A method for monitoring the state of a hybrid tower fan

Abstract

The present application relates to wind power generation monitoring technical field, disclose a kind of monitoring method of mixed tower wind turbine state, comprising: obtaining the monitoring data including variable pitch torque current variation rate, actual pitch angle and tower top orthogonal acceleration sequence;Based on zero-lift critical limit, establish feedforward decoupling rule, lock time window when blade synchronous feathering;When actual pitch angle touches critical limit and torque current variation rate meets set threshold, determine eigenvalue breakpoint, divide window and peel off aerodynamic damping modulation data;Acceleration sequence in free decay section is converted into dynamic displacement response sequence, deducts the geometric pseudo-displacement component caused by nacelle rack deformation to identify orthogonal natural frequency, the present application locks time domain segmentation point and aerodynamic turning point, separates and eliminates residual aerodynamic damping modulation interference, improves structural damage feature extraction precision, guarantees wind turbine stable operation.

Description

Technical Field

[0001] This invention relates to the field of wind power generation monitoring technology, and more specifically, to a method for monitoring the status of hybrid tower wind turbines. Background Technology

[0002] In the current operation and maintenance of wind power equipment, the hybrid tower, which is composed of concrete and steel structure, is the core component that bears the load. Its structural safety when bearing alternating loads is directly related to the stability of the whole machine. The conventional strategy is to capture the physical vibration sequence of the tower structure to analyze and characterize the first-order natural frequency that represents the stiffness decay of the tower.

[0003] Not only are there limitations in hardware structure adjustments, but existing technologies also have shortcomings in software control methods. For example, Chinese invention patent application CN116146437A discloses a method for measuring the natural frequency of a wind turbine tower and related components, which intercepts attenuated vibration signals during a specific shutdown period when the generator speed is lower than the grid-connected rated speed. The core premise of this scheme is fundamentally mismatched with the actual operating conditions of grid-connected power generation: this technology implicitly relies on an idealized free attenuation boundary where the unit is completely disconnected from the grid and the flow field tends to be static; however, under the actual dynamic operating conditions of grid-connected power generation or rapid feathering, the time-varying load interference introduced by unsteady wake distortion, and the transient dynamic flexible load of the nacelle and frame caused by rapid pitch changes, will be intertwined with the natural vibration of the tower in multiple dimensions. Control logic that relies solely on shutdown commands and speed determination is insufficient. The inability to accurately identify the physical boundary of transient load unloading makes it easy to misjudge residual aerodynamic damping modulation interference as intrinsic structural dynamic characteristics, leading to frequency analysis distortion. Under grid-connected power generation operation, the unsteady wake distortion generated by the large-scale pitch motion of the blades introduces time-varying load interference components into the tower top structure. This interference component interweaves with the inherent vibration response of the tower structure in multiple dimensions, causing the true physical natural vibration characteristics to be completely hidden. If vibration signals are directly collected, the traditional method of dividing the window based on the geometric characteristics of the vibration waveform cannot accurately identify the physical boundary of transient load unloading, resulting in the calculated first-order natural frequency being accompanied by non-destructive, unrealistic drift. The conventional approach is to forcibly isolate the load by adding a multi-channel external monitoring sensor array, which increases the operation and maintenance costs and difficulty in daily operation.

[0004] Therefore, the technical problem to be solved by this invention is how to eliminate the conventional path that relies on external sensors or conventional waveform geometry slices, combine pitch history with aerodynamic physical variables, and utilize the existing main control unit and airborne sensors of the system to achieve coordinated decoupling of unsteady load, transient deformation and foundation constraints on vibration response under continuous power generation, and accurately extract the true natural frequency that reflects the tower stiffness degradation. Summary of the Invention

[0005] This invention provides a method for monitoring the status of a hybrid tower fan, comprising the following steps: Step S101: Obtain the operation monitoring data of the wind turbine control system. The operation monitoring data includes the pitch torque current variation rate, the actual pitch angle sequence, and the tower top orthogonal bidirectional acceleration sequence. Step S102: Based on the zero-lift pitch angle critical limit, establish aerodynamic load feedforward decoupling rules. When the timing start point of the synchronous feathering action of all blades is captured, lock the dynamic transition time window with a time increment of 15s to 45s. When the actual pitch angle sequence touches the zero-lift pitch angle critical limit and the absolute value of the pitch torque current variation rate is greater than the set drag threshold, establish the intrinsic breakpoint as the timing node after shifting the discrete time point that touches the zero-lift pitch angle critical limit backward by the time cascade lag increment. The time cascade lag increment is the product of the torque-time conversion constant and the pitch torque current variation rate. Using the intrinsic breakpoint as the timing boundary, divide the dynamic transition time window into the first aerodynamic damping modulation segment and the second pure structural intrinsic free decay segment, and separate and remove the acceleration data in the first aerodynamic damping modulation segment. Step S103: The orthogonal bidirectional acceleration sequence at the top of the tower in the pure structural intrinsic free decay section is converted into a dynamic displacement response sequence using the discrete integral operator. At the transient node of the total pitch torque, the geometric pseudo-displacement component caused by the deformation of the nacelle and frame is subtracted to repair the initial amplitude maximum reference and generate the intrinsic free decay response sequence of the hybrid tower. The intrinsic free decay response sequence of the hybrid tower is analyzed to identify the longitudinal true natural frequency and the transverse true natural frequency of the hybrid tower structure.

[0006] Preferably, step S103 specifically includes the following sub-steps: Step S1031, synchronously analyze the intrinsic free decay response sequence of the mixed tower, independently calculate the longitudinal true natural frequency and the transverse true natural frequency, and extract the numerical difference between the longitudinal true natural frequency and the transverse true natural frequency; Step S1032, when the numerical difference is greater than the set boundary variation threshold, identify the external geometric constraint boundary variation caused by uneven settlement of the foundation, and introduce a weight correction operator when calculating the relative frequency drift rate to remove the frequency evolution component caused by the geometric constraint change.

[0007] Preferably, step S102 specifically includes the following sub-steps: step S1021, importing the aerodynamic design data of the wind turbine blades into the wind turbine main control unit; step S1022, extracting the zero-lift pitch angle critical value from the aerodynamic design data, converting the zero-lift pitch angle critical value into a feedforward separation threshold, so as to establish an aerodynamic load feedforward decoupling rule.

[0008] Preferably, the repair of the initial amplitude maximum reference in step S103 specifically includes the following sub-steps: Step S1033, by obtaining the pitch torque current variation rate to characterize the transient dynamic flexible load of the nacelle frame caused by rapid feathering; Step S1034, by combining the actual pitch angle sequence and the tower top orthogonal bidirectional acceleration sequence in the pure structural intrinsic free decay section, the dynamic displacement response sequence is calculated by the discrete integral operator; Step S1035, at the transient node of the total pitch torque, the geometric pseudo-displacement component generated by the transient dynamic flexible load of the nacelle frame is calculated, and the geometric pseudo-displacement component is subtracted from the dynamic displacement response sequence to complete the in-situ repair of the initial amplitude maximum reference, eliminating the low-frequency tilting interference introduced by the local elastic deformation to the nacelle accelerometer used to collect the tower top orthogonal bidirectional acceleration sequence.

[0009] Preferably, the introduction of a weight correction operator to remove frequency evolution components caused by geometric constraint variations specifically includes the following sub-steps: Step S1036, during the calculation of the relative frequency drift rate, when the orthogonal frequency deviation value does not exceed the set boundary variation threshold, the basic parameter calculation channel is kept running; Step S1037, when the orthogonal frequency deviation value exceeds the set boundary variation threshold, the parameter compensation channel is activated to separate and remove the frequency evolution components caused by external geometric constraint boundary variations from the total frequency drift, and the remaining pure frequency drift rate is mapped to the stiffness degradation state of the hybrid tower body.

[0010] Preferably, mapping the remaining pure frequency drift rate to the stiffness degradation state of the hybrid tower body specifically includes the following sub-steps: Step S1038, continuously record the pure frequency drift rate of the wind turbine during the operating cycle, and establish a time series benchmark that evolves with operating time and load accumulation; Step S1039, extract the changing trend characteristics of the time series benchmark to calculate a quantitative index characterizing the stiffness reduction rate of the hybrid tower connection part; Step S10310, when the quantitative index exceeds the set safety threshold, output the corresponding state prediction and early warning signal of microcrack propagation and stiffness degradation inside the concrete.

[0011] Preferably, in acquiring the operation monitoring data of the wind turbine control system, the sampling frequency of the nacelle accelerometer for acquiring the orthogonal bidirectional acceleration sequence at the top of the tower is set to not less than 100Hz, and the single interception time length of the operation monitoring data is set to not less than 600s, so as to ensure that the segmented pure structural intrinsic free decay section has a complete free decay history.

[0012] Preferably, step S103 specifically includes the following sub-steps: Step S10311, compare the identified longitudinal true natural frequency and transverse true natural frequency with the design natural frequency in the initial state of the hybrid tower, and calculate the total frequency drift; Step S10312, search the set hybrid tower damage mapping relationship library, match the fatigue damage degree of the concrete connection part corresponding to the total frequency drift, and determine the structural health level of the hybrid tower structure.

[0013] Preferably, after determining the structural health level of the hybrid tower structure, the following steps are included: Step S10313, when the structural health level reaches the danger control level, the main control unit of the wind turbine sends a vibration damping adjustment command to the pitch drive unit of the wind turbine. By changing the pitch rate and speed control trajectory within the operating wind speed range, the wind turbine avoids the resonance frequency band of the hybrid tower structure, thus ensuring the safe operation of the hybrid tower wind turbine.

[0014] The present invention has the following beneficial effects: 1. In the monitoring of the condition of hybrid tower wind turbines, by transforming the critical limit of zero-lift pitch angle in the aerodynamic physical design of the wind turbine blades into a data feedforward interception rule, the time domain signal segmentation point is continuously aligned with the aerodynamic characteristic inflection point in the pitch process. The main control unit feedforward cuts off and removes the flow field distortion and residual aerodynamic damping modulation interference caused by large-scale feathering, ensuring that the frequency analysis waveform is transformed into a free decay response dominated by the stiffness and mass distribution of the hybrid tower itself. This solves the constraint of residual aerodynamic loads masking intrinsic dynamic parameters when relying on waveform geometric gradient segmentation, improves the accuracy of structural damage feature extraction under continuous power generation conditions, and thus ensures the long-term safe and stable operation of the high tower structure.

[0015] 2. To address the elastic deformation interference of the nacelle and frame caused by rapid feathering under strong wind conditions, the main control unit synchronously acquires the pitch torque current variation rate of the pitch drive unit and, combined with the actual pitch angle sequence, converts the tower top acceleration into a dynamic displacement response using a discrete integral operator. At the transient node of the total pitch torque, the geometric pseudo-displacement component caused by the frame bounce is subtracted, and the initial amplitude maximum reference in the initial stage of the free decay section is repaired in situ. This avoids the technical problem of local elastic deformation introducing low-frequency nonlinear tilt signals to the airborne accelerometer, ensures the physical accuracy of the logarithmic decay rate operator in solving the intrinsic damping ratio of the mixed tower, and avoids state judgment deviations caused by transient geometric nonlinear interference.

[0016] 3. Focusing on the two-dimensional spatiotemporal topological correlation of forward and backward acceleration and lateral acceleration, the main control unit synchronously analyzes the orthogonal bidirectional acceleration decoupling sequence, calculates the intrinsic frequencies of the orthogonal bidirectional acceleration respectively, and calculates the asymmetric anisotropic characteristic index based on the numerical difference. When the index exceeds the preset threshold, it identifies the alienation of the external geometric constraint boundary caused by uneven settlement of the foundation. In the calculation of relative frequency drift rate, the parameter compensation channel is activated and the weight correction operator is introduced to separate and eliminate the frequency evolution component caused by the change of geometric constraints, avoid the cross interference of environmental constraint alienation on the damage monitoring of the mixed tower body, and eliminate the defects of traditional unidirectional monitoring methods that cause false alarms under uneven working conditions. Attached Figure Description

[0017] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings, in which several embodiments of the invention are illustrated by way of example and not limitation, wherein: Figure 1 This is a flowchart of the wind turbine condition monitoring data segmentation and frequency identification process of the present invention; Figure 2 This is a topology diagram of the aerodynamic load decoupling and pseudo-displacement subtraction of the present invention. Detailed Implementation

[0018] The principles and spirit of the present invention will now be described with reference to several exemplary embodiments in conjunction with the accompanying drawings. It should be understood that these embodiments are provided merely to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to fully convey the scope of the present invention to those skilled in the art.

[0019] A method for monitoring the status of a hybrid tower fan includes the following steps: Step S101: Obtain the operation monitoring data of the wind turbine control system. The operation monitoring data includes the pitch torque current variation rate, the actual pitch angle sequence, and the tower top orthogonal bidirectional acceleration sequence. Step S102: Based on the zero-lift pitch angle critical limit, establish aerodynamic load feedforward decoupling rules. When the timing start point of the synchronous feathering action of all blades is captured, lock the dynamic transition time window with a time increment of 15s to 45s. When the actual pitch angle sequence touches the zero-lift pitch angle critical limit and the absolute value of the pitch torque current variation rate is greater than the set drag threshold, establish the intrinsic breakpoint as the timing node after shifting the discrete time point that touches the zero-lift pitch angle critical limit backward by the time cascade lag increment. The time cascade lag increment is the product of the torque-time conversion constant and the pitch torque current variation rate. Using the intrinsic breakpoint as the timing boundary, divide the dynamic transition time window into the first aerodynamic damping modulation segment and the second pure structural intrinsic free decay segment, and separate and remove the acceleration data in the first aerodynamic damping modulation segment. Step S103: The orthogonal bidirectional acceleration sequence at the top of the tower in the pure structural intrinsic free decay section is converted into a dynamic displacement response sequence using the discrete integral operator. At the transient node of the total pitch torque, the geometric pseudo-displacement component caused by the deformation of the nacelle and frame is subtracted to repair the initial amplitude maximum reference and generate the intrinsic free decay response sequence of the hybrid tower. The intrinsic free decay response sequence of the hybrid tower is analyzed to identify the longitudinal true natural frequency and the transverse true natural frequency of the hybrid tower structure.

[0020] Preferably, step S103 specifically includes the following sub-steps: Step S1031, synchronously analyze the intrinsic free decay response sequence of the mixed tower, independently calculate the longitudinal true natural frequency and the transverse true natural frequency, and extract the numerical difference between the longitudinal true natural frequency and the transverse true natural frequency; Step S1032, when the numerical difference is greater than the set boundary variation threshold, identify the external geometric constraint boundary variation caused by uneven settlement of the foundation, and introduce a weight correction operator when calculating the relative frequency drift rate to remove the frequency evolution component caused by the geometric constraint change.

[0021] Preferably, step S102 specifically includes the following sub-steps: step S1021, importing the aerodynamic design data of the wind turbine blades into the wind turbine main control unit; step S1022, extracting the zero-lift pitch angle critical value from the aerodynamic design data, converting the zero-lift pitch angle critical value into a feedforward separation threshold, so as to establish an aerodynamic load feedforward decoupling rule.

[0022] Preferably, the repair of the initial amplitude maximum reference in step S103 specifically includes the following sub-steps: Step S1033, by obtaining the pitch torque current variation rate to characterize the transient dynamic flexible load of the nacelle frame caused by rapid feathering; Step S1034, by combining the actual pitch angle sequence and the tower top orthogonal bidirectional acceleration sequence in the pure structural intrinsic free decay section, the dynamic displacement response sequence is calculated by the discrete integral operator; Step S1035, at the transient node of the total pitch torque, the geometric pseudo-displacement component generated by the transient dynamic flexible load of the nacelle frame is calculated, and the geometric pseudo-displacement component is subtracted from the dynamic displacement response sequence to complete the in-situ repair of the initial amplitude maximum reference, eliminating the low-frequency tilting interference introduced by the local elastic deformation to the nacelle accelerometer used to collect the tower top orthogonal bidirectional acceleration sequence.

[0023] Preferably, the introduction of a weight correction operator to remove frequency evolution components caused by geometric constraint variations specifically includes the following sub-steps: Step S1036, during the calculation of the relative frequency drift rate, when the orthogonal frequency deviation value does not exceed the set boundary variation threshold, the basic parameter calculation channel is kept running; Step S1037, when the orthogonal frequency deviation value exceeds the set boundary variation threshold, the parameter compensation channel is activated to separate and remove the frequency evolution components caused by external geometric constraint boundary variations from the total frequency drift, and the remaining pure frequency drift rate is mapped to the stiffness degradation state of the hybrid tower body.

[0024] Preferably, mapping the remaining pure frequency drift rate to the stiffness degradation state of the hybrid tower body specifically includes the following sub-steps: Step S1038, continuously record the pure frequency drift rate of the wind turbine during the operating cycle, and establish a time series benchmark that evolves with operating time and load accumulation; Step S1039, extract the changing trend characteristics of the time series benchmark to calculate a quantitative index characterizing the stiffness reduction rate of the hybrid tower connection part; Step S10310, when the quantitative index exceeds the set safety threshold, output the corresponding state prediction and early warning signal of microcrack propagation and stiffness degradation inside the concrete.

[0025] Preferably, in acquiring the operation monitoring data of the wind turbine control system, the sampling frequency of the nacelle accelerometer for acquiring the orthogonal bidirectional acceleration sequence at the top of the tower is set to not less than 100Hz, and the single interception time length of the operation monitoring data is set to not less than 600s, so as to ensure that the segmented pure structural intrinsic free decay section has a complete free decay history.

[0026] Preferably, step S103 specifically includes the following sub-steps: Step S10311, compare the identified longitudinal true natural frequency and transverse true natural frequency with the design natural frequency in the initial state of the hybrid tower, and calculate the total frequency drift; Step S10312, search the set hybrid tower damage mapping relationship library, match the fatigue damage degree of the concrete connection part corresponding to the total frequency drift, and determine the structural health level of the hybrid tower structure.

[0027] Preferably, after determining the structural health level of the hybrid tower structure, the following steps are included: Step S10313, when the structural health level reaches the danger control level, the main control unit of the wind turbine sends a vibration damping adjustment command to the pitch drive unit of the wind turbine. By changing the pitch rate and speed control trajectory within the operating wind speed range, the wind turbine avoids the resonance frequency band of the hybrid tower structure, thus ensuring the safe operation of the hybrid tower wind turbine.

[0028] Example 1: The current method for monitoring the condition of a hybrid tower wind turbine operates in the environment of a hybrid tower wind turbine generator set connected to the grid for power generation. The hybrid tower structure is subjected to alternating aerodynamic loads, blade rotational excitation forces, and environmental disturbances. Its connection parts are at risk of stiffness degradation, prestressing tension loss, and concrete microcrack propagation. Due to the interweaving of aerodynamic noise during power generation, 1P and 3P frequency disturbances accompanying the blade passing frequency, and the low-order structural natural frequencies of the hybrid tower itself, the characteristic signals representing structural damage are often masked by high-energy random excitation loads, making it difficult to directly extract the dynamic parameters of the hybrid tower under continuous power generation conditions. Offline impact testing or cable force detection causes the unit to shut down and results in power generation loss, making it impossible to dynamically capture sudden damage tendencies during operation.

[0029] When the wind turbine control system issues a pitch command and the main control unit detects the timing start point t of the synchronous rapid feathering action of all blades, start At that time, the main control unit locks the transition time window T, which lasts for a time increment of 15s to 45s. win Its starting time is t start The termination time is , where t end t is the window termination time. start As the starting point of the time series, The time window span is the transition time window T. win Inside, the main control unit collects the vibration acceleration response sequence of the tower top through accelerometers arranged on the nacelle base. The sequence includes the forward and backward acceleration sequence of the tower top obtained at a sampling frequency of 50Hz. With lateral acceleration sequence Simultaneously, the actual pitch angle sequence of the three blades of the wind turbine is introduced. The main control unit reads the pre-stored blade zero-lift physical characteristic pitch angle reference value S.zero and the actual pitch angle sequence Compared with the benchmark value S zero Expand continuous numerical comparison; when the actual pitch angle sequence Increase until the reference value S is reached zero At discrete time points, the main control unit introduces a data feedforward interception operator to forcibly establish that discrete time point as the intrinsic breakpoint t for aerodynamic decoupling. cut Therefore, the main control unit uses the intrinsic breakpoint t cut As a rigid timing boundary, the transition time window T is... win The system is divided into a front aerodynamic damping modulation section and a rear pure structural intrinsic free decay section. Acceleration data from the front aerodynamic damping modulation section is directly stripped and discarded, while the acceleration response sequence from the rear pure structural intrinsic free decay section is transmitted to subsequent demodulation. Based on the non-equilibrium aerodynamic state of the blade surface during extreme complex shear winds in non-ideal wind farm operation, which causes a time lag at the actual zero-lift point, the main control unit synchronously acquires the pitch torque current variation rate dl / dt of the pitch drive unit. When the actual pitch angle sequence... Touching the reference value S zero When the absolute value of the pitch torque current variation rate dl / dt is greater than the preset drag threshold, the main control unit completes the time cascade lag fine adjustment and sets the intrinsic breakpoint t. cut Shift backward by a fixed step increment t add To extend the scope of front-end data stripping, a fixed step increment t is used. add The specific calculation formula is as follows: , where t add The time increment required for translation, The torque-time conversion constant is preset in the controller, and dl / dt is the currently acquired pitch torque-current variation rate.

[0030] After extracting the acceleration response sequence within the pure structural intrinsic free decay range, in order to eliminate the low-frequency nonlinear tilting interference introduced to the nacelle accelerometer by the transient change in flexible load on the nacelle frame caused by rapid feathering under extreme wind conditions, the main control unit uses a discrete integral operator to convert the tower top forward and backward acceleration sequence into a single unit. With lateral acceleration sequence The result is converted into a displacement response sequence, and the geometric pseudo-displacement component caused by frame bounce is subtracted at the transient node of the total pitch torque. Based on the initial amplitude maximum value benchmark of the initial stage of the free decay section, the intrinsic free decay response sequence of the hybrid tower is calculated. The calculation process of the geometric pseudo-displacement component is as follows: the main control unit performs one-step integration in the time domain using the obtained pitch torque current variation rate to obtain a characteristic quantity reflecting the transient impact intensity of the pitch torque. This characteristic quantity is multiplied by the dynamic flexibility calibration coefficient of the nacelle frame to obtain the initial pseudo-displacement offset generated by the elastic jump. This dynamic flexibility calibration coefficient is pre-measured through static cantilever load test during the unit assembly stage, and its value is fixed at 0.00096 m / Ampere. The calculated offset is used as the geometric pseudo-displacement component. At the time node when the total pitch torque changes transiently, it is directly subtracted from the original displacement response sequence output by the discrete integral operator to restore the balance axis of the filtered waveform to the zero mean level. The main control unit simultaneously analyzes the intrinsic free decay response sequence of the hybrid tower and independently calculates the longitudinal natural frequency f using the zero crossover point counting operator. cur_x With transverse natural frequency f cur_y Longitudinal natural frequency f cur_x The specific calculation formula is as follows: , where f cur_x Where N is the longitudinal natural frequency. zero t represents the absolute number of times the vibrational acceleration response sequence crosses the zero level within the intrinsically free decay segment of a pure structure. end t is the window termination time. cut For the calibrated intrinsic breakpoint time, the main control unit simultaneously uses the logarithmic decay rate operator to solve for the intrinsic damping ratio of the hybrid tower. , The specific calculation formula is as follows: ,in, The intrinsic damping ratio, Let π be the constant value of a circle, and A1 be the first amplitude maximum value within the repaired free decay segment. M This represents the Mth amplitude maximum, where M is a fixed amplitude level selection value, limited to integers between 5 and 10. Based on the heterogeneous consolidation of the foundation soil and anisotropic settlement of the foundation caused by long-term unidirectional wind-induced vibration of the hybrid tower structure, this settlement constraint distortion leads to asymmetric implicit drift of the orthogonal bidirectional natural frequencies. To avoid confusion between the external constraint variation characteristics and the tower body stiffness degradation characteristics, the main control unit introduces an asymmetric anisotropic characteristic index K. asym The specific calculation formula is as follows: , where K asym f is an asymmetric anisotropy characteristic index. cur_x f is the longitudinal natural frequency. cur_y Here, f0 is the transverse natural frequency, and f0 is the reference value of the first-order natural frequency of the wind turbine calibrated under no-damage standard during the initial operation. When index K asymWhen the deviation exceeds the preset boundary variation threshold of 0.008, the system identifies the external geometric constraint boundary variation caused by uneven foundation settlement. It automatically initiates parameter compensation routing and introduces a weight correction operator to forcibly remove the frequency evolution component caused by foundation constraint changes, thereby ensuring the specificity of damage monitoring for the hybrid tower body. Specifically, the execution steps of the weight correction operator and component stripping are as follows: the main control unit calculates the absolute difference between the asymmetric anisotropy characteristic index and the preset boundary variation threshold, multiplies this difference by a pre-calibrated base settlement stiffness sensitivity coefficient, and obtains the result of the asymmetric foundation settlement variation. The drift components of the transverse and longitudinal natural frequencies caused by the variation are determined by a sensitivity coefficient calculated from historical settlement observation data and fixed at 0.045 Hz. By directly subtracting the calculated drift components from the arithmetic mean of the currently measured longitudinal and transverse natural frequencies, the non-structural damage frequency evolution components caused by the distortion of external constraints can be forcibly eliminated. The corrected pure value is then output as the current first-order intrinsic natural frequency to the subsequent relative frequency drift rate calculation channel. Within the parameter compensation channel, the main control unit outputs the first-order intrinsic natural frequency f after removing the environmental constraint distortion components. cur Compare with the baseline value f0, and according to the calculation formula Calculate the relative frequency drift rate ,in, f is the relative frequency drift rate, f0 is the reference value of the first-order natural frequency of the damage-free standard, and f cur The current first-order intrinsic natural frequency is determined after foundation settlement compensation calibration. During this process, the main control unit supports state assessment by calling a hybrid tower damage mapping database pre-stored in non-volatile memory. This database is constructed using a bi-column associative topology data structure. Its input column is a one-dimensional matrix of relative frequency drift rate discretized with a step size of 0.001, and its output column is a pre-calibrated structural fatigue damage scale value and safety warning level code. The database's data matrix is ​​calibrated and initialized using finite element fatigue simulation tests on a full-size hybrid tower structure, combined with 20 sets of historical fracture sample data. This establishes a deterministic mapping relationship between the calculated pure relative frequency drift rate and the microcrack propagation depth at the internal connection points of the concrete. The state decision unit makes conditional judgments based on the obtained indicators. When the relative frequency drift rate... When the relative frequency drift rate is less than or equal to 0.02, the hybrid tower structure is considered to be in a healthy state; when the relative frequency drift rate is less than or equal to 0.02, the hybrid tower structure is considered to be in a healthy state. When the relative frequency drift rate is greater than 0.02 and less than or equal to 0.05, the hybrid tower structure is judged to have suffered minor damage, with localized microcracks evolving in the concrete at the corresponding connection points. The system automatically sends a secondary early warning signal and a minor maintenance suggestion to the management terminal to establish operational decision support; when the relative frequency drift rate... When the value is greater than 0.05, structural stiffness degradation is determined to have occurred. The main control unit triggers a safety protection mechanism, issuing vibration damping adjustment commands and a safety shutdown protection program to the wind turbine's pitch drive unit. By limiting the pitch rate and speed control trajectory of the pitch drive unit, the wind turbine avoids the resonant frequency band of the hybrid tower structure, thus resolving the masking effect of continuous aerodynamic excitation on the intrinsic dynamic parameters of the tower. This is used to automatically monitor and protect against stiffness degradation and internal crack evolution during continuous power generation. Specifically, the closed-loop adjustment process of the vibration damping adjustment command and speed control trajectory is as follows: the main control unit dynamically locks a frequency based on the currently identified degraded first-order intrinsic natural frequency. The resonant physical frequency dead zone is centered at a frequency of 0.05 Hz. The speed control trajectory optimization algorithm automatically adjusts the generator speed command to limit the wind turbine rotation excitation frequency outside the aforementioned resonant dead zone, while limiting the pitch rate to within 1.5 degrees per second, thereby forcibly reducing the transient impact amplitude generated during large-scale aerodynamic load unloading. Through this active limitation of the control loop, the energy input of external aerodynamic excitation at the tower's natural frequency point is reduced. At the physical level, it provides an active electromagnetic and aerodynamic combined damping feedback for the hybrid tower structure, enabling the system to effectively suppress the resonant response even under damage conditions, ensuring the inherent mechanical safety of the entire machine during shutdown transition periods.

[0031] Example 2: A current method for monitoring the status of hybrid-tower wind turbines operates in the context of grid-connected hybrid-tower wind turbine generators. A semi-physical simulation verification platform for a 6 MW onshore wind turbine generator provides the test environment. The test signal source input is loaded with a multi-dimensional composite vibration signal stream containing wind shear inhomogeneity disturbances, blade passing frequency excitation, and Gaussian white noise with a signal-to-noise ratio of 20 dB. The basic data source of the test environment is a 140-meter-high concrete and steel composite hybrid tower structure constructed from a free, high-damping, full-size nonlinear finite element structural model. This is coupled with the aerodynamic load vibration response time sequence obtained by solving the Navier-Stokes equations under the unsteady flow field conditions of a conventional wind farm operating under grid-connected conditions for large-capacity, high-tower hybrid-tower wind turbine generators. The sensor used for orthogonal acceleration acquisition at the tower top is calibrated with a measurement range of ±5 m / s². 2 Its measurement resolution is better than 0.001 m / s 2 The sampling frequency is fixed at 50Hz. This sampling frequency was determined by considering two physical factors: the spectral bandwidth of the vibration signal and the computational load of the main control unit. A balance was struck between meeting the technical objective of avoiding signal aliasing and maintaining a reasonable data throughput load for the controller. The transition time window T of the main control unit... win Dynamic span in Set to 30 seconds, it falls within the range of 15 to 45 seconds; this dynamic span... The value of the value is determined by considering the duration of the free decay of the intrinsic damping of the structure and the timing constraints of the subsequent pitch reset action. This ensures sufficient attenuation of the tower vibration while avoiding subsequent aerodynamic re-excitation waveforms. To verify the filtering capability of the monitoring method against multidimensional interference and its sensitivity to structural damage, a multidimensional control system consisting of the sample group of this invention, a control group, a partially missing control group, and an out-of-range control group was set up in the experiment. Simultaneously, three different damage intensity gradients were set for the concrete and steel composite hybrid tower structure. The first level is the healthy state with no stiffness loss, and its standard first-order natural frequency reference value f0 is calibrated to 0.362Hz. The second level is the mild damage state with the evolution of local microcracks in the concrete at the connection point, and its actual structural frequency becomes 0.351Hz. The third level is the damage state with structural stiffness degradation, and its actual structural frequency becomes 0.334Hz. In the first control sequence after the start of the experiment, the main control unit retrieves the data of the healthy state with no loss in the first level. When the timing start t of the blade starting synchronous rapid feathering action is captured, the main control unit retrieves the data of the healthy state with no loss in the first level. start When the transition time window T is 10.00s, win The end time The value is set at 40.00s, where t end t is the window termination time. start As the starting point of the time series, The time window span is 30 seconds, during which the actual pitch angle sequence... The pitch rate increases continuously from the initial power generation operating angle towards the feathering direction at a rate of 4.0° / s, reaching the preset zero-lift physical characteristic pitch angle reference value S at the timing node of 26.25s. zero At 85.00°, the non-equilibrium flow field caused by wind shear generates additional rotational drag on the pitch drive unit. The collected pitch torque current variation rate dl / dt reaches 12.50 A / s, which is greater than the preset drag threshold of 10.00 A / s. The sample group of this invention automatically intervenes through the data feedforward interception operator and substitutes it into the formula. The fixed step increment t is calculated. add The value is 1.00s, where t add The time increment required for translation, The torque-time conversion constant preset in the controller is set to 0.08 s / A / s, and dl / dt is the pitch torque current variation rate, with a measured value of 12.50 A / s. Therefore, the initial intrinsic breakpoint of the sample group of this invention is shifted from 26.25 s to 27.25 s to establish the calibrated intrinsic breakpoint t. cutThis approach completely cuts off and discards the acceleration data of the aerodynamic damping modulation segment before 27.25s at the time domain boundary, and transmits the data of the pure structural intrinsic free decay segment between 27.25s and 40.00s to the subsequent demodulation. In contrast, the control group uses the traditional waveform geometric gradient segmentation method, which cannot intercept the unsteady wake distortion flow field disturbance during the pitch process. Its time domain segmentation point lags to 32.10s, and the intercepted decay response waveform contains residual aerodynamic damping modulation components that are not fully decoupled. At the same time, some missing control groups have shut down the geometric pseudo-displacement components for the transient changes in the flexible load of the nacelle frame. The corrected calculation shows that the initial amplitude maximum reference of the displacement response sequence produces an upward bias jump of 0.012m at the transient node of the total variable pitch frame, which causes the equilibrium axis of the free decay waveform to bend nonlinearly.

[0032] The present invention analyzes the retained pure structural intrinsic free decay segment acceleration response sequence using a zero-crossing point counting operator and a logarithmic decay rate operator. Under the first level of lossless and healthy conditions, the absolute number N of the vibration acceleration response sequence crossing the zero level within the pure structural intrinsic free decay segment is measured. zero It is 9 times, substituting into the formula The longitudinal natural frequency f is calculated. cur_x The transverse natural frequency f is 0.3615Hz. cur_y It is 0.3611Hz, where f cur_x Where N is the longitudinal natural frequency. zero t represents the absolute number of times the vibrational acceleration response sequence crosses the zero level within the intrinsically free decay segment of a pure structure. end The window ends at time t, which is 40.00 seconds. cut The calibrated intrinsic breakpoint time is 27.25 s, and then the formula is used... Calculate the asymmetric anisotropy characteristic index K asym The value is 0.0011, where K asym f is an asymmetric anisotropy characteristic index. cur_x The longitudinal natural frequency is 0.3615 Hz, f cur_y The transverse natural frequency is 0.3611 Hz, and f0 is the reference value for the first-order natural frequency under no-damage standard, which is 0.362 Hz. Since this index is lower than the set boundary variation threshold of 0.008, the system determines that the current bidirectional frequency is symmetrical and there is no abnormal foundation anisotropic settlement. The output first-order intrinsic natural frequency f is... cur The frequency was set at 0.3613 Hz, according to the formula. Calculate the relative frequency drift rate It is 0.0019, where, f0 is the relative frequency drift rate, and f0 is the reference value of the first-order natural frequency of the damage-free standard, which is 0.362 Hz. cur This is the current first-order intrinsic natural frequency, with a value of 0.3613 Hz.

[0033] When switching to the second-level mild damage state, the continuous superposition of Gaussian white noise and blade frequency perturbation caused the relative frequency drift rate data obtained from the traditional control group to oscillate violently between 0.011 and 0.028, failing to provide stable demodulation results. However, the sample group of this invention exhibited stable suppression of environmental background noise interference, and its calculated first-order eigenfrequency f... cur The frequency converged stably to 0.3508 Hz, corresponding to the relative frequency drift rate. The calculated result is 0.0309, which falls within the range of 0.02 to 0.05 for minor damage assessment. This accurately drives the state decision unit to output a secondary early warning signal and minor maintenance recommendations to the downstream management terminal, thus establishing operational decision support. Under the third-level severe damage condition, the first-order intrinsic natural frequency f measured by the sample group of this invention... cur The frequency shift becomes 0.3341Hz, and the calculated relative frequency shift rate is... The value becomes 0.0771, exceeding the structural damage critical trigger line of 0.05, triggering the wind turbine main control protection circuit and sending a safety shutdown protection program to the pitch drive unit to limit the pitch rate. To further verify the performance inflection point and nonlinear working effect within the transition time window range defined by the method of this invention, the out-of-range control group was tested by artificially changing the transition time window T. win A comparative test was conducted on the span, and the test data showed that when the dynamic span was increased... When shortened to 10s, the insufficient time-domain analysis length and the finite number of points limited by the zero-crossing point counting operator under discrete sampling cause quantization resolution truncation errors, resulting in a decrease in the calculated longitudinal natural frequency f under the second-level mild damage state. cur_x Bias drift occurs, making it impossible to clearly distinguish from a healthy state, exhibiting a performance failure saturation surface at low temporal resolution; while when the dynamic span is increased... When the window extends to 60s, due to the excessive extension of the window tail and its intrusion into the yaw-wind reset sequence after the wind turbine's rapid feathering, the newly occurring unsteady aerodynamic load excites the tower structure again, causing a sudden nonlinear amplitude jump in the acceleration response waveform. This disrupts the physical boundary model of pure structural free decay, leading to a change in the hybrid tower intrinsic damping ratio calculated by the logarithmic decay rate operator. The value changes from the normal 0.015 to 0.068, thus forming a mathematical demodulation degradation inflection point. The fact that the quantization performance degrades outside the boundary of the limited window confirms that the transition time window range of 15s to 45s is the engineering optimal working window that balances frequency resolution accuracy and the physical boundary of pure free decay. By adopting the monitoring method of hybrid tower wind turbine status, without relying on high-cost external dedicated hardware arrays, and only using the onboard accelerometer and variable pitch status monitoring signals, it is possible to achieve accurate identification and intrinsic safety protection of the stiffness degradation state and internal texture evolution of hybrid towers in the high-noise industrial environment of continuous power generation.

[0034] Example 3: This example combines Figures 1 to 2 A method for monitoring the status of a hybrid tower fan is described, such as... Figure 1 As shown, the wind turbine control system and the acquisition of wind turbine control system operation monitoring data have an input-output relationship. After acquiring operation monitoring data including pitch torque current variation rate, actual pitch angle sequence and tower top orthogonal bidirectional acceleration sequence, the step of establishing aerodynamic load feedforward decoupling rules is executed, and the process of locking dynamic transition time window is entered. Within this window, the operation of establishing intrinsic breakpoint and dividing time window is performed to separate and remove the front-end acceleration data. Then, the introduced discrete integral operator is used as input and the remaining sequence is converted into dynamic displacement response sequence. Then, the interference is eliminated by subtracting geometric pseudo displacement components to repair the benchmark, and finally the monitoring target of identifying the true natural frequencies of longitudinal and transverse vibration is realized.

[0035] like Figure 2 As shown, the technical architecture of the monitoring method is supported by four core technical dimensions: operational monitoring data, aerodynamic load decoupling, intrinsic breakpoint establishment, and pseudo-displacement component subtraction. These dimensions work together to ultimately identify the true longitudinal and lateral natural frequencies. The operational monitoring data section includes basic monitoring variables such as the pitch torque current variation rate, the actual pitch angle sequence, and the tower top orthogonal bidirectional acceleration sequence. The aerodynamic load decoupling section includes feedforward control elements such as the zero-lift pitch angle critical limit, the dynamic transition time window, and the front-end aerodynamic damping modulation section. The intrinsic breakpoint establishment section integrates time-series boundary correction parameters such as the set drag threshold, the torque-time conversion constant, and the time cascade lag increment. The pseudo-displacement component subtraction section coordinates benchmark repair operators such as geometric pseudo-displacement components, the transient node of the total pitch torque, and the discrete integral operator. All technical elements cooperate with each other and converge to the core output object, thereby achieving accurate analysis of the true natural frequencies of the hybrid tower structure.

[0036] Example 4: The current method for monitoring the condition of a hybrid tower wind turbine operates in a grid-connected environment for hybrid tower wind turbine generators. The application scenario is based on long-term aging conditions. The hybrid tower structure is subjected to alternating aerodynamic loads, blade rotational excitation forces, and environmental disturbances. Its connection points are susceptible to stiffness degradation, prestressing tension loss, and the propagation of microcracks in the concrete. This is due to the creep relaxation of the prestressed steel cables caused by stress alternation during long-term operation, and the low-frequency zero-point temperature drift of the accelerometer due to alternating ambient temperature, resulting in a forward and backward acceleration sequence at the tower top. With lateral acceleration sequence When the baseline drift of the displacement trend term is superimposed, directly applying the quadratic integral will cause the dynamic parameter calculation results to diverge, and the bias component will mask the degradation characteristics of the first-order natural frequency of the structure and cause operational safety risks.

[0037] To eliminate bias interference caused by hardware aging and sensor zero drift, the main control unit automatically triggers a timeliness guarantee and reconstruction channel for the discrete integral operator and set boundary variation thresholds under specific calibration conditions where the wind speed is below 3 m / s and the fan is in free-running mode. This is used to correct the reference physical parameters in situ. The main control unit reads the static bias residual value output by the accelerometer, calculates the temperature drift signal of the current time segment through the discrete difference mean operator with a sliding time window, and attaches a fourth-order Butterworth high-pass digital filter with a cutoff frequency set to 0.05 Hz to the pre-stage of the discrete integral operator to truncate the bias components below the cutoff frequency in the time domain, thus completing the forward and backward acceleration sequence at the top of the tower. With lateral acceleration sequence Baseline correction.

[0038] Before converting the orthogonal bidirectional acceleration sequence at the tower top into a dynamic displacement response sequence using a discrete integral operator, the main control unit calls a fourth-order Butterworth high-pass digital filter with a cutoff frequency set between 0.02 Hz and 0.05 Hz to process the acceleration sequence, truncating the low-frequency temperature drift signal at the sensor zero point caused by alternating ambient temperature. During the initial baseline calibration before unit commissioning, the yaw mechanism is driven to generate a 90-degree forward and reverse yaw step reversal action. The transient inertial torque generated at the moment of braking is used as the pulse excitation source. The static offset residual value output by the accelerometer at the nacelle base is collected, and the pitch feedback torque of the pitch drive unit is read synchronously. The elastic deformation of the nacelle frame corresponding to each ampere pitch current change is calculated using a multi-point linear regression equation to determine the benchmark value of the dynamic flexibility calibration coefficient, eliminating low-frequency tilting interference caused by local elastic deformation. Specifically, the execution logic of the sliding time window discrete difference averaging operator is as follows: the main control unit sets a discrete sliding time window with a length of 600 seconds and a step size of 1 second, continuously buffering the sampling data points of the accelerometer under free-stop conditions; the operator performs first-order forward difference on adjacent discrete sampling points within the window to remove high-frequency vibration components and extract the local trend slope of the signal; then, it calculates the arithmetic mean of all calculated slope sequences to characterize the low-frequency sensor zero-point drift velocity caused by alternating ambient temperature; finally, by multiplying the average drift velocity by the current cumulative running time, the absolute value of the temperature drift signal in the current time segment can be accurately calculated, thereby providing a reliable baseline calibration input; the main control unit retrieves the historical database under non-damage conditions, and based on the anisotropic distribution characteristics, uses the longitudinal natural frequency f... cur_x With transverse natural frequency f cur_y Within a 48-hour normal power generation window, a predetermined multiple of the standard deviation is used to dynamically reconstruct the boundary variation threshold. The main control unit retrieves historical natural frequency data under the initial undamaged state during commissioning, statistically analyzes the time-series values ​​of the longitudinal and transverse true natural frequencies within the 48-hour normal power generation window, calculates the root mean square error of the difference between the longitudinal and transverse true natural frequencies, and sets three times the root mean square error as the boundary variation threshold. When the asymmetric anisotropy characteristic index exceeds the threshold, the main control unit activates the parameter compensation channel, calculates the absolute difference between the index and the threshold, multiplies the absolute difference by a pre-calibrated base settlement stiffness sensitivity coefficient, obtains the transverse and longitudinal natural frequency drift components caused by foundation asymmetric variation, and subtracts the drift component from the arithmetic mean of the currently measured longitudinal and transverse true natural frequencies to eliminate non-structural damage frequency evolution components caused by external constraint distortion, thus setting the boundary variation threshold T. asym The specific calculation formula is as follows: , among which, T asym To determine the boundary variation threshold obtained from the reconstruction, The root mean square error of the orthogonal bidirectional natural frequency difference under historical no-damage conditions is used as a basis for eliminating the influence of gear wear on the torque-time conversion constant of the pitch drive unit. The generated dynamic nonlinear deformation disturbance is addressed by the main control unit within a specific operating segment where the blade feathering rate is fixed at 4.0° / s. This is achieved by applying a discrete electromagnetic excitation step signal of known amplitude and acquiring the transient damped response curve of the pitch torque current variation rate dl / dt. The convergence time of the response curve is then linearly fitted to the standard dynamic decay characteristic curve using the least squares method to complete the torque-time conversion constant. In-situ numerical updates are performed. During the operation segment where the blade feathering rate is fixed at 4.0 degrees per second, the main control unit injects a discrete electromagnetic step current signal with an amplitude of 5 amperes and a pulse width of 50 milliseconds into the pitch current loop. The pitch motor control loop outputs a current damping response curve. The main control unit collects the convergence time of the curve and inputs it and the wear-free standard dynamic decay time line into the least squares iterative fitting process to calculate the electrical decay time difference caused by nonlinear clearance wear of the mechanical transmission. Based on the time difference, the torque-time conversion constant correction increment is calculated, and the torque-time conversion constant inside the non-volatile memory is updated as the input parameter for calculating the time cascade lag increment. In practical applications, the physical mechanism by which the current damping response driven by the above electromagnetic excitation signal can map mechanical wear is that when the gears of the pitch drive unit experience mechanical wear, the clearance on the transmission side increases. This results in the superposition of nonlinear gap clutter in the transmission reaction torque experienced by the pitch motor in the constant speed feathering section. This mechanical load fluctuation is directly fed back to the electromagnetic torque loop of the motor through the motor shaft. When the main control unit injects a discrete electromagnetic step current signal of 5 amps, the response curve of the motor electromagnetic control loop is not only affected by the inherent impedance and inductive reactance of the circuit, but also deeply coupled with the back electromotive force fluctuation caused by the nonlinear gap of the mechanical transmission. By extracting the convergence time of the response curve, the decay time sequence characteristics of the mechanical gap clutter in the electrical loop are essentially captured. By performing least squares linear fitting with the preset standard dynamic decay time line, the electrical decay time difference can be decoupled and quantitatively converted into the correction increment of the torque time conversion constant. Thus, without disassembling the unit, the pitch torque timing conversion channel containing mechanical transmission errors can be reconstructed in situ.

[0039] Through the collaboration of a fourth-order Butterworth high-pass digital filter and a bias signal, the resulting displacement response sequence converges to a zero-mean trajectory in the time domain, and the geometric pseudo-displacement components... The residual variance was reduced to below 0.0001 m², completely filtering out the spurious frequency drift components caused by sensor zero drift, and the calculated relative frequency drift rate was thus improved. The measurement level is maintained at a stable 0.0021 under non-destructive conditions, accurately aligning with the first-level health status judgment benchmark. This parameter reconstruction and adaptive calibration process eliminates the non-structural stiffness bias caused by hardware performance degradation during long-term operation. The state preservation decision unit has specific attributes for identifying the propagation of microcracks and stiffness degradation in the internal structure of the hybrid tower. This enables the system to output decision data with specific reliability when facing the dual conditions of environmental parameter changes and component aging. This data is used to provide automatic monitoring and safety protection for long-term stiffness degradation and internal crack evolution in conventional wind farms where large-capacity high-tower hybrid tower wind turbine generators are connected to the grid.

[0040] Example 5: When the system faces the pre-commissioning and grid connection of a new unit, due to the differences in the stiffness of the foundation soil and the manufacturing tolerances of the tower steel, the main control unit initiates the initial baseline calibration process before the unit is put into operation to determine the reference value f0 of the first-order natural frequency of the non-destructive standard. During the static time period when the ambient wind speed is below 1.5 m / s and there is no grid-connected load, the main control unit issues a command to the yaw drive mechanism to drive the nacelle to perform a 90° forward and reverse yaw step reversal action. The transient inertial torque applied to the top of the tower at the moment of yaw start and brake is used as a pulse excitation source to excite the hybrid tower structure to generate a pure free decay vibration response.

[0041] Within a 30-second window following the excitation by the transient inertial moment, the main control unit collects acceleration response data at the top of the tower using accelerometers, filters out noise components with amplitudes below 0.002 m / s², and uses the discrete Fourier transform operator to convert the time-domain waveform into a frequency-domain power spectral density curve. By searching for the energy maxima peak points within the 0.1 Hz to 1.0 Hz frequency band of the curve, the natural vibration frequency of the structure under the current constraints is identified. This process is repeated five times, and the arithmetic mean is calculated. This average value is then written in-situ into the storage area as the reference value f0 for the first-order natural frequency without damage. This eliminates the interference of initial physical environment differences on damage identification and ensures the relative frequency drift rate under subsequent power generation conditions. The calculations have a consistent alignment reference.

[0042] The above description is only a few preferred embodiments of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, technical solutions formed by replacing the above-mentioned features with the technical features with similar functions disclosed in the embodiments of the present invention.

Claims

1. A method for monitoring the status of a hybrid tower fan, characterized in that, Includes the following steps: Step S101: Obtain the operation monitoring data of the wind turbine control system. The operation monitoring data includes the pitch torque current variation rate, the actual pitch angle sequence, and the tower top orthogonal bidirectional acceleration sequence. Step S102: Based on the zero-lift pitch angle critical limit, establish aerodynamic load feedforward decoupling rules. When the timing start point of the synchronous feathering action of all blades is captured, lock the dynamic transition time window with a time increment of 15s to 45s. When the actual pitch angle sequence touches the zero-lift pitch angle critical limit and the absolute value of the pitch torque current variation rate is greater than the set drag threshold, establish the intrinsic breakpoint as the timing node after shifting the discrete time point that touches the zero-lift pitch angle critical limit backward by the time cascade lag increment. The time cascade lag increment is the product of the torque-time conversion constant and the pitch torque current variation rate. Using the intrinsic breakpoint as the timing boundary, divide the dynamic transition time window into the first aerodynamic damping modulation segment and the second pure structural intrinsic free decay segment, and separate and remove the acceleration data in the first aerodynamic damping modulation segment. Step S103: The orthogonal bidirectional acceleration sequence at the top of the tower in the pure structural intrinsic free decay section is converted into a dynamic displacement response sequence using the discrete integral operator. At the transient node of the total pitch torque, the geometric pseudo-displacement component caused by the deformation of the nacelle and frame is subtracted to repair the initial amplitude maximum reference and generate the intrinsic free decay response sequence of the hybrid tower. The intrinsic free decay response sequence of the hybrid tower is analyzed to identify the longitudinal true natural frequency and the transverse true natural frequency of the hybrid tower structure.

2. The method for monitoring the status of a hybrid tower fan according to claim 1, characterized in that, Step S103 specifically includes the following sub-steps: Step S1031, synchronously analyze the intrinsic free decay response sequence of the mixed tower, independently calculate the longitudinal true natural frequency and the transverse true natural frequency, and extract the numerical difference between the longitudinal true natural frequency and the transverse true natural frequency; Step S1032, when the numerical difference is greater than the set boundary variation threshold, identify the external geometric constraint boundary variation caused by uneven foundation settlement, and introduce a weight correction operator when calculating the relative frequency drift rate to remove the frequency evolution component caused by the geometric constraint change.

3. The method for monitoring the status of a hybrid tower fan according to claim 1, characterized in that, Step S102 specifically includes the following sub-steps: Step S1021, import the aerodynamic design data of the wind turbine blades into the wind turbine main control unit; Step S1022, extract the zero-lift pitch angle critical value from the aerodynamic design data, convert the zero-lift pitch angle critical value into a feedforward separation threshold, so as to establish the aerodynamic load feedforward decoupling rule.

4. The method for monitoring the status of a hybrid tower fan according to claim 1, characterized in that, The repair of the initial amplitude maximum reference in step S103 specifically includes the following sub-steps: Step S1033, by obtaining the pitch torque current variation rate to characterize the transient dynamic flexible load of the nacelle frame caused by rapid feathering; Step S1034, by combining the actual pitch angle sequence and the tower top orthogonal bidirectional acceleration sequence in the pure structural intrinsic free decay section, the dynamic displacement response sequence is calculated by the discrete integral operator; Step S1035, at the transient node of the total pitch torque, the geometric pseudo-displacement component of the nacelle frame caused by the transient dynamic flexible load of the nacelle frame is calculated, and the geometric pseudo-displacement component is subtracted from the dynamic displacement response sequence to complete the in-situ repair of the initial amplitude maximum reference, eliminating the low-frequency tilting interference introduced by the local elastic deformation to the nacelle accelerometer used to collect the tower top orthogonal bidirectional acceleration sequence.

5. The method for monitoring the status of a hybrid tower fan according to claim 2, characterized in that, The introduction of a weight correction operator to remove frequency evolution components caused by geometric constraint variations specifically includes the following sub-steps: Step S1036, during the calculation of the relative frequency drift rate, when the orthogonal frequency deviation value does not exceed the set boundary variation threshold, the basic parameter calculation channel is kept running; Step S1037, when the orthogonal frequency deviation value exceeds the set boundary variation threshold, the parameter compensation channel is activated to separate and remove the frequency evolution components caused by external geometric constraint boundary variations from the total frequency drift, and the remaining pure frequency drift rate is mapped to the stiffness degradation state of the hybrid tower body.

6. The method for monitoring the status of a hybrid tower fan according to claim 5, characterized in that, Mapping the remaining pure frequency drift rate to the stiffness degradation state of the hybrid tower body specifically includes the following sub-steps: Step S1038, continuously record the pure frequency drift rate of the wind turbine during the operating cycle, and establish a time series benchmark that evolves with operating time and load accumulation; Step S1039, extract the changing trend characteristics of the time series benchmark to calculate a quantitative index characterizing the stiffness reduction rate of the hybrid tower connection part; Step S10310, when the quantitative index exceeds the set safety threshold, output the corresponding state prediction and early warning signal of microcrack propagation and stiffness degradation inside the concrete.

7. The method for monitoring the status of a hybrid tower fan according to claim 1, characterized in that, In acquiring the operation monitoring data of the wind turbine control system, the sampling frequency of the nacelle accelerometer for acquiring the orthogonal bidirectional acceleration sequence at the top of the tower is set to no less than 100Hz, and the single intercept time length of the operation monitoring data is set to no less than 600s, so as to ensure that the segmented pure structural intrinsic free decay section has a complete free decay history.

8. The method for monitoring the status of a hybrid tower fan according to claim 1, characterized in that, Step S103 specifically includes the following sub-steps: Step S10311, compare the identified longitudinal true natural frequency and transverse true natural frequency with the design natural frequency in the initial state of the hybrid tower, and calculate the total frequency drift; Step S10312, search the set hybrid tower damage mapping relationship library, match the fatigue damage degree of the concrete connection part corresponding to the total frequency drift, and determine the structural health level of the hybrid tower structure.

9. The method for monitoring the status of a hybrid tower fan according to claim 8, characterized in that, After determining the structural health level of the hybrid tower structure, the following steps are included: Step S10313, when the structural health level reaches the danger control level, the main control unit of the wind turbine sends a vibration damping adjustment command to the pitch drive unit of the wind turbine. By changing the pitch rate and speed control trajectory within the operating wind speed range, the wind turbine avoids the resonance frequency band of the hybrid tower structure, thus ensuring the safe operation of the hybrid tower wind turbine.

Images