instantaneous limit stability state of the unit, while the equivalent fatigue load of structural dynamics determines the long-term fatigue life of key load-bearing components. During control parameter adjustment, excessive suppression of platform motion often leads to frequent and violent actuation of the pitch actuator or tuned mass damper, thereby increasing the local fatigue load at the tower base. Conversely, excessive pursuit of reducing structural fatigue may allow the platform to oscillate excessively, potentially causing the entire unit to overturn. Therefore, this embodiment establishes a dynamic balance between the standard deviation of the motion response and the equivalent fatigue load through closed-loop optimization iteration.

[0048] The control system uses the standard deviation of motion response to assess the severity of the swaying in the rigid body multibody response of the platform. In this embodiment, the standard deviation of motion response... The mathematical formula for calculation is expressed as follows: ; In the formula, The standard deviation of the motion response; The time-domain sampling sequence of the dynamic response is specifically obtained by using the instantaneous pitch or roll angle sampling data of the floating foundation. This represents the statistical mean of the time-domain sampled sequences within the current statistical period. This represents the total number of samples sampled in the time domain. In specific implementations, the total number of samples sampled in the time domain... The selection is determined based on the safety assessment wind condition time period specified in the International Electrotechnical Commission (IEC) standards. Typically, a single assessment time window is set to 600 seconds, combined with a discrete sampling frequency of 20Hz to 50Hz, and the total number of time-domain samples... The value ranges from 12,000 to 30,000 to ensure that the statistical characteristics can fully cover the low-frequency resonance cycle.

[0049] While evaluating the rigid body motion of the platform, the control system utilizes structural dynamics equivalent fatigue loads to assess the long-term vibration reduction lifespan extension characteristics. Structural dynamics equivalent fatigue loads, as a core indicator for evaluating long-term vibration reduction effectiveness, are calculated by inverting the calculations of key load-bearing components using the rainflow counting method and linear damage theory. In this embodiment, the key load-bearing component is specifically selected as the critical section at the bottom of the wind turbine tower. Equivalent fatigue loads The mathematical expression is: ; In the formula, The equivalent fatigue load of the critical section under specific working conditions; For specific working conditions The first one extracted by the rainflow counting method The actual number of cycles for each load amplitude range; For the first The nonlinear load amplitude corresponding to each cycle is obtained by separating the time-domain bending moment sequence collected by the strain sensor installed at the bottom section of the tower using the rainflow counting method; The Wöhler fatigue inverse slope index is the reciprocal of the slope of the curve representing the relationship between material stress amplitude and failure cycle number in engineering technology. When the key load-bearing component is a steel tower structure, ... The value is usually 3 or 4. When the evaluation object is changed to a composite material blade... The value is 10; The equivalent benchmark cycle frequency specified in the standard is typically set to 1×10⁻⁶ within the assessment cycle. 7 .

[0050] Obtaining the standard deviation of motion response With equivalent fatigue load Subsequently, the control system constructs a comprehensive objective function aimed at minimizing both the motion response standard deviation and the equivalent fatigue load. Since the motion response standard deviation and the equivalent fatigue load differ in physical dimensions and numerical magnitudes, to avoid a single index dominating the optimization result due to an excessively large base value, this embodiment introduces a dimensional normalization factor to establish the comprehensive objective function. The mathematical expression of the comprehensive objective function is as follows: ; In the formula, To comprehensively evaluate the index values; The initial platform motion response standard deviation baseline value without the two-stage cascaded filtering and hardware structure dynamic vibration reduction system enabled; This corresponds to the initial equivalent fatigue load reference value at the bottom of the tower; and The preset physical weighting coefficients, and satisfying The specific proportions are adjusted according to the control requirements of the unit, which prioritize safety and stability or structural lifespan.

[0051] The control system uses a built-in multi-objective particle swarm optimization algorithm or grid step optimization algorithm to iteratively solve the comprehensive objective function. The algorithm uses a comprehensive evaluation index value. Reaching the minimum value within the current iteration step is the convergence criterion. During the parameter optimization iteration process, the control system performs a scan update within the preset parameter search space.

[0052] The specific optimization boundary condition is set as follows: notch center angular frequency. The search range is limited to between 0.03 rad / s and 0.15 rad / s, and the notch bandwidth is... The search range is limited to 0.02 rad / s to 0.08 rad / s; the tuning stiffness of the tuned mass damper. With damping coefficient The optimization range is linearly scanned around the physical parameters corresponding to the first-order natural frequency of the tower. Through the coordinated operation of software algorithms and hardware devices, the control system calculates and updates the optimal parameter combination, thereby completing the vibration reduction control of the multibody dynamic response of the floating wind power system.

[0053] Specific application examples: This embodiment uses a 15MW semi-submersible floating wind turbine as an example to demonstrate the specific numerical calculation process of the two-stage cascaded filtration and hardware inertial counterbalancing collaborative control of the present invention.

[0054] System parameter initialization and data acquisition: The discrete sampling frequency is set to 20Hz, which is the discrete sampling period. Based on the information collected by the sensing module and prior structural information: Low-frequency rigid body pitch resonance center angular frequency of floating body The first-order natural frequency of the flexible tower is... (about ), tower equivalent modal mass .

[0055] Numerical calculation of the two-stage cascaded filtering algorithm (taking a certain time step as an example): Set the cutoff frequency of the first-order discrete low-pass filter. According to digital control theory, the adjustment coefficient is set to... .

[0056] Assuming the received pitch error input sample value at the current moment The historical output value of the first-level filter at the previous moment. .

[0057] Substituting into the time-domain differential control equation of the first-order discrete low-pass filter: At this point, the sampled values ​​of the first-order smoothed reference control sequence output... .

[0058] Enter the second-stage narrowband notch filter and set the notch bandwidth. .

[0059] Internal control parameters can be solved by inverse equation. : ; Assuming the historical input value at this moment Historical output commands .

[0060] Substitute .

[0061] Substituting into the time-domain difference formula of a second-order discrete notch filter: ;

[0062] The calculation result is the positive damping pitch control command output after removing the low-frequency oscillation component.

[0063] Numerical calculation of hardware inertial hedging (tuned mass damper): The dynamic inertial mass is taken as 3% of the tower modal mass, that is .

[0064] Combined with the first-order natural frequency of the tower ( Matching tuning stiffness .

[0065] Set the optimal damping ratio Matching damping coefficient .

[0066] At a certain moment of excitation, the sensor measures the displacement of the internal mass block relative to the tower. relative speed Calculate the reaction inertial impulse force: ; The backward-applied shear force will be coupled as a control force to the global time-domain control equations, absorbing the forward tilting kinetic energy of the tower.

[0067] The comprehensive objective function is calculated within the evaluation window. (sampling points) Baseline state (this invention is not enabled): Equivalent fatigue load at the base of the tower .

[0068] Current iteration state (with this invention enabled): The current iteration state is calculated using the motion response standard deviation formula. Using rainflow counting to invert current .

[0069] Selecting weighting coefficients Substituting into the comprehensive objective function: ; because This indicates that the method not only achieves dual vibration reduction, but the adaptive boundary optimization algorithm will also continue to update the control and physical parameters to a smaller value region until convergence.

[0070] Experimental verification and effect comparison: according to Figure 3The performance comparison results before and after multi-index optimization show that, under the same wind and wave environment excitation, the control method of this invention has a reducing effect on both the floating foundation motion and the tower structure load. Data shows that, compared with the baseline control group without vibration reduction control, the standard deviation of the floating foundation pitch angle in the experimental group of this invention is 49.9% of the baseline value, the maximum pitch angle is reduced to 52.6% of the baseline value, and the equivalent fatigue load on the key section at the bottom of the flexible tower is reduced to 66.1% of the baseline value. This indicates that by adjusting the low-frequency components in the pitch circuit through a two-stage discrete narrowband notch filter, and in conjunction with the inertial counterforce of the hardware execution unit, the platform's sway amplitude and the fatigue load on the tower structure can be reduced.

[0071] Figure 4 The time-domain dynamic response of the floating foundation's pitch angle is demonstrated over a testing period ranging from 200 to 800 seconds. The baseline control group, due to aerodynamic negative damping and wave coupling, exhibits a low-frequency oscillating pitch angle, with a local maximum value reaching 9.80°. However, after employing the method of this invention, the pitch angle fluctuation range shown by the solid line is reduced throughout the entire time period, with the maximum extreme value limited to within 5.15°. The comparison of the time-domain waveforms reflects that this invention, by adjusting the turbine blade pitch angle and the relative displacement generated by the hardware mass block, can effectively compensate for the platform's motion in the time domain.

[0072] Figure 5 The convergence trajectory of the comprehensive objective function during the optimization iteration process is presented. In the early stages of the algorithm, since the control parameters and physical parameters have not yet achieved optimal matching, the comprehensive evaluation index remains at the initial baseline value of 1.00. As the algorithm scans and updates the parameter space using the minimization of this comprehensive index value as the criterion, the index value gradually decreases with the increase of the iteration step, stabilizing at 0.68 after the 25th step. This trend indicates that the adaptive boundary optimization algorithm can enable the system to adaptively approach the optimal parameters of the dual-objective equilibrium, realizing closed-loop tuning and adaptive control of the multibody dynamic response.

Claims

1. A multi-body dynamic response vibration reduction and control method for floating wind power systems, characterized in that, Includes the following steps: The instantaneous pitch angular velocity and pitch angular displacement of the floating foundation, as well as the dynamic displacement velocity and displacement of the top of the flexible tower, are obtained. The collected data are converted into dynamic displacement, velocity and acceleration vectors of the multibody flexible structure, and global time-domain control equations containing various loads and control forces are established. Receive pitch error input sequence or platform motion feedback signal, and perform two-level cascaded filtering in combination with the global time domain control equation to calculate and output positive damping pitch control command and hardware actuation parameters; According to the positive damping pitch control command, the pitch angle of the wind turbine blades is adjusted by the pitch drive mechanism, and at the same time, according to the hardware actuation parameters, the relative displacement is generated by the internal mass block, and an inertial counterforce in the opposite direction is applied to the top of the flexible tower. The system periodically collects and executes the pitch angle of the wind turbine blades and the multibody dynamic response data after applying the inertial counterforce action. It constructs a comprehensive objective function that includes the standard deviation of the rigid body motion response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower. It runs an adaptive boundary optimization algorithm to dynamically correct the control parameters and physical parameters in the hardware actuation parameters used for the second-level cascaded filtering, and re-inputs the corrected parameters for closed-loop iteration.

2. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, The steps of converting the collected data into dynamic displacement, velocity, and acceleration vectors of a multibody flexible structure include: Based on the spatial geometric topological relationship between the components, a coordinate system transformation matrix is ​​established. The instantaneous pitch angular velocity, pitch angular displacement, and dynamic displacement velocity and displacement of the top of the flexible tower in the local coordinate system are uniformly mapped to a global reference coordinate system defined at the still water surface, with the initial overall center of gravity of the floating wind power system as the global coordinate origin.

3. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, The steps for establishing global time-domain control equations that include various loads and control forces include: The generalized time-varying mass matrix, time-varying damping matrix, and stiffness matrix of the floating wind power system are multiplied by the dynamic acceleration vector, dynamic velocity vector, and dynamic displacement vector, respectively, and a balance relationship is established. The various loads include the calculated aerodynamic load vector, wave hydrodynamic load vector, and mooring nonlinear constraint reaction force vector; The physical parameters of mass, damping, and stiffness corresponding to the hardware actuation parameters are directly coupled to the time-varying damping matrix and stiffness matrix of the floating wind power system or included in the control force.

4. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, In the process of performing a two-stage cascaded filter, the first-stage filtering steps include: The adjustment coefficient is determined by combining the discrete sampling period of the control loop with the set cutoff frequency. The current pitch error input sampling value and the previous filter history output value are input into the time-domain differential control equation of the first-order discrete low-pass filter for processing, and the corresponding reference control sequence is output. The cutoff frequency is set within a frequency band that is lower than the high-frequency measurement noise frequency of the sensor and higher than the basic response frequency of the pitch control loop.

5. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 4, characterized in that, In the process of performing a two-level cascaded filter, the two-level filtering steps include: The internal control parameters that satisfy the natural logarithmic relationship and the notch bandwidth of the notch filter are set. Combined with the discrete sampling period and the notch center angular frequency of the floating foundation, the time-domain difference formula of the discrete narrowband notch filter is run using the current and historical input reference control sequence sample values ​​and the historical calculation values ​​of the output command to calculate and generate the positive damping pitch control command.

6. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, The steps of adjusting the pitch angle of the wind turbine blades using the pitch drive mechanism according to the positive damping pitch control command include: The bottom-level pitch servo mechanism interprets the positive damping pitch control command as the target pitch angle to drive the blade deflection. When it detects that the floating foundation is tilting forward at a low frequency, it drives the blade to deflect in the feathering direction and reduces the lift coefficient of the blade airfoil section and the aerodynamic thrust acting on the wind turbine.

7. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, The step of applying an inertial counterforce in the opposite direction to the top of the flexible tower by generating a relative displacement through an internal mass block according to the hardware actuation parameters includes: The dynamic inertial mass of the internal mass block is set according to the equivalent modal mass of the wind turbine tower, and the tuning stiffness and damping coefficient in the hardware actuation parameters are matched in combination with the first natural frequency of the flexible tower. When external excitation induces vibration, the internal mass block generates a relative displacement opposite to the direction of movement of the tower top based on the tuned stiffness and damping. The inertial force of the internal mass block and the frictional force of the damper element are combined to form the inertial counterforce.

8. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, The steps for constructing a comprehensive objective function that includes the standard deviation of the rigid body motion response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower include: The standard deviation of the rigid body motion response is calculated using the time-domain sampling sequence within the specified evaluation window period. The equivalent fatigue load is calculated using the time-domain bending moment sequence collected by the strain sensor installed at the bottom section of the tower, through the rainflow counting method and linear damage theory. By introducing a dimensional normalization factor and a physical weighting coefficient, the extracted standard deviation of the rigid body motion response and the equivalent fatigue load are respectively compared with the benchmark value when vibration reduction control is not enabled. A dual-objective collaborative optimization mechanism is established, and a corresponding comprehensive objective function is constructed.

9. The multi-body dynamic response vibration reduction and control method for floating wind power systems according to claim 1, characterized in that, The steps of dynamically correcting the control parameters and physical parameters in the hardware actuation parameters for the second-level cascaded filtering by running an adaptive boundary optimization algorithm include: Run the built-in multi-objective particle swarm optimization algorithm or grid step optimization algorithm, and use the minimum value of the comprehensive objective function as the convergence criterion. Within the preset parameter space, the center angular frequency and notch bandwidth of the second-order discrete notch filter, as well as the tuning stiffness and damping coefficient in the physical parameters, are updated by linear step scan.

10. A multi-body dynamic response vibration reduction control system for a floating wind power system, characterized in that, The method for multibody dynamic response vibration reduction and control of floating wind power systems as described in any one of claims 1 to 9 includes: The sensing module is used to acquire the instantaneous pitch angular velocity and pitch angular displacement of the floating foundation, as well as the dynamic displacement velocity and displacement of the top of the flexible tower. It converts the acquired data into dynamic displacement, velocity and acceleration vectors of the multi-body flexible structure and establishes global time-domain control equations that include various loads and control forces. The calculation module is used to receive the pitch error input sequence or platform motion feedback signal, and perform two-level cascaded filtering in combination with the global time domain control equation to calculate and output the positive damping pitch control command and hardware actuation parameters. The execution module is used to adjust the pitch angle of the wind turbine blades using the pitch drive mechanism according to the positive damping pitch control command, and at the same time generate relative displacement through the internal mass block according to the hardware actuation parameters, and apply an inertial counterforce in the opposite direction to the top of the flexible tower. The optimization module is used to periodically collect multibody dynamic response data after the adjustment of the wind turbine blade pitch angle and the application of the inertial counterforce action. It constructs a comprehensive objective function that includes the standard deviation of the rigid body motion response of the floating foundation and the equivalent fatigue load of the key bearing section of the flexible tower. It runs an adaptive boundary optimization algorithm to dynamically correct the control parameters and physical parameters in the hardware actuation parameters used for the second-level cascaded filtering. The corrected parameters are then re-input into the calculation module and the execution module for closed-loop iteration.

Images