Stability Analysis Methods, Equipment, and Media for Offshore Wind Turbines Based on Operating Parameters

Abstract

This invention provides a method, equipment, and medium for stability analysis of offshore wind turbines based on operating parameters, belonging to the field of new energy technology. It solves the technical problem that the operational stability of offshore wind turbines cannot be effectively monitored due to the difficulty in real-time identification of oscillation risks in weak grid environments. The technical solution is as follows: S1, collecting electrical and mechanical parameters generated by the offshore wind turbine during operation; S2, preprocessing the electrical and mechanical parameters; S3, constructing a dynamic mapping model to characterize the relationship between the wind turbine and the oscillation response; S4, calculating the integral result of the oscillation amplitude and the oscillation energy within the time window. The beneficial effects of this invention are: by directly analyzing the turbine's operating parameters, this method can promptly detect potential operational instability states, providing decision support for maintenance personnel and helping to improve the operational safety of offshore wind turbines in weak grid environments.

Description

Technical Field

[0001] This invention relates to the field of new energy technology, and in particular to a method, equipment and medium for stability analysis of offshore wind turbines based on operating parameters. Background Technology

[0002] With the rapid development of renewable energy, offshore wind power has gradually become an important pillar of global energy structure transformation. Due to the stability of offshore wind resources and high area utilization efficiency, offshore wind power projects are being deployed at an accelerated pace globally. However, offshore wind farms are typically connected to the terrestrial power grid via long-distance submarine cables, resulting in high grid impedance and weak voltage support, creating a typical weak grid environment. Under these conditions, wind turbines are more susceptible to external disturbances and internal control interactions, leading to operational stability issues, including subsynchronous oscillations, control system coupled oscillations, and high-frequency instability. Therefore, operational stability analysis of offshore wind turbines has become one of the key technologies for ensuring safe grid connection and efficient operation of wind power.

[0003] Currently, the analysis of wind turbine stability mainly employs modeling and simulation-based methods, with the impedance method being the most widely used. This method establishes a frequency domain model of the wind turbine's output impedance and the grid's equivalent impedance, analyzes the amplitude and phase relationship between the two at specific frequencies, and uses the Nyquist criterion to determine the system's stability. The impedance method can accurately reflect the interaction characteristics between the turbine and the grid, especially advantageous in studying subsynchronous oscillations and controller interactions. However, this method is highly dependent on model and controller parameters, which are difficult to obtain, and it is not suitable for rapid diagnosis and analysis during field operation. Another common method is eigenvalue analysis, which is based on solving for eigenvalues ​​in a state-space linearized model of the wind power system to evaluate dynamic stability under small disturbances. Although eigenvalue analysis has a mature theoretical foundation and provides intuitive and clear results, it also suffers from complex modeling, time-consuming computation, and difficulty in handling nonlinear and large-disturbance conditions. In addition, time-domain simulation methods are also widely used in the research and verification of the dynamic behavior of wind power systems, but they rely on a complete simulation platform and detailed model configuration, requiring significant computational resources and making online application difficult.

[0004] Given the shortcomings of traditional methods in terms of real-time performance, model dependence, and system adaptability, a class of stability analysis methods based on operating parameters has gradually emerged in recent years. These methods do not rely on detailed models of the turbine controller but directly collect raw data generated by the wind turbine during operation, such as voltage, current, power, speed, and torque. Through signal processing techniques such as frequency domain analysis, time-frequency transformation, or pattern recognition, frequency components representing the dynamic characteristics of the system are extracted, with particular attention paid to the changing trends of subsynchronous and supersynchronous oscillation frequency bands. Based on this, a mapping relationship between parameter changes and oscillation characteristics is established to achieve real-time assessment and early warning of the wind turbine's operating status. Compared to traditional methods, operating parameter-based analysis is easier to deploy, applicable to various types of turbines, and possesses good versatility and real-time performance, making it particularly suitable for online monitoring and dynamic response management in weak grid environments. Therefore, researching a method and system oriented towards operating parameters, with oscillation feature extraction capabilities and stability assessment functions, has significant engineering application value and practical significance. Summary of the Invention

[0005] The purpose of this invention is to provide a method, equipment, and medium for stability analysis of offshore wind turbines based on operating parameters. This method can promptly identify potential unstable operating conditions by directly analyzing the operating parameters of the turbine, providing decision support for operation and maintenance personnel and helping to improve the operational safety of offshore wind turbines in weak grid environments.

[0006] The inventive concept of this invention is as follows: The method includes: real-time acquisition of electrical parameters (including voltage, current, and power) and mechanical parameters (including speed and torque) of grid-connected wind turbine units; processing the acquired operating parameters using frequency domain analysis to extract characteristic frequency components of subsynchronous and supersynchronous oscillations; establishing a dynamic mapping model between operating parameters and oscillation characteristics based on the extracted oscillation characteristics; identifying parameters using the least squares method to assess the unit's operational stability; and generating an operational instability warning signal when the oscillation energy index exceeds a set threshold. The system includes a data acquisition module, a feature analysis module, and a status assessment module, which work collaboratively to achieve real-time monitoring and analysis of the unit's operating status.

[0007] To achieve the aforementioned objectives, the present invention employs the following technical solution: a method, equipment, and medium for stability analysis of offshore wind turbines based on operating parameters, comprising:

[0008] The electrical and mechanical parameters generated by the offshore wind turbine during operation are collected. The electrical parameters include three-phase voltage, current, instantaneous active power and reactive power at the grid connection point, and the mechanical parameters include generator speed, torque and state quantities such as pitch angle and yaw angle.

[0009] The electrical and mechanical parameters are preprocessed, and data noise is removed by moving average filtering. Frequency domain decomposition is used to extract the subsynchronous and supersynchronous oscillation feature components present during the wind turbine's operation, constructing the following feature vector set:

[0010] (1)

[0011] in, This is the subsynchronous frequency component; This refers to the supersynchronous frequency component; A This represents the oscillation amplitude. This is the phase offset. It is oscillating energy;

[0012] A dynamic mapping model is constructed to characterize the relationship between wind turbine generators and oscillation response. This mapping model is based on the following functional relationship:

[0013] (2)

[0014] in, t For time, Rotational speed; Electromagnetic torque; This refers to the phase difference between the generator set and the power grid. The rate of change of rotational speed reflects the intensity of the mechanical disturbance. The electromagnetic torque change rate reflects the response to electrical disturbances; These are the characteristic weighting coefficients for the rate of change of rotational speed, the rate of change of electromagnetic torque, and the phase difference, respectively. For high-frequency residual disturbances;

[0015] Based on the integral of the oscillation amplitude and the oscillation energy within the time window, and the total energy... The oscillation energy ratio index was obtained through comparative calculation. and with preset threshold Compare;

[0016] when At that time, an early warning signal for unstable unit operation is generated. The early warning signal includes: oscillation frequency energy distribution characteristic parameters, oscillation energy change trend, and deviation degree of key operating parameters.

[0017] The analysis of electrical and mechanical parameters includes the following steps:

[0018] The collected electrical and mechanical parameter signals are respectively subjected to moving average filtering, with a filtering window width of N. The filtered signals are represented as follows:

[0019] (3)

[0020] in: The original electrical parameter signal; This is the original mechanical parameter signal; N The width of the sliding window; The sampling interval time. These are the filtered electrical and mechanical signals, respectively.

[0021] Perform Fast Fourier Analysis on the filtered signal to calculate the power spectral density, and extract the oscillatory components to construct the feature vector:

[0022] (4)

[0023] in; The power spectral density of the electrical parameter signal; The power spectral density of the mechanical parameter signal; T To analyze the length of the time window; f π is the frequency; j is the imaginary unit; π is the value of pi. These are the threshold coefficients for electrical and mechanical parameters, respectively; their values ​​range from 0.05 to 0.2, with an optimal value of 0.1.

[0024] The construction and parameter identification process of the dynamic mapping model is as follows:

[0025] A dynamic mapping model is constructed to characterize the relationship between the wind turbine and the oscillation response, as shown in Equation (2);

[0026] Parameter identification is performed using recursive least squares with a forgetting factor:

[0027] First, perform initialization settings and define initial values ​​for the parameter vector. ;

[0028] Initial values ​​of the covariance matrix (I is a 3x3 identity matrix);

[0029] Forgetting factor (Value range: 0.95~0.99); the optimal value is 0.98.

[0030] Sampling period .

[0031] Real-time parameter updates ( k (Time calculation)

[0032] Constructing a data vector: , T Indicates matrix transpose;

[0033] Calculate prior error ;

[0034] Update the gain matrix ;

[0035] Parameter estimation update: ;

[0036] Covariance matrix update: .

[0037] The coefficient of determination is used to evaluate the model fit:

[0038] (5)

[0039] in, These are the model's predicted values; This represents the average amplitude of the oscillations. N To verify the window length, the coefficient of determination is required. Otherwise, a model reset will be triggered.

[0040] According to the oscillation amplitude With window time The integral result of the internal oscillation energy is used to calculate the oscillation energy ratio index. :

[0041] (6)

[0042] in, The average value of the oscillation energy under rated operating conditions of the unit is taken as the reference energy benchmark value; The integration time window, ranging from 5 to 10 power frequency cycles, is used to compare the oscillation energy with a preset threshold. Compare them.

[0043] when Generate an early warning signal for unstable unit operation, the early warning signal including:

[0044] Oscillation frequency energy distribution characteristic parameters:

[0045]

[0046] in, Energy for subsynchronous and supersynchronous frequency bands.

[0047] Oscillation energy change trend:

[0048] Deviation of key operating parameters:

[0049]

[0050] in, These are the initial values ​​for speed, torque, and phase.

[0051] To achieve the above-mentioned objectives, the present invention also provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, the processor being used to call and run the computer program stored in the memory to perform the method described in any of the preceding claims.

[0052] To achieve the above-mentioned objectives, the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method described in any of the preceding claims.

[0053] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0054] 1. This invention provides a method and system for stability analysis of offshore wind turbines based on operating parameters. It can effectively extract potential subsynchronous and supersynchronous oscillation characteristics in the system by performing frequency domain analysis on actual collected electrical parameters (including voltage, current, and power) and mechanical parameters (including speed and torque) without relying on detailed controller models or grid topology information. This enables real-time monitoring and stability assessment of the wind turbine's operating status. The method establishes a dynamic mapping model between operating parameters and oscillation behavior, accurately reflecting the dynamic interaction characteristics during turbine operation. When the oscillation energy index exceeds a set threshold, a timely warning signal is generated, helping to identify potential instability states in advance and improving the operational safety of wind farms.

[0055] 2. Compared with existing methods based on impedance modeling, eigenvalue calculation, or simulation analysis, this invention eliminates the need for complex modeling processes, offering higher real-time performance and field adaptability. It is particularly suitable for offshore wind farm environments with complex grid conditions and incomplete modeling information. Furthermore, the analysis system constructed in this invention is highly modular, integrating multiple functional units such as data acquisition, feature analysis, and state assessment. This facilitates deployment at wind farm control centers or edge nodes, enabling unified monitoring and dynamic management of multiple wind turbine units, demonstrating excellent engineering practicality and promotional value. Attached Figure Description

[0056] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0057] Figure 1 This is a flowchart of the stability analysis of offshore wind turbines in an embodiment of the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0059] Example 1: The technical solution provided in this example is: a stability analysis method for offshore wind turbines based on operating parameters, including:

[0060] The electrical and mechanical parameters generated by the offshore wind turbine during operation are collected. The electrical parameters include three-phase voltage, current, instantaneous active power and reactive power at the grid connection point, and the mechanical parameters include generator speed, torque and state quantities such as pitch angle and yaw angle.

[0061] The electrical and mechanical parameters are preprocessed, and data noise is removed by moving average filtering. Frequency domain decomposition is used to extract the subsynchronous and supersynchronous oscillation feature components present during the wind turbine's operation, constructing the following feature vector set:

[0062]

[0063] in, This is the subsynchronous frequency component; This refers to the supersynchronous frequency component; A This represents the oscillation amplitude. This is the phase offset. It is oscillating energy;

[0064] A dynamic mapping model is constructed to characterize the relationship between wind turbine generators and oscillation response. This mapping model is based on the following functional relationship:

[0065] (1)

[0066] in, Rotational speed; Electromagnetic torque; This refers to the phase difference between the generator set and the power grid. The rate of change of rotational speed reflects the intensity of the mechanical disturbance. The electromagnetic torque change rate reflects the response to electrical disturbances; These are the characteristic weighting coefficients for the rate of change of rotational speed, the rate of change of electromagnetic torque, and the phase difference, respectively, obtained based on historical operating data; For high-frequency residual disturbances;

[0067] The oscillation energy ratio index is calculated based on the integral result of the oscillation amplitude and the oscillation energy within the time window. and with preset threshold Compare;

[0068] when At that time, an early warning signal for unstable unit operation is generated. The early warning signal includes: oscillation frequency energy distribution characteristic parameters, oscillation energy change trend, and deviation degree of key operating parameters.

[0069] Example 2: This example further develops upon Example 1 by analyzing the electrical and mechanical parameters:

[0070] The collected electrical and mechanical parameter signals are processed by moving average filtering, with a filtering window width of N. The filtered signals are represented as follows:

[0071] (2)

[0072] in: The original electrical parameter signal; This is the original mechanical parameter signal; N The width of the sliding window; This is the sampling interval time; These are the filtered electrical and mechanical signals, respectively.

[0073] Fast Fourier analysis is performed on the filtered electrical and mechanical signals to calculate the power spectral density, and oscillatory components are extracted to construct eigenvectors.

[0074] (3)

[0075] in; The power spectral density of the electrical parameter signal; The power spectral density of the mechanical parameter signal; T To analyze the length of the time window; f For frequency; This is the threshold coefficient (ranging from 0.05 to 0.2).

[0076] Example 3: This example further designs the construction and parameter identification process of the dynamic mapping model based on Example 2:

[0077] A dynamic mapping model is constructed to characterize the relationship between wind turbine and oscillation response. The mapping model is based on the following functional relationship, as shown in Equation (1).

[0078] Parameter identification is performed using recursive least squares with a forgetting factor:

[0079] First, perform initialization settings and define initial values ​​for the parameter vector. ;

[0080] Initial values ​​of the covariance matrix (I is a 3x3 identity matrix);

[0081] Forgetting factor (Value range: 0.95~0.99);

[0082] Sampling period .

[0083] Parameters updated in real time (calculated at time k):

[0084] Constructing a data vector: ;

[0085] Calculate prior error ;

[0086] Update the gain matrix ;

[0087] Parameter estimation update: ;

[0088] Covariance matrix update: .

[0089] The coefficient of determination is used to evaluate the model fit:

[0090] (4)

[0091] in, These are the model's predicted values; This represents the average amplitude of the oscillations. N To verify the window length, the coefficient of determination is required. Otherwise, a model reset will be triggered.

[0092] Example 4: This example is a further design based on Example 3:

[0093] According to the oscillation amplitude With window time The integral result of the internal oscillation energy is used to calculate the oscillation energy ratio index. :

[0094] (5)

[0095] in, The average value of the oscillation energy under the rated operating conditions of the unit is used as a reference energy benchmark. The integration time window, ranging from 5 to 10 power frequency cycles, is used to compare the oscillation energy with a preset threshold. Compare them.

[0096] Example 5: This example is a further design based on Example 4:

[0097] when Generate an early warning signal for unstable unit operation, the early warning signal including:

[0098] Oscillation frequency energy distribution characteristic parameters:

[0099]

[0100] in, Energy for subsynchronous and supersynchronous frequency bands.

[0101] Oscillation energy change trend:

[0102] Deviation of key operating parameters:

[0103]

[0104] in, These are the initial values ​​for speed, torque, and phase.

[0105] Example 6: An electronic device including a memory and a processor, the memory storing a computer program, the processor being used to call and run the computer program stored in the memory to perform the method as described in any of the preceding examples.

[0106] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method as described in any of the preceding claims.

[0107] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A stability analysis method for offshore wind turbines based on operating parameters, characterized in that, include: S1. Collect electrical and mechanical parameters generated by offshore wind turbines during operation. Electrical parameters include three-phase voltage, current, instantaneous active power and reactive power at the grid connection point. Mechanical parameters include generator speed, torque, pitch angle and yaw angle. S2. The electrical and mechanical parameters are preprocessed, data noise is removed by moving average filtering, and the subsynchronous and supersynchronous oscillation feature components existing in the operation of the offshore wind turbine are extracted by frequency domain decomposition to construct the following feature vector set: （1） in, This is the subsynchronous frequency component; This refers to the supersynchronous frequency component; A This represents the oscillation amplitude. This is the phase offset. It is oscillating energy; S3. Construct a dynamic mapping model to characterize the relationship between offshore wind turbines and oscillation response, the mapping model being based on the following functional relationship: （2） in, Rotational speed; Electromagnetic torque; This refers to the phase difference between the generator set and the power grid. The rate of change of rotational speed reflects the intensity of the mechanical disturbance. The electromagnetic torque change rate reflects the response to electrical disturbances; These are the characteristic weighting coefficients for the rate of change of rotational speed, the rate of change of electromagnetic torque, and the phase difference, respectively, obtained based on historical operating data; For high-frequency residual disturbances; S4. Calculate the oscillation energy ratio index based on the integral result of the oscillation amplitude and the oscillation energy within the time window. and with preset threshold Compare; when At that time, an early warning signal for unstable operation of offshore wind turbines is generated. The early warning signal includes: oscillation frequency energy distribution characteristic parameters, oscillation energy change trend, and deviation degree of key operating parameters.

2. The offshore wind turbine stability analysis method according to claim 1, characterized in that, In step S1, the analysis of electrical and mechanical parameters includes the following steps: S11. Perform moving average filtering on the collected electrical and mechanical parameter signals respectively. The filtering window width is N. The filtered signals are represented as follows: （3） in: The original electrical parameter signal; This is the original mechanical parameter signal; N The width of the sliding window; This is the sampling interval time; These are the filtered electrical signal and the filtered mechanical signal, respectively. S12. Perform Fast Fourier Analysis on the filtered electrical and mechanical signals, calculate the power spectral density, and extract the oscillating components to construct the feature vector: （4） in; The power spectral density of the electrical parameter signal; The power spectral density of the mechanical parameter signal; T To analyze the length of the time window; f For frequency; These are the threshold coefficients for electrical and mechanical parameters, respectively, with values ​​ranging from 0.05 to 0.

2.

3. The offshore wind turbine stability analysis method according to claim 2, characterized in that, In step S3, the construction and parameter identification process of the dynamic mapping model includes the following steps: S31. Construct a dynamic mapping model to characterize the relationship between offshore wind turbine and oscillation response, as shown in Equation (2); S32. Parameter identification is performed using the recursive least squares method with a forgetting factor: First, perform initialization settings and define initial values ​​for the parameter vector. ; Initial values ​​of the covariance matrix I is a 3x3 identity matrix; Forgetting factor Its value ranges from 0.95 to 0.99; Sampling period ; Real-time parameter k-time calculation: S33. Constructing a data vector: ; Calculate prior error ; Update the gain matrix ; Parameter estimation update: ; Covariance matrix update: ; The coefficient of determination is used to evaluate the model fit: （5） in, These are the model's predicted values; This represents the average amplitude of the oscillations. N To verify the window length, the coefficient of determination is required. Otherwise, a model reset will be triggered.

4. The offshore wind turbine stability analysis method according to claim 3, characterized in that: In step S4, based on the oscillation amplitude With window time The integral result of the internal oscillation energy is used to calculate the oscillation energy ratio index. : （6） in, The average value of the oscillation energy under rated operating conditions of the unit is taken as the reference energy benchmark value; The integration time window, ranging from 5 to 10 power frequency cycles, is used to compare the oscillation energy with a preset threshold. Compare them.

5. The offshore wind turbine stability analysis method according to claim 4, characterized in that: In step S4 when Generate an early warning signal for unstable unit operation, the early warning signal including: Oscillation frequency energy distribution characteristic parameters; （7） in, Energy for subsynchronous and supersynchronous frequency bands; Oscillation energy change trend: ; Deviation of key operating parameters: ； in, These are the initial values ​​for each parameter.

6. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor being used to invoke and run the computer program stored in the memory to perform the method as described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, A computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1-5.

Images