Offshore wind power transient control method and system

Abstract

This invention provides a transient control method and system for offshore wind power. The method includes: collecting voltage, current, and frequency data of the offshore wind power system; performing data preprocessing to obtain operating status information; analyzing fault characteristics based on the operating status information; establishing equipment overcurrent capacity constraints and converter modulation constraints, and constructing a discrete-time robust predictive control barrier function; constructing a wind turbine dynamic model considering mechanical-electrical coupling characteristics, and executing an iterative learning predictive control algorithm; configuring control parameters, executing a transient control strategy, and dynamically adjusting it. This invention, by comprehensively considering equipment constraints and mechanical-electrical coupling characteristics, achieves optimized transient control of the offshore wind power system under various fault conditions, improving the system's grid support capability and operational reliability.

Description

Technical Field

[0001] This invention relates to the field of power system control technology, specifically to a transient control method and system for offshore wind power. Background Technology

[0002] As an important component of clean energy, the stability and reliability of offshore wind power's grid connection have a significant impact on the safe operation of the power system. Offshore wind power systems need to possess rapid response and support capabilities when facing grid faults and disturbances.

[0003] Currently, common wind power grid connection control technologies mainly include two methods: voltage source control and current source control. Voltage source control has better grid support capabilities, but it is prone to overcurrent problems under fault conditions; current source control has good overcurrent protection capabilities, but its grid support capabilities are weaker.

[0004] The most relevant existing technology employs a control strategy based on a virtual synchronous machine, simulating the characteristics of a synchronous generator to achieve grid-connected operation of wind turbines. This technology achieves wind turbine support for grid frequency and voltage by designing virtual inertia and damping parameters.

[0005] However, this technology has the following problems when facing symmetrical / asymmetrical short-circuit faults in the system: 1) It does not fully consider the overcurrent capacity constraints of the equipment, which may lead to damage to the converter; 2) Under asymmetrical faults, the converter is prone to overmodulation, which affects the control effect; 3) It lacks in-depth consideration of the mechanical-electrical coupling characteristics of the wind turbine, which affects the transient performance of the system.

[0006] Therefore, a new transient control strategy for offshore wind power is needed that can fully leverage the grid support capabilities of offshore wind power systems while ensuring equipment safety. Summary of the Invention

[0007] To address the problems existing in the prior art, this invention provides a method and system for transient control of offshore wind power. This method can achieve optimized transient control of the offshore wind power system under various fault conditions while ensuring equipment safety, thereby improving the grid support capability and operational reliability of the system.

[0008] The technical solution of this invention is: A transient control method for offshore wind power includes: The voltage, current, and frequency data of the offshore wind power system are collected, and the voltage, current, and frequency data are filtered and time-stamped to obtain the operating status information of the offshore wind power system. Based on the operating status information of the offshore wind power system, a sliding time window detection algorithm is used to determine whether the offshore wind power system has a fault. If a fault occurs, the fault type is analyzed by symmetric component decomposition to obtain fault characteristic parameters. Based on the fault characteristic parameters, equipment overcurrent capacity constraints and converter modulation constraints are established, a robust predictive control barrier function for discrete-time offshore wind power system is constructed, and a constraint model framework is generated. Based on the aforementioned constraint model framework, a wind turbine dynamic model considering mechanical-electrical coupling characteristics is constructed. The wind turbine dynamic model is then used to execute an iterative learning predictive control algorithm to generate a sequence of control commands. Configure control parameters according to the control command sequence, execute transient control strategies, and dynamically adjust the control parameters based on real-time monitoring results.

[0009] Preferably, voltage, current, and frequency data of the offshore wind power system are collected, and the voltage, current, and frequency data are filtered and time-stamped to obtain the operating status information of the offshore wind power system, including: The voltage, current and frequency data of the offshore wind power system are collected as input, the voltage deviation rate is calculated, and voltage assessment data is obtained. The frequency deviation rate is calculated based on the voltage evaluation data as input, and the frequency evaluation data is obtained. Based on the voltage assessment data and the frequency assessment data as input, a weighted average algorithm is executed to perform data aggregation analysis and output the operating status information of the offshore wind power system.

[0010] Preferably, based on the operating status information of the offshore wind power system, a sliding time window detection algorithm is used to determine whether a fault has occurred in the offshore wind power system. If a fault occurs, the fault type is analyzed through symmetric component decomposition to obtain fault characteristic parameters, including: The system receives the operating status information of the offshore wind power system as input, compares the deviation between the voltage of the offshore wind power system and the voltage threshold, and outputs the voltage fault detection result. The offshore wind power system's operating status information is used as input, and the deviation between the offshore wind power system's current and the rated current value is compared to output the current fault detection result. Based on the voltage fault detection results and the current fault detection results as input, a fault determination matrix algorithm is executed for comprehensive analysis to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault characteristic parameters are output, including the fault type, fault severity, and fault duration.

[0011] Preferably, a comprehensive analysis is performed based on the voltage fault detection results and the current fault detection results to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault characteristic parameters are obtained, including: The voltage and current of the offshore wind power system are used as inputs to perform positive-sequence component, negative-sequence component and zero-sequence component decomposition, and the component decomposition data is output. Using the component decomposition data as input, the ratio of the negative-order component to the base value is calculated, and asymmetric fault judgment data is output. The fault type is determined by executing a fault type identification algorithm based on the asymmetric fault judgment data as input, the fault feature parameters are updated, and the results are output for subsequent constraint model construction.

[0012] Preferably, constraints on equipment overcurrent capacity and converter modulation are established to construct a robust predictive control barrier function for the discrete-time offshore wind power system, generating a constraint model framework, including: Based on the fault characteristic parameters as input, the current of the offshore wind power system is limited to no more than 1.2 times the rated current, and an overcurrent constraint condition is output. Based on the overcurrent constraint condition as input, the converter modulation ratio is limited within the linear modulation region, and the modulation constraint condition is output. Based on the overcurrent constraint and the modulation constraint as input, the control barrier function construction algorithm is executed to output the constraint model framework.

[0013] Preferably, based on the constraint model framework, a dynamic model of the wind turbine considering the mechanical-electrical coupling characteristics is constructed, including: Using the aforementioned constraint model framework as input, the wind turbine, drive shaft, and generator rotor are constructed as a multi-mass block structure model, and the inertia distribution data is output. Based on the inertia distribution data as input, the spring-damping relationship construction algorithm is executed to establish the spring-damping connection relationship and output mechanical dynamic characteristic data. Based on the mechanical dynamic characteristic data and electromagnetic characteristic parameters as input, a coupled model construction algorithm is executed to output the dynamic response characteristics of the wind turbine dynamic model.

[0014] Preferably, the method further includes: executing an iterative learning prediction control algorithm based on the dynamic response characteristics, including: Using the aforementioned dynamic response characteristics as input, the dynamic response of the offshore wind power system is predicted based on the state-space equation, and the predicted response data is output. The deviation between the predicted response data and the expected response is calculated as input, a parameter optimization algorithm is executed, and the updated control parameter values ​​are output. The control parameters are optimized by an iterative optimization algorithm based on the updated control parameter values ​​as input, and the control command sequence is output.

[0015] Preferably, configuring control parameters according to the control command sequence includes: The system receives the control command sequence as input, configures the current loop and voltage loop parameters of the converter, and outputs the initial control parameters. The initial control parameters are used as input to perform a piecewise linear interpolation algorithm, and the output is the smoothed control parameters. Using the smoothed control parameters as input, the execution parameter verification algorithm checks whether they are within the rated parameter range and outputs valid control parameters.

[0016] Preferably, the transient control strategy includes: Based on the control parameters as input, the control mode selection algorithm is executed to select either the positive sequence voltage support control mode or the negative sequence voltage suppression control mode, and the control mode signal is output. Based on the control mode signal as input, a voltage transient control algorithm is executed, and voltage control data is output. The frequency transient control algorithm is executed based on the voltage control data as input, outputs frequency control data, and dynamically adjusts the control parameters based on the frequency control data.

[0017] The present invention also provides a transient control system for offshore wind power, comprising: The data acquisition module is used to collect voltage, current and frequency data of the offshore wind power system, filter the data and synchronize the timestamp to obtain the operating status information of the offshore wind power system. The fault analysis module is used to determine whether the offshore wind power system has a fault based on the operating status information of the offshore wind power system, using a sliding time window detection algorithm, and to obtain fault characteristic parameters by analyzing the fault type through symmetric component decomposition. The constraint modeling module is used to establish equipment overcurrent capacity constraint and converter modulation constraint models based on the fault characteristic parameters, construct the robust predictive control barrier function of the discrete-time offshore wind power system, and obtain the constraint model framework. The control strategy module is used to construct a dynamic model of the wind turbine that considers the mechanical-electrical coupling characteristics based on the constraint model framework, execute an iterative learning predictive control algorithm, and generate a sequence of control commands. The control execution module is used to configure control parameters according to the control instruction sequence, execute transient control strategies, and dynamically adjust the control parameters based on real-time monitoring results.

[0018] The present invention has the following beneficial effects: 1. This invention proposes a multi-objective control method based on the robust predictive control barrier function of discrete-time systems. This method can achieve coordinated optimization of multiple control objectives such as voltage support and frequency support while ensuring the safe operation of the system.

[0019] 2. This invention designs an iterative learning predictive control strategy for constrained uncertain systems. By considering equipment overcurrent capacity constraints and modulation constraints, it achieves stable operation of wind turbine units under different fault types and grid support.

[0020] 3. This invention proposes a dynamic modeling method for wind turbines that considers the "mechanical-electrical" coupling characteristics, enabling accurate assessment and control of the active power support capacity of wind turbines under different operating states and time scales. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the embodiments will be briefly described below. These drawings are incorporated in and constitute a part of this specification. They illustrate embodiments conforming to this disclosure and, together with the specification, serve to explain the technical solutions of this disclosure. It should be understood that the following drawings only show some embodiments of this disclosure and should not be considered as limiting the scope. Those skilled in the art can obtain other related drawings based on these drawings without creative effort.

[0022] Figure 1 A flowchart of the offshore wind power transient control method provided by the present invention; Figure 2 A detailed flowchart of the system status monitoring steps provided by the present invention; Figure 3 A detailed flowchart of the fault feature analysis steps provided by this invention; Figure 4 A detailed flowchart of the constraint modeling steps provided for this invention; Figure 5 A detailed flowchart of the design steps for the multi-objective control strategy provided by this invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. The components of the embodiments of this disclosure described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this disclosure provided in the accompanying drawings is not intended to limit the scope of the claimed disclosure, but merely represents selected embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without inventive effort are within the scope of protection of this disclosure.

[0024] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0025] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Example

[0028] refer to Figure 1 The present invention provides a transient control method for offshore wind power, comprising the following steps: Step S1, System Status Monitoring: Collect voltage, current and frequency data of the offshore wind power system, filter and synchronize the voltage, current and frequency data to obtain the operating status information of the offshore wind power system.

[0029] System status monitoring is a fundamental component of the entire offshore wind power transient control strategy. Its purpose is to acquire and preprocess real-time operational data of the offshore wind power system, providing accurate and reliable data support for subsequent fault analysis and control strategy design. In the offshore wind farm environment, due to the instability of the marine climate and its remote location from land, acquiring high-quality system status data presents numerous challenges. To overcome these challenges, this invention employs a distributed data acquisition network, deploying high-precision measurement equipment at key nodes within the wind farm to collect data such as voltage, current, and frequency in real time.

[0030] The acquired raw data typically contains various noise and interference signals, requiring effective filtering. This system employs a combination of wavelet transform and Kalman filtering to effectively remove measurement noise while preserving the dynamic characteristics of the signal. Compared to traditional filtering algorithms, this method offers better time-frequency characteristics and the ability to handle non-stationary signals, making it particularly suitable for processing transient and transient signals commonly found in wind power systems.

[0031] Meanwhile, due to the vast distribution of offshore wind farms, data from different collection points exhibit time asynchrony, which can severely affect the accuracy of subsequent analysis. This invention employs GPS-based synchronization timestamp technology to strictly synchronize all collected data, ensuring data consistency. The system uses a precise clock synchronization mechanism to mark the collected data with accurate timestamps, controlling the error to the microsecond level, thus meeting the time accuracy requirements for power system transient process analysis.

[0032] The data, after filtering and time synchronization, will be used as input for system status assessment. After further calculation and analysis, complete operational status information of the offshore wind power system will be formed, providing a data foundation for subsequent fault detection and control strategy formulation.

[0033] like Figure 2 As shown, step S1 specifically includes: Step S1.1: Collect voltage, current and frequency data of the offshore wind power system as input, calculate the voltage deviation rate, and obtain voltage assessment data; The system first receives real-time voltage, current, and frequency data from a distributed data acquisition network. Voltage data includes the amplitude and phase information of the three-phase voltages; current data includes the amplitude and phase information of the three-phase currents; and frequency data reflects the instantaneous frequency state of the system. After preliminary preprocessing, the system calculates the voltage deviation rate, which is the percentage deviation between the actual voltage and the rated voltage. The calculation formula is:

[0034] The system calculates the deviation rate for each of the three phases of voltage and further analyzes the three-phase imbalance. The voltage deviation rate is a crucial indicator for judging the system's voltage quality and should normally remain within ±5%. Exceeding this range may indicate a voltage anomaly. The calculated voltage deviation rates constitute the voltage assessment data and serve as input for subsequent frequency deviation rate calculations.

[0035] Step S1.2: Calculate the frequency deviation rate based on the voltage evaluation data as input to obtain the frequency evaluation data; Based on the voltage assessment data, the system further calculates the frequency deviation rate. Frequency is a key parameter for the stable operation of a power system, and its deviation directly reflects the power balance of the system. The formula for calculating the frequency deviation rate is:

[0036] For the Chinese power grid, the rated frequency is 50Hz, and the frequency deviation should be controlled within ±0.2Hz during normal operation. The system uses the sliding window method to calculate the rate of change of frequency (ROCOF), which is an important indicator for evaluating system inertia and stability. The frequency deviation rate and ROCOF together constitute the frequency assessment data, providing an important basis for the comprehensive assessment of the system status.

[0037] Step S1.3: Based on the voltage assessment data and the frequency assessment data as input, perform a weighted average algorithm to aggregate and analyze the data, and output the operating status information of the offshore wind power system.

[0038] After obtaining voltage and frequency assessment data, the system needs to integrate this data into comprehensive information describing the system state. This step employs a weighted average algorithm for data aggregation analysis. This algorithm assigns weights to different parameters based on their importance to the system state assessment, constructing a comprehensive evaluation index.

[0039] Specifically, the algorithm first assigns weights to key indicators such as voltage deviation rate, frequency deviation rate, and ROCOF, typically with a weight of 0.5 for voltage, 0.3 for frequency, and 0.2 for ROCOF. Then, it calculates a weighted average to obtain a comprehensive system status index. Simultaneously, the system analyzes the trends of these indicators to determine if the system status is deteriorating. By setting different threshold ranges, the system classifies the operating status into four levels: normal, alert, abnormal, and emergency.

[0040] The weighted average algorithm demonstrates good robustness when processing multi-source heterogeneous data, effectively reducing the impact of individual parameter anomalies on the overall assessment. The final output of offshore wind power system operating status information includes status level, key indicator values ​​and their changing trends, providing a comprehensive and accurate basis for subsequent fault detection and analysis.

[0041] In this embodiment, the system status monitoring step involves deploying multiple measurement devices to collect real-time data such as system voltage, current, frequency, and power. The collected data undergoes filtering and synchronization timestamp processing to ensure accuracy and time consistency. Then, a system status evaluation index system is established, including voltage deviation rate, frequency deviation rate, etc., and these indicators are calculated in real time to output system operating status information. This information will provide a foundation for subsequent fault detection and analysis.

[0042] Step S2, Fault Feature Analysis: Based on the operating status information of the offshore wind power system, a sliding time window detection algorithm is used to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault type is analyzed by symmetric component decomposition to obtain fault feature parameters.

[0043] Fault characteristic analysis is a crucial step in the transient control of offshore wind power. Its goal is to quickly and accurately detect system faults and analyze their characteristics, providing a basis for the selection and execution of subsequent control strategies. Offshore wind power systems face a complex and diverse range of faults, including symmetrical short-circuit faults, asymmetrical short-circuit faults, and frequency disturbances. Different types of faults require different control strategies. Therefore, accurate fault detection and characteristic analysis are essential for the stable operation of the system.

[0044] In this method, a sliding time window detection algorithm is used for fault detection. This algorithm monitors changes in key electrical quantities in real time within a dynamically sliding observation window in the time domain. Compared with traditional threshold detection, the sliding window algorithm considers time continuity and can effectively reduce the false positive rate. The window size is typically set to 100-200 ms to balance detection speed and accuracy. When parameters such as voltage and current within the window exceed a preset threshold range, the system triggers a fault detection signal.

[0045] After a fault is detected, the system needs to further determine the fault type. This method uses symmetrical component theory to decompose the three-phase voltage and current into positive-sequence, negative-sequence, and zero-sequence components. By analyzing the amplitude and phase relationship of these components, the system can determine whether the fault is symmetrical or asymmetrical. For example, when there are significant negative-sequence voltage and current components (usually exceeding 5% of the base value), it indicates that an asymmetrical fault has occurred in the system; when the three-phase voltages decrease simultaneously and the negative-sequence component is small, it is determined to be a symmetrical fault.

[0046] After determining the basic fault type, the system further calculates fault characteristic parameters, including fault depth (voltage drop), fault duration, and fault phase information. These parameters are calculated using the least squares estimation method to improve the accuracy of parameter estimation and enhance anti-interference capabilities. These fault characteristic parameters will serve as crucial inputs for subsequent control strategy design, directly influencing the system's fault response strategy.

[0047] like Figure 3 As shown, step S2 specifically includes: Step S2.1: Receive the operating status information of the offshore wind power system as input, compare the deviation between the voltage of the offshore wind power system and the voltage threshold, and output the voltage fault detection result; In the first step of fault characteristic analysis, the system first receives operating status information from the system status monitoring module, focusing on voltage parameters. Voltage parameters of offshore wind power systems include the amplitude and phase information of the three-phase voltages; these parameters are primary indicators for judging grid faults. The system compares real-time voltage measurements with preset voltage thresholds to calculate the degree of deviation. Voltage threshold settings are typically based on grid operating standards; for example, for the Chinese power grid, the normal operating voltage deviation should generally not exceed ±7% of the rated voltage.

[0048] The voltage deviation is calculated using a sliding time window method, which considers not only the instantaneous deviation but also its duration. Specifically, the system sets a sliding window of 100-200 ms to continuously monitor the voltage deviation within the window. When the deviation exceeds a threshold and the duration exceeds a set value (usually 10 ms), the system determines it as a potential voltage fault. To enhance the reliability of the detection, the system also monitors the voltage change rate. When the voltage changes rapidly (e.g., more than 10% change within 5 ms), a voltage fault alarm will be triggered even if the deviation has not yet reached the duration requirement.

[0049] In addition, the system analyzes the three-phase voltage imbalance, a crucial indicator for diagnosing asymmetrical faults. The three-phase imbalance is calculated based on the ratio of the maximum difference between the three-phase voltages to their average value. When this ratio exceeds 5%, it indicates a potential asymmetrical fault in the system. Finally, the system comprehensively considers voltage deviation, rate of change, and imbalance to output voltage fault detection results, including a fault assessment flag (yes / no), a preliminary fault type determination (symmetrical / asymmetrical), and a fault severity assessment.

[0050] Step S2.2: Take the operating status information of the offshore wind power system as input, compare the deviation between the current of the offshore wind power system and the rated current value, and output the current fault detection result; While performing voltage fault detection, the system also needs to analyze current parameters, which is an important supplement to determining the fault type and severity. Current data is extracted from the operating status information, including the amplitude, phase, and harmonic content of the three-phase current. Similar to voltage analysis, the system first compares the measured current value with the rated current value and calculates the degree of deviation.

[0051] Current fault detection has unique characteristics because different types of faults lead to different current response patterns. For example, short-circuit faults typically cause a sharp increase in current, while some open-circuit faults may cause a decrease in current. Therefore, the system is equipped with bidirectional threshold monitoring, which monitors whether the current exceeds the upper threshold (typically 1.2 times the rated value) or falls below the lower threshold (typically 0.5 times the rated value). When the current exceeds these threshold ranges and the duration exceeds the set value, the system determines that the current is abnormal.

[0052] In addition to amplitude monitoring, the system also analyzes the rate of change of current (di / dt), a key indicator for judging the initial stage of a fault. At the moment a short-circuit fault occurs, the rate of change of current typically shows a significant peak. By monitoring the di / dt value, the system can quickly identify potential problems in the early stages of a fault and trigger protection mechanisms in advance. Simultaneously, the system analyzes the balance of the three-phase currents and calculates the proportion of the negative-sequence current component, serving as another important basis for judging asymmetrical faults.

[0053] Ultimately, the system comprehensively considers current amplitude deviation, rate of change, and phase sequence components to output current fault detection results, including fault judgment indicators, preliminary fault type assessment, and current limit exceedance degree. These results, together with voltage fault detection results, will provide a basis for the final fault determination.

[0054] Step S2.3: Based on the voltage fault detection results and the current fault detection results as input, execute the fault determination matrix algorithm for comprehensive analysis to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, output the fault characteristic parameters, including fault type, fault severity, and fault duration.

[0055] After acquiring the voltage and current fault detection results, the system needs to perform comprehensive analysis to make a final fault determination. This step uses a fault determination matrix algorithm, which takes the fault characteristics of voltage and current as input, and outputs the final fault diagnosis result through preset determination rules.

[0056] The fault determination matrix is ​​a two-dimensional decision table. The horizontal axis represents voltage fault detection results, and the vertical axis represents current fault detection results. Each cell in the matrix defines the fault type corresponding to a specific voltage-current combination. For example, when a significant voltage drop and a large current increase are detected, the system determines it to be a short-circuit fault; when the voltage drops but the current does not change significantly, it may be determined to be a remote fault or a weak connection fault; when both voltage and current are normal but the frequency deviates, it may be determined to be a system frequency disturbance. Through this matrix mapping method, the system can efficiently and accurately determine the fault type.

[0057] Based on the identified fault type, the system further calculates the fault severity. For voltage faults, severity is typically defined by the percentage voltage drop, such as minor faults (10-30% voltage drop), moderate faults (30-60% voltage drop), and severe faults (over 60% voltage drop). For current faults, severity is determined by the multiple by which the current exceeds the rated value. Simultaneously, the system records the fault start time and current time, and calculates the fault duration; these are crucial parameters for assessing fault impact and selecting control strategies.

[0058] Ultimately, the system outputs complete fault characteristic parameters, including fault type (e.g., three-phase short circuit, single-phase ground fault, two-phase short circuit), fault severity (mild, moderate, severe), and fault duration. These parameters serve as key inputs for subsequent constraint modeling and control strategy design, ensuring the system can adopt the optimal control response for specific fault types. It is worth noting that the fault determination process is a continuous update process; the system constantly corrects the fault determination results based on new monitoring data, ensuring the control strategy can adapt to the dynamic changes in faults.

[0059] In this embodiment, the fault feature analysis step is based on a sliding time window detection algorithm. By comparing the deviation of key electrical quantities (such as voltage and current) with preset thresholds, it determines whether a system fault has occurred. When a voltage drop exceeding 10% or a current surge exceeding 1.2 times the rated value is detected, the system triggers a fault detection signal.

[0060] Subsequently, the system employs the symmetrical component method to analyze the fault type. By performing symmetrical component decomposition on the three-phase voltage and current, positive-sequence, negative-sequence, and zero-sequence components are obtained. By analyzing the amplitude and phase relationship of these components, it can be determined whether the fault type is symmetrical or asymmetrical. For example, when a significant negative-sequence component is detected (typically greater than 5% of the base value), it indicates that an asymmetrical fault has occurred in the system; when the three-phase voltages decrease simultaneously and the negative-sequence component is small, it is determined to be a symmetrical fault.

[0061] After determining the fault type, the system further calculates fault characteristic parameters. For voltage faults, the fault depth (i.e., the percentage of voltage drop) and fault duration are calculated; for frequency faults, the frequency deviation and rate of change (ROCOF) are calculated. These characteristic parameters are calculated using the least squares estimation method to improve the accuracy of parameter estimation and the ability to resist interference.

[0062] Based on fault characteristic parameters, the system determines the priority of control objectives. Under symmetrical fault conditions, priority is given to ensuring the system's voltage support capability; under asymmetrical fault conditions, both positive-sequence voltage support and negative-sequence voltage suppression must be considered simultaneously. The system output sequence of control objectives includes priority information, providing a basis for the design of subsequent control strategies.

[0063] Step S2.4: Based on the voltage fault detection results and the current fault detection results, a comprehensive analysis is performed to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault characteristic parameters are obtained, including: The voltage and current of the offshore wind power system are used as inputs to perform positive-sequence component, negative-sequence component and zero-sequence component decomposition, and the component decomposition data is output. Using the component decomposition data as input, the ratio of the negative-order component to the base value is calculated, and asymmetric fault judgment data is output. The fault type is determined by executing a fault type identification algorithm based on the asymmetric fault judgment data as input, the fault feature parameters are updated, and the results are output for subsequent constraint model construction.

[0064] In this embodiment, when the system initially determines that a fault exists through voltage and current deviation detection, it is necessary to further determine the specific type of fault, which is crucial for the selection of subsequent control strategies. The system first performs symmetrical component decomposition on the acquired three-phase voltage and current signals, which is achieved by applying specific mathematical transformations to the three-phase signals. Specifically, the system uses the Fortescue transform to decompose the three-phase signals into positive-sequence, negative-sequence, and zero-sequence components. The positive-sequence component represents the system's operating state under balanced conditions, the negative-sequence component mainly reflects the degree of system imbalance, and the zero-sequence component is closely related to the system's grounding method.

[0065] After component decomposition, the system focuses on analyzing the magnitude of the negative-sequence components. By calculating the ratio of the negative-sequence components to the baseline value (usually the system's rated value), the system can quantify the degree of fault asymmetry. In practical applications, when the ratio of the negative-sequence components to the baseline value exceeds a preset threshold (usually 5%), the system classifies it as an asymmetric fault; when this ratio is small and the positive-sequence components are significantly reduced, it is classified as a symmetric fault. This fault type identification method based on symmetric components has high accuracy and reliability.

[0066] After determining the fault type, the system uses a specially designed fault type identification algorithm to further refine the fault category. Based on pattern recognition principles, this algorithm comprehensively considers features such as the phase relationship between voltage and current, and the amplitude variation trend of each component, enabling it to distinguish specific fault types such as single-phase grounding, two-phase short circuit, two-phase ground fault, and three-phase short circuit. The algorithm's output is the updated fault characteristic parameters, containing detailed information such as fault type, fault severity, and fault phase information. These parameters are directly used in subsequent constraint model construction to ensure that the control strategy can make an optimal response to specific types of faults.

[0067] For offshore wind power systems, accurate fault identification is crucial because different types of faults have significantly different impacts on system stability, requiring different control strategies. For example, in the case of asymmetrical faults, the system needs to consider both positive-sequence voltage support and negative-sequence voltage suppression; while in the case of symmetrical faults, the system should prioritize ensuring the rapid recovery of positive-sequence voltage. Through detailed fault characteristic analysis in this step, the system can provide accurate fault information for subsequent constraint modeling and control strategy design, improving the system's fault response capabilities.

[0068] Step S3, Constraint Modeling: Based on the fault characteristic parameters, establish equipment overcurrent capacity constraints and converter modulation constraints, construct a robust predictive control barrier function for discrete-time offshore wind power system, and generate a constraint model framework.

[0069] Constraint modeling is a crucial step in ensuring the safety and feasibility of control strategies. Under fault conditions, offshore wind power systems need to guarantee equipment safety while providing grid support. This requires control strategies to consider various physical and operational constraints, such as equipment overcurrent capacity constraints and converter modulation constraints. This step involves establishing a comprehensive constraint model based on fault characteristic parameters, providing boundary conditions for subsequent control strategy optimization.

[0070] The first step is to establish constraints on the equipment's overcurrent capacity. The overcurrent capacity of a wind turbine converter is typically 1.2-1.5 times its rated current. During fault periods, a certain degree of overload is permissible for short periods, but it must be strictly controlled within the equipment's tolerance range. The constraint model uses a soft constraint approach, meaning that under normal operating conditions, the current is strictly limited to the rated value, while under fault conditions, short-term overcurrent is allowed but not exceeding the equipment's limits. This soft constraint approach ensures equipment safety while providing sufficient control flexibility.

[0071] Secondly, the converter modulation constraints are established. The modulation ratio of the converter is a key parameter for controlling the output voltage, and under normal circumstances, it should be kept within the linear modulation region. However, under asymmetrical fault conditions, in order to simultaneously provide positive-sequence voltage support and negative-sequence voltage suppression, the converter may need to output a larger voltage vector, causing the modulation ratio to exceed the linear region and enter an overmodulation state. Overmodulation introduces low-order harmonics, affecting power quality. Therefore, the model avoids overmodulation by limiting the modulation ratio within a safe range (e.g., no more than 0.907 in SPWM mode and no more than 1.15 in SVPWM mode).

[0072] After establishing these fundamental constraints, this method innovatively introduces a discrete-time robust predictive control barrier function. The barrier function is a special mathematical structure whose value increases rapidly as the system state approaches the constraint boundary, thus naturally forming a protection mechanism for the constraints during the optimization process. The barrier function used in this method has a logarithmic form, generating a large penalty value at the constraint boundary to ensure that the system state always remains within the safe operating region. Considering the uncertainty of system parameters, the barrier function also incorporates robust design, ensuring the robustness of the control strategy by considering the worst-case scenario of parameter perturbations.

[0073] Finally, all constraints are integrated into a unified constraint model framework. This framework employs convex optimization theory, representing each constraint as the intersection of convex sets, ensuring the solvability of the constraint framework. Simultaneously, by introducing slack variables, potential conflicts between constraints are addressed, improving the system's adaptability. The integrated constraint model framework provides clear boundary conditions and safety guarantees for the design of multi-objective control strategies.

[0074] like Figure 4 As shown, step S3 specifically includes: Step S3.1: Based on the fault characteristic parameters as input, limit the current of the offshore wind power system to no more than 1.2 times the rated current, and output the overcurrent constraint condition; The first step in constraint modeling is to establish a current-capacity constraint model for the equipment. This step takes fault characteristic parameters as input and sets appropriate current limits based on the fault type and severity to ensure that the system provides grid support without compromising equipment safety. For offshore wind power systems, the converter is a critical power electronic interface device, and its current-capacity directly determines the upper limit of the system's support capacity under fault conditions.

[0075] Based on industry standards and equipment specifications, the fundamental constraint set for this system is that the current of the offshore wind power system should not exceed 1.2 times the rated current. This limitation takes into account factors such as the current carrying capacity of semiconductor devices, cooling conditions, and safety margins. At the same time, the system also considers the time characteristics of faults, allowing for higher current peaks to occur within a very short time (usually within 10ms), but such peaks should not exceed 1.5 times the rated current to prevent damage to devices due to heat accumulation.

[0076] The mathematical expression of overcurrent constraints uses inequality constraints, which limit the current amplitude injected into the grid by the converter. For a three-phase system, the constraints are applied to the current components in the abc three-phase system or the dq coordinate system. In particular, under asymmetrical fault conditions, the system needs to inject both positive-sequence and negative-sequence currents simultaneously. In this case, the constraints need to consider the vector superposition of the two sequence current components to ensure that the combined current does not exceed the limit value.

[0077] Ultimately, the system outputs complete overcurrent constraints, including the upper limit of current amplitude, dynamic adjustment mechanism, and emergency handling strategy. These constraints directly influence the design of subsequent control strategies, ensuring that the generated control commands do not cause equipment overcurrent. In practical applications, the system will also dynamically adjust the overcurrent limit based on factors such as equipment temperature and operating time, further improving the reliability of equipment protection.

[0078] Step S3.2: Based on the overcurrent constraint condition as input, limit the converter modulation ratio within the linear modulation region, and output the modulation constraint condition; Based on the established current constraints, the system needs to further consider the modulation constraints of the converter. The modulation ratio is a key parameter in converter control, determining the amplitude and phase of the output voltage. An excessively high modulation ratio will cause the converter to enter the overmodulation region, introducing low-order harmonics and affecting power quality; an excessively low modulation ratio will prevent full utilization of the DC bus voltage, limiting the converter's output capacity. Therefore, setting appropriate modulation constraints is crucial to ensuring system performance.

[0079] There are various modulation strategies for converters, the most common being sinusoidal pulse width modulation (SPWM) and space vector pulse width modulation (SVPWM). For SPWM, the upper limit of the linear modulation region is 0.907 (i.e., 2 / π); for SVPWM, the linear modulation region can be extended to about 1.15. This system sets a corresponding upper limit for the modulation ratio based on the modulation method used to ensure that the converter always operates within the linear modulation region.

[0080] Modulation constraints and current constraints are closely related because the modulation ratio directly affects the converter's output voltage, which in turn determines the injected current through the network impedance. Therefore, the system needs to comprehensively consider both types of constraints to ensure their consistency. Specifically, when the system needs to inject a large current to provide grid support, the converter may need to output a higher voltage, which requires a higher modulation ratio. If the modulation ratio approaches or reaches its upper limit, the system will prioritize ensuring the modulation constraint and adjust the current injection strategy accordingly to avoid entering the overmodulation region.

[0081] Under asymmetrical fault conditions, the converter needs to output both positive-sequence and negative-sequence voltage components simultaneously, which places higher demands on modulation capabilities. The system adopts a sequence-based modulation management strategy, prioritizing positive-sequence voltage support and then providing negative-sequence voltage suppression within the remaining modulation capability range. This priority management ensures that the system meets multiple objective requirements without violating modulation constraints.

[0082] The system outputs complete modulation constraints, including the modulation ratio upper limit, dynamic allocation strategy, and emergency handling mechanism. These constraints, together with the overcurrent constraints, constitute the boundary conditions for the control strategy design.

[0083] Step S3.3: Based on the overcurrent constraint and the modulation constraint as input, execute the control barrier function construction algorithm to output the constraint model framework.

[0084] After establishing overcurrent and modulation constraints, the system needs to integrate these constraints into the control algorithm. Traditional methods typically employ hard constraints, directly adding inequality constraints to the optimization problem. However, this approach can make the optimization problem difficult to solve, especially when the constraints approach critical values. To address this issue, this invention innovatively introduces a control barrier function construction algorithm, transforming hard constraints into flexible constraints, thereby improving the stability and efficiency of the optimization solution.

[0085] A control barrier function is a special mathematical structure that rapidly increases in value as the system state approaches the constraint boundary, generating a "push" that drives the state away from the boundary. This system uses a logarithmic barrier function, whose mathematical expression is:

[0086] Where x represents the system state variable, g(x) represents the constraint function (such as current or modulation ratio constraint), ε is the constraint limit, and α is the weighting coefficient. When g(x) approaches ε, the barrier function value tends to infinity, effectively preventing the system from violating the constraints.

[0087] The construction of the barrier function takes into account the uncertainties of the system. In practical applications, system parameters (such as network impedance) may contain errors, which can affect the accuracy of the constraints. To enhance the robustness of the control, the system adopts robust predictive control theory, considering the worst-case scenario of parameter disturbances to ensure that the constraints are satisfied under all possible operating conditions. Specifically, the system establishes a set of parameter uncertainties, calculates the maximum value that the constraint function can reach within this set, and constructs the barrier function based on this value, thereby achieving robust protection against uncertainties.

[0088] For discrete-time systems, the barrier function also needs to consider the influence of the sampling period. Since control decisions are made at discrete time points, while the system state changes continuously between sampling points, there may be situations where constraints are satisfied at some sampling point but violated between sampling points. To address this issue, the system introduces discrete-time robust predictive control theory, which, by considering the state evolution between sampling points, ensures that constraints are satisfied throughout the entire control period.

[0089] The system transforms overcurrent constraints and modulation constraints into corresponding barrier functions and integrates them into a unified constraint model framework. This framework employs a convex optimization form to ensure the solvability of the optimization problem. Simultaneously, by introducing slack variables, potential conflicts between constraints are addressed, improving the system's adaptability. This constraint model framework serves as the foundation for subsequent control strategy design, ensuring that the generated control commands meet both the grid support requirements and guarantee the safe and stable operation of equipment.

[0090] This constraint handling method based on barrier functions offers better numerical stability and solution efficiency compared to traditional hard constraint methods, making it particularly suitable for handling complex multi-objective control problems. Furthermore, its robust design approach, which considers uncertainties, enhances the reliability and adaptability of the control strategy in practical applications.

[0091] In this embodiment, the constraint modeling first establishes a current-capacity constraint model for the equipment. Considering that the current-capacity of a wind power converter is typically 1.2-1.5 times its rated current, the model uses a soft constraint approach to describe the current limit. That is, under normal operating conditions, the current is strictly limited to not exceeding the rated value, while under fault conditions, short-term overcurrent is allowed but not exceeding the equipment limit. The model is described using a quadratic programming form, which ensures the feasibility of the constraints and avoids oscillations in the control system.

[0092] The converter modulation constraint model primarily considers overmodulation issues. Under asymmetrical fault conditions, the converter may enter the overmodulation region in order to simultaneously inject positive and negative sequence currents. The model introduces a modulation ratio constraint to ensure that the output voltage vector always remains within the linear modulation region of the converter. Specifically, the model limits the modulation ratio to no more than 0.907 (for SPWM modulation) or no more than 1.15 (for SVPWM modulation).

[0093] The design of the robust predictive control barrier function for the discrete-time system is the core of this step. The barrier function adopts a logarithmic form, generating a large penalty value at the constraint boundaries to ensure that the system state always remains within the safe operating region. Considering the uncertainty of system parameters, the barrier function also incorporates robustness design, ensuring the robustness of the control strategy by considering the worst-case scenario of parameter disturbances.

[0094] Finally, the system integrates the various constraint models into a unified constraint framework. Using convex optimization theory, each constraint condition is represented as the intersection of convex sets, ensuring the solvability of the constraint framework. Simultaneously, by introducing slack variables, potential conflicts between constraints are addressed, improving the system's adaptability. The integrated constraint framework serves as a crucial input for subsequent multi-objective control strategy design.

[0095] Step S4, Multi-objective control strategy design: Based on the constraint model framework, construct a wind turbine dynamic model that considers mechanical-electrical coupling characteristics, and use the wind turbine dynamic model to execute an iterative learning predictive control algorithm to generate a control command sequence.

[0096] The core of this invention is the design of a multi-objective control strategy. Its goal is to design an optimal control strategy that coordinates and achieves multiple control objectives while satisfying constraints. Offshore wind power systems, when facing faults, need to simultaneously consider multiple objectives such as voltage support, frequency support, and mechanical load control, which may conflict with each other. This step achieves coordinated optimization of multiple objectives by establishing an accurate dynamic model of the wind turbine and combining iterative learning predictive control algorithms.

[0097] Constructing a dynamic model of the wind turbine is fundamental to achieving precise control. Unlike traditional models, the dynamic model proposed in this invention comprehensively considers the mechanical and electrical characteristics of the wind turbine and their coupling relationships. The model adopts a multi-mass block structure, treating the rotor, drive shaft, and generator rotor as independent blocks of rotational inertia, connected by a spring-damping system. This structure accurately describes the elastic and damping characteristics of the transmission system and reflects torsional vibration phenomena in the transmission chain. Simultaneously, the model also considers the electromagnetic characteristics of the generator, establishing coupling equations between the mechanical and electrical systems, enabling it to describe the impact of electrical disturbances on the mechanical system and the feedback of changes in the mechanical system to the electrical output.

[0098] Considering the unique environment of offshore wind power, the model also incorporates platform vibration and hydrodynamic factors. Offshore wind turbines are typically installed on floating or fixed platforms, and platform vibration affects turbine stability. The model establishes the coupling relationship between platform vibration and turbine motion by introducing platform dynamics equations. Furthermore, taking into account the impact of waves and currents on the platform, the model includes a hydrodynamic module capable of simulating external load variations under different sea conditions. These considerations make the model more consistent with the actual operating conditions of offshore wind power, improving the targeting and effectiveness of the control strategy.

[0099] Based on a complete dynamic model of the wind turbine, this method designs an iterative learning predictive control algorithm. This algorithm combines the predictive power of model predictive control with the adaptability of iterative learning control, enabling continuous optimization of control parameters based on historical operating data. The core of the algorithm consists of three parts: a predictive model, a cost function, and an optimization solver. The predictive model, based on state-space equations, can predict the dynamic response of the system within a future time window. The cost function comprehensively considers multiple control objectives such as voltage support, frequency support, power stability, and mechanical stress, and adopts a weighted sum form, with the weight coefficients dynamically adjusted according to the priority of the control objectives. The optimization solver employs an improved interior-point method algorithm combined with a warm-start strategy to ensure the algorithm's real-time performance and convergence.

[0100] By executing an iterative learning predictive control algorithm, the system can generate an optimal sequence of control commands. These commands include key parameters such as the converter's modulation ratio and phase angle, which directly determine the wind turbine's response characteristics under fault conditions. The generation of the command sequence considers multiple time steps, ensuring that the control strategy not only focuses on the current state but also takes into account possible future system changes, thus improving the control's foresight and stability.

[0101] like Figure 5 As shown, step S4 specifically includes: Step S4.1: Using the constraint model framework as input, construct the wind turbine, drive shaft, and generator rotor into a multi-mass block structure model, and output the inertia distribution data; The primary task in designing a multi-objective control strategy is to establish an accurate dynamic model of the wind turbine, and a multi-mass block structural model is an effective method for describing the dynamic characteristics of the wind turbine's mechanical system. In this step, the mechanical parts of the wind turbine are decomposed into several rigid blocks with concentrated inertia, mainly including the rotor, drive shaft, and generator rotor. This decomposition accurately reflects the dynamic characteristics of each component in the wind turbine's drive system, providing a foundation for subsequently establishing a complete mechanical-electrical coupling model.

[0102] The wind turbine rotor is the core component of a wind turbine, directly receiving wind energy and converting it into mechanical energy. In modeling, the rotor is considered a mass block with a large moment of inertia. The system calculates the rotor's moment of inertia based on its geometric parameters (such as rotor diameter, number and shape of blades) and material properties. Offshore wind turbine rotors typically have diameters ranging from 150 to 220 meters, three blades, and are primarily made of glass fiber reinforced composites or carbon fiber composites. These parameters collectively determine the rotor's moment of inertia, typically on the order of 10⁷ to 10⁸ kg·m², making it the largest mass block in the entire system.

[0103] The drive shaft is a key component connecting the wind turbine and the generator. While it may be simplified in direct-drive wind turbines, it includes a low-speed shaft, a gearbox, and a high-speed shaft in geared turbines. During modeling, the system selects an appropriate topology based on the transmission structure type. For geared turbines, the drive shaft is divided into two inertia blocks: a low-speed end and a high-speed end, connected by the gearbox. The system calculates the rotational inertia of these components, considering factors such as material properties, geometry, and transmission ratio. Typically, the inertia of the low-speed shaft is on the order of 10⁴–10⁵ kg·m², and that of the high-speed shaft is on the order of 10²–10³ kg·m².

[0104] The generator rotor is the last major inertia block in the mechanical system and is directly coupled to the electrical system. The rotational inertia of the rotor is calculated based on the generator type (e.g., permanent magnet synchronous generator, doubly-fed induction generator) and capacity. For large offshore wind turbines, the generator rotor inertia is typically on the order of 10³-10⁴ kg·m². The inertial characteristics of the generator rotor directly affect the mechanical response under electrical disturbances and are a key part of the mechanical-electrical coupling analysis.

[0105] After modeling each inertia block, the system outputs complete inertia distribution data, including the inertia value, speed ratio, and topological connection relationship of each component. This data will serve as input for the next step of establishing the spring-damped connection relationship, together forming the dynamic model of the wind turbine mechanical system. Compared with the single-mass block model, the multi-mass block model can more accurately reflect the torsional vibration phenomenon and load distribution in the transmission system, providing a precise theoretical basis for the dynamic response analysis of the wind turbine under fault conditions.

[0106] Step S4.2: Based on the inertia distribution data as input, execute the spring damping relationship construction algorithm to establish the spring damping connection relationship and output the mechanical dynamic characteristic data; After determining the inertia distribution of each mass block, the next step is to establish the connection relationships between these mass blocks to accurately describe the dynamic characteristics of the system. In the wind turbine drive system, the connections between components are not completely rigid, but possess certain elastic and damping characteristics. These characteristics have a significant impact on the dynamic response of the system, especially under disturbances such as grid faults. Therefore, this step establishes an elastic and damping connection model between the mass blocks by executing a spring-damping relationship construction algorithm.

[0107] The establishment of spring-damping relationships first requires determining the system's topology. For typical wind turbines, the topology is usually in series, with the rotor, drive shaft, and generator rotor connected sequentially. In systems with gearboxes, the structure is more complex, requiring consideration of the gearbox's transmission ratio and efficiency. Based on the wind turbine type and transmission method, a corresponding topology connection model is established to clarify the connection paths and transmission relationships between each mass block.

[0108] The elastic elements connecting the various mass blocks are characterized by stiffness coefficients (or spring constants), measured in N·m / rad, which describe the magnitude of the torque generated per unit torsional angle. The system calculates the stiffness coefficients of each connection point based on the material properties (such as elastic modulus), geometric parameters (such as shaft diameter and length), and structural characteristics of the drive shaft. For large offshore wind turbines, the stiffness of the low-speed shaft is typically in the range of 10⁸–10⁹ N·m / rad, while the stiffness of the high-speed shaft is in the range of 10⁶–10⁷ N·m / rad. Furthermore, the system also considers the gear meshing stiffness of the gearbox, which significantly affects the resonance characteristics of the transmission system.

[0109] Damping elements are characterized by their damping coefficients, measured in N·m·s / rad, which describe the magnitude of the damping torque generated per unit angular velocity. Damping originates from various factors, including the internal damping of materials, bearing friction, and gear meshing losses. The system estimates the damping coefficients of each connection point using a combination of theoretical analysis and empirical data. Since damping coefficients are typically difficult to measure directly, the system employs modal analysis to inversely deduce damping parameters based on the system's vibration characteristics and attenuation rate. For large wind turbines, the damping ratio of the transmission system is usually in the range of 0.01–0.1.

[0110] When establishing the spring-damping relationship, the system also considers nonlinear factors, such as nonlinear variations in clearance, friction, and stiffness. These nonlinear factors may cause the system to exhibit different dynamic characteristics under different operating conditions. The system incorporates these nonlinear effects into the model through methods such as piecewise linear approximation or polynomial fitting, thereby improving the model's accuracy and applicability.

[0111] After establishing the spring-damping relationship, the system outputs mechanical dynamic characteristic data, including the stiffness coefficient, damping coefficient, transmission ratio, and nonlinear characteristic parameters of each connection. This data comprehensively describes the dynamic characteristics of the wind turbine's mechanical system and, together with the electromagnetic characteristic parameters, will be used to construct a complete dynamic model of the wind turbine. An accurate mechanical dynamic characteristic model is crucial for analyzing the impact of grid faults on the wind turbine's mechanical system and for optimizing control strategies to mitigate mechanical stress.

[0112] Step S4.3: Based on the mechanical dynamic characteristic data and electromagnetic characteristic parameters as input, execute the coupled model construction algorithm to output the dynamic response characteristics of the wind turbine dynamic model; After completing the mechanical system modeling, the next step is to combine the mechanical system with the electrical system to establish a complete mechanical-electrical coupling model. This step is one of the core innovations of this invention because traditional control methods often consider the mechanical and electrical systems separately, ignoring their mutual influence, leading to unsatisfactory control results. This step, by executing a coupling model construction algorithm, achieves the organic integration of mechanical and electrical characteristics, improving the accuracy of the model and the effectiveness of the control strategy.

[0113] The construction of the coupled model first requires electromagnetic characteristic parameters. For a typical wind turbine, these parameters mainly include the generator's inductance, resistance, flux linkage, and electromagnetic torque constant. The system selects a suitable electromagnetic model based on the generator type (e.g., permanent magnet synchronous generator PMSG, doubly-fed induction generator DFIG). For PMSG, the main parameters include stator resistance, d-axis and q-axis inductance, and permanent magnet flux linkage; for DFIG, they include stator and rotor resistance, mutual inductance, and self-inductance. These parameters collectively determine the generator's electromagnetic characteristics and dynamic response.

[0114] The core of the coupled model is establishing the interaction between the mechanical and electrical systems. In wind power systems, this interaction is mainly achieved through electromagnetic torque and mechanical speed. Electromagnetic torque is the key variable connecting the two systems; it is generated by the electrical system and acts on the mechanical system. Conversely, mechanical speed is determined by the mechanical system and, in turn, acts on the electrical system, affecting the generation of electromagnetic torque. By establishing the electromagnetic torque equation and the mechanical motion equation, the system forms a closed-loop coupled model.

[0115] For a permanent magnet synchronous generator, the electromagnetic torque equation is:

[0116] Where p is the number of pole pairs, Ld and Lq are the d-axis and q-axis inductances, id and iq are the d-axis and q-axis currents, and ψf is the permanent magnet flux linkage.

[0117] The mechanical equations of motion describe the changes in the motion state of each mass block under the action of electromagnetic torque and wind turbine torque:

[0118] Where J is the system's equivalent inertia, ω is the rotational speed, Tm is the wind turbine torque, Te is the electromagnetic torque, and B is the damping coefficient.

[0119] In practical modeling, the system considers more complex factors, such as the dynamic characteristics of the converter, the influence of grid impedance, and the time delay of the control system. These factors are incorporated into the model through additional state variables and equations, forming a more complete system description. Especially for offshore wind power systems, the unique influences of the marine environment, such as platform vibration and wave loads, must also be considered. These factors are added to the model through additional disturbance terms, improving the model's adaptability to actual operating conditions.

[0120] After the coupled model was constructed, dynamic response characteristic analysis was performed on the system, including frequency response analysis, small-signal stability analysis, and time-domain transient simulation. These analysis results formed the dynamic response characteristic data of the wind turbine's dynamic model, including the system's natural frequency, damping ratio, gain margin, phase margin, and time constant. This characteristic data comprehensively describes the dynamic behavior of the wind turbine under different disturbances, providing a theoretical basis for subsequent control strategy design.

[0121] An accurate mechanical-electrical coupling model is key to achieving effective wind power transient control. By considering the interaction between the mechanical and electrical systems, the system can more accurately predict the impact of grid faults on wind turbines and design optimal control strategies that provide grid support while protecting the mechanical safety of the wind turbines.

[0122] Step S4.4: Using the dynamic response characteristics as input, predict the dynamic response of the offshore wind power system based on the state-space equation, and output the predicted response data; After establishing a complete dynamic model of the wind turbine, the next step is to apply the model to a predictive control algorithm to predict the system's future dynamic response. This step is fundamental to model predictive control; by accurately predicting the system's future behavior, control decisions can be made in advance, resulting in better control performance. This step uses the dynamic response characteristics as input and predicts the system's dynamic response within a future time window using state-space equations, providing a basis for subsequent control optimization.

[0123] State-space equations are the standard form for describing dynamic systems, consisting of two parts: state equations and output equations.

[0124] Where x is the state variable vector, u is the control input vector, y is the system output vector, and A, B, C, and D are the system matrices.

[0125] For offshore wind power systems, state variables typically include the angles and angular velocities of the mechanical system, the currents and voltages of the electrical system, and the DC bus voltage of the converter. Control inputs include control signals such as the modulation ratio and phase angle of the converter. System outputs include key indicators such as the active and reactive power injected into the grid, the grid voltage support, and the frequency response.

[0126] The prediction process first requires determining the current state of the system. The system obtains the current state variable values ​​through a state observer or direct measurement. For state variables that cannot be directly measured, the system uses techniques such as Kalman filtering or Luenberger observers for state estimation to ensure an accurate initial state is obtained.

[0127] After obtaining the current state, the system predicts the system state evolution over multiple future time steps based on the state-space equations. The prediction time window is typically 10-20 control cycles, covering the main dynamic response process of the system. Considering the nonlinear characteristics of the system, the prediction process may require linearization or the use of nonlinear prediction algorithms. The system considers various constraints at each prediction step to ensure that the predicted state does not violate safety restrictions such as overcurrent constraints and modulation constraints.

[0128] The system prediction considers not only deterministic inputs but also potential disturbances and uncertainties. For offshore wind power systems, the main sources of disturbances include wind speed variations, wave loads, and grid voltage fluctuations. The system employs a robust prediction method that considers the worst-case effects of disturbances, ensuring sufficient accuracy and reliability of the prediction results under various disturbance conditions.

[0129] The predicted results include the complete trajectories of state and output variables within the prediction time window. Particular attention is paid to the system's response characteristics to grid faults, including voltage recovery processes, changes in active and reactive power, frequency support effects, and mechanical stress levels. This predicted data will serve as the basis for subsequent control optimization, helping the system select the optimal control strategy.

[0130] Accurate system response prediction is a prerequisite for effective control. Through prediction based on detailed models, the system can "see" the response effects under different control strategies in advance, thereby making more informed control decisions, avoiding problems that may arise from blind control, and improving the system's support capability and safety under fault conditions.

[0131] Step S4.5: Calculate the deviation between the predicted response data and the expected response as input, execute the parameter optimization algorithm, and output the updated control parameter values; After obtaining the predicted response data, the system needs to evaluate the differences between these responses and the desired responses, and optimize the control parameters based on these differences. The goal of this step is to find the set of control parameters that makes the system response closest to the desired objective, achieving multi-objective optimal control. This step calculates the deviation between the predicted and desired responses, executes the parameter optimization algorithm, and generates updated values ​​for the control parameters.

[0132] First, the system needs to define the expected response. Under power grid fault conditions, the expected response mainly includes the following aspects: 1) Voltage support target: During a fault, the system should provide appropriate reactive power to support the grid voltage and restore it to a safe range as soon as possible.

[0133] 2) Frequency support target: The system should provide active power support under frequency disturbance conditions, suppress frequency deviation, and improve system inertia and damping.

[0134] 3) Negative sequence suppression target: Under asymmetrical fault conditions, the system should provide negative sequence current injection to suppress negative sequence voltage components and improve the three-phase balance of the power grid.

[0135] 4) Mechanical protection objective: The control strategy should avoid causing excessive loads and stresses on the wind turbine mechanical system to protect equipment safety.

[0136] The system defines corresponding evaluation indicators and weights for each objective, constructing a comprehensive objective function. The objective function typically takes the form of a weighted squared error, which is the sum of the squared deviations of each objective multiplied by its weight coefficients. The weight coefficients are dynamically adjusted based on the priority of the control objectives. For example, in the event of a severe fault, the weight of the voltage support objective may increase; when mechanical stress approaches its limit, the weight of the mechanical protection objective will increase accordingly.

[0137] After determining the objective function, the system executes a parameter optimization algorithm to find the control parameters that minimize the objective function. The optimized control parameters mainly include the converter's control gain (such as the proportional and integral coefficients of the PI controller), virtual impedance parameters, and current distribution coefficients. The optimization algorithm employs a modified gradient descent method, determining the direction and magnitude of parameter adjustment by calculating the partial derivatives of the objective function with respect to each control parameter.

[0138] The parameter optimization process considers several constraints, including parameter change rate limits, parameter range limits, and stability constraints. Parameter change rate limits ensure the smoothness of parameter updates and avoid system instability caused by abrupt changes; parameter range limits ensure that parameters remain within an effective engineering range; and stability constraints, through small-signal stability analysis, ensure that the optimized parameters maintain system stability.

[0139] For complex nonlinear systems, optimization algorithms may encounter the problem of finding local optima. To overcome this difficulty, the system employs a multi-starting-point search strategy, optimizing from different initial points and comparing the results obtained from different starting points, selecting the globally optimal or a relatively good local optimum as the final result.

[0140] Finally, the system outputs optimized control parameter updates, including gain adjustments for each controller, updated virtual impedance values, and adjustments to the current distribution strategy. These updated values ​​will be used in the next step for iterative optimization of the control parameters, ultimately generating the optimal control command sequence.

[0141] Step S4.6: Optimize the control parameters using the updated control parameter values ​​as input by executing an iterative optimization algorithm, and output the control command sequence.

[0142] After obtaining the updated values ​​of the control parameters, the system executes an iterative optimization algorithm to continuously optimize the control parameters until a satisfactory control effect is achieved. Iterative optimization is a key step in improving control performance. Through multiple trials and adjustments, the system can overcome the influence of model errors and achieve more precise control. This step executes the iterative optimization algorithm based on the updated control parameter values, and finally outputs an optimized control command sequence.

[0143] The core of iterative optimization is the learning mechanism. The system learns from historical control processes, adjusts its control strategy, and gradually improves control accuracy. Specifically, the system records the difference between the actual response and the target response after each control step, analyzes the patterns and rules of these differences, and identifies the mapping relationship between control parameters and system response. In this way, the system can find control parameters that are closer to the optimal value after multiple iterations.

[0144] The iterative optimization algorithm employs an adaptive learning rate strategy. In the initial optimization phase, the system uses a large learning rate to quickly approach the optimal region; as the number of iterations increases, the learning rate gradually decreases to achieve fine-tuning. Simultaneously, the algorithm introduces a momentum factor, utilizing historical update information to accelerate the convergence process and avoid getting trapped in local optima.

[0145] To improve the robustness of the optimization, the system employs a constrained Lyapunov optimization method. This method ensures the stability and convergence of the system during the optimization process by constructing a Lyapunov function. Simultaneously, the constraints ensure that the optimized parameters still meet the system's safety and stability requirements. This method is particularly suitable for handling complex control problems with multiple constraints, enabling performance optimization while guaranteeing system safety.

[0146] During the iterative optimization process, the system also considers the coordination of control objectives at different time scales. The electrical and mechanical systems have different time constants; the electrical system responds quickly (milliseconds), while the mechanical system responds more slowly (seconds). The system employs a hierarchical optimization strategy, optimizing electrical control parameters at the fast time scale and coordinating mechanical protection strategies at the slow time scale, thus achieving coordinated control across multiple time scales.

[0147] The termination conditions for iterative optimization include the objective function value falling below a preset threshold, parameter changes being less than a specified precision, or reaching the maximum number of iterations. When one of these conditions is met, the system outputs the final optimized control command sequence. These commands include complete control signals within the prediction time window, such as the converter's modulation ratio, phase angle, and reference values ​​for the current and voltage loops.

[0148] The control command sequence also includes emergency response strategies to handle unforeseen anomalies. The system is pre-designed with various contingency measures, such as rapid load reduction, emergency current limiting, and fault ride-through mode switching, to cope with possible extreme situations and ensure the system operates safely and reliably under various conditions.

[0149] Through iterative optimization algorithms, the system can comprehensively consider grid support requirements and equipment safety requirements to generate the optimal control strategy, achieving optimized transient control of offshore wind power systems under fault conditions. This learning-based optimization method overcomes the influence of model uncertainty and improves the adaptability and robustness of control, which is one of the core innovations of this invention.

[0150] In this embodiment, constructing a dynamic model of the wind turbine is fundamental to achieving precise control. This model comprehensively considers the mechanical and electrical characteristics of the wind turbine, including the aerodynamic characteristics of the rotor, the dynamic characteristics of the transmission system, the electromagnetic characteristics of the generator, and electromechanical coupling effects. The model adopts a multi-mass block structure, treating the rotor, drive shaft, and generator rotor as independent blocks of rotational inertia, connected by a spring-damping system. Furthermore, considering the characteristics of offshore wind power, the model also incorporates platform vibration and hydrodynamic effects, making it more consistent with actual operating conditions.

[0151] Based on a dynamic model, an iterative learning predictive control algorithm framework is designed. This framework comprises three core components: a predictive model, a cost function, and an optimization solver. The predictive model, based on state-space equations, predicts the dynamic response of the system within a future time window. By introducing an iterative learning mechanism, the system can learn from historical operating data and optimize control parameters, thereby improving control accuracy. Specifically, at the end of each control cycle, the system updates the control parameters based on the deviation between the actual and expected responses, achieving a gradual improvement in control performance.

[0152] The establishment of a multi-objective optimization problem requires comprehensive consideration of multiple control objectives, such as voltage support and frequency support. The objective function adopts a weighted sum form, with the weight coefficients dynamically adjusted according to the priority of the control objectives. The voltage support objective mainly focuses on the rapid recovery and stability of the voltage amplitude; the frequency support objective considers the dynamic response and steady-state deviation of the system frequency. In addition, practical engineering constraints such as power fluctuation constraints and mechanical stress constraints also need to be considered. The mathematical description of the optimization problem adopts a quadratic programming form to ensure the real-time performance and reliability of the solution.

[0153] Control command sequences are generated by solving an optimization problem. The solution process employs an improved interior-point method algorithm, which features fast convergence and good numerical stability. To improve real-time performance, a warm-start strategy is used, utilizing the solution from the previous time step as the initial value for the current time step. The generated control commands include key parameters such as the converter's modulation ratio and phase angle, which directly determine the transient response characteristics of the wind turbine.

[0154] Step S5, Transient control execution: Configure control parameters according to the control instruction sequence, execute the transient control strategy, and dynamically adjust the control parameters based on real-time monitoring results.

[0155] Transient control execution is the step of putting the generated control strategy into practice. Its goal is to ensure that control commands are executed accurately and promptly, and to make dynamic adjustments based on actual results. This step includes three key processes: control parameter configuration, control strategy execution, and dynamic adjustment, which together ensure the stable operation of the offshore wind power system under fault conditions and its grid support capability.

[0156] Control parameter configuration is the first step in executing the control strategy. Based on the received control command sequence, the system configures the converter's control parameters, including PI parameters for the current loop and voltage loop, and virtual impedance parameters. To ensure smooth parameter changes, a piecewise linear interpolation method is used during configuration to avoid system oscillations caused by abrupt parameter changes. Simultaneously, the system performs parameter validity checks to ensure that the configured parameters are within the equipment's allowable range, preventing equipment damage or control failure due to unreasonable parameters.

[0157] The transient control strategy employs a hierarchical control structure. The outer layer control is responsible for power allocation and operating mode switching, determining the priority and allocation ratio of voltage and frequency support based on the fault type and system state. The inner layer control implements specific voltage and current regulation, directly driving the converter's switching signal generation. During a fault, the system automatically switches control modes according to the fault type. For example, in symmetrical fault conditions, positive-sequence voltage support control is used; in asymmetrical fault conditions, both positive-sequence voltage support and negative-sequence voltage suppression control are activated simultaneously. The execution of the control strategy requires high-speed processing capabilities, typically with a control cycle of 0.1 ms (10 kHz frequency) to ensure rapid response to grid faults.

[0158] Real-time monitoring and dynamic adjustment are crucial for ensuring effective control. The system uses a high-speed data acquisition device to monitor the control execution in real time, including electrical quantities (voltage, current, power) and mechanical quantities (speed, torque). The data acquisition frequency is 20kHz, higher than the control execution frequency, ensuring that detailed changes in the control process are captured. Based on the monitoring results, the system employs an adaptive control mechanism to dynamically adjust the control parameters. The adjustment process is based on fuzzy control rules, determining the parameter adjustment amount according to the magnitude and trend of the deviation. For example, when a slow voltage recovery rate is detected, the system increases the proportional gain of the voltage control loop; when excessive mechanical stress is detected, the response speed of the power support is reduced. This adaptive mechanism ensures that the control system maintains good performance under different operating conditions.

[0159] Step S5 specifically includes: Step S5.1: Receive the control command sequence as input, configure the current loop and voltage loop parameters of the converter, and output the initial control parameters; The first step in transient control execution is configuring controller parameters based on the optimized control commands. This step transforms the theoretical control strategy into practically executable control commands, serving as a crucial link between theory and practice. This step receives the control command sequence, configures the converter's key control parameters, and ensures the accurate implementation of the control strategy.

[0160] Converter control systems typically employ a hierarchical structure, including outer-loop control (such as power control and voltage control) and inner-loop control (current control). The current loop is the most basic control level, directly controlling the current waveform output by the converter; the voltage loop serves as the outer loop, generating reference values ​​for the current loop. The parameter settings of these two control loops have a decisive impact on the dynamic response characteristics of the system.

[0161] Current loop control typically employs a PI (Proportional-Integral) controller, whose parameters include a proportional coefficient Kp and an integral coefficient Ki. The proportional coefficient determines the control response speed, while the integral coefficient affects the ability to eliminate steady-state errors. Under fault conditions, the current loop response speed needs to be particularly fast to promptly track changes in the reference value and provide rapid grid support. The system configures appropriate Kp and Ki values ​​based on the control command sequence; typically, the Kp value increases during faults to improve response speed.

[0162] The voltage loop control also employs a PI controller structure, but its dynamic characteristics differ from those of the current loop. The voltage loop typically has a slower response speed to avoid dynamic interference with the current loop and ensure system stability. The system sets the Kp and Ki values ​​of the voltage loop according to the control command sequence and configures the voltage reference value generation strategy. Under asymmetrical fault conditions, the voltage reference value needs to include both positive and negative sequence components to suppress negative sequence voltage.

[0163] In addition to the basic PI parameters, the system also needs to be configured with other key control parameters, such as virtual impedance parameters, current limiting parameters, and reference calculation module parameters. Virtual impedance is used to simulate the characteristics of a synchronous generator, improving the system's stability and grid-connected adaptability; the current limiting parameter ensures that the current does not exceed the safety limit; and the reference calculation module dynamically generates reference values ​​for voltage and current based on the grid status and control objectives.

[0164] During parameter configuration, the system also considers switching strategies between different control modes. Depending on the fault type and severity, the system may need to switch between different control modes, such as switching from normal operation mode to voltage support mode or frequency support mode. The configuration includes the switching trigger conditions, the smooth transition strategy for switching, and the preset parameter values ​​for each mode.

[0165] After parameter configuration is complete, the system outputs initial control parameters, including parameter settings for all controllers, amplitude limits, and mode selection signals. These parameters will serve as inputs for subsequent smoothing and parameter verification, ultimately forming executable control commands.

[0166] Step S5.2: Execute a piecewise linear interpolation algorithm on the initial control parameters as input, and output the smoothed control parameters. After obtaining the initial control parameters, the system needs to undergo smoothing to ensure the continuity of parameter changes and avoid system oscillations caused by abrupt parameter changes. This step executes a piecewise linear interpolation algorithm to connect discrete parameter points into a continuous parameter curve, achieving a smooth transition of parameters.

[0167] The core idea of ​​piecewise linear interpolation is to establish a linear transition relationship between parameter change points. For example, if the proportional coefficient Kp of the controller needs to change from the initial value K1 to the target value K2, the system will not change the parameter instantaneously, but will transition linearly over a specified time interval, i.e.:

[0168] Here, t1 and t2 are the start and end times of the transition.

[0169] Parameter smoothing requires consideration of several factors. First is the transition time setting. Too short a transition time may cause parameters to change too quickly, leading to system oscillations; too long a transition time will delay system response and reduce control effectiveness. The system dynamically adjusts the transition time based on the nature and magnitude of parameter changes. Generally, the greater the parameter change, the longer the transition time should be to ensure system stability.

[0170] Secondly, there is the issue of coordination among multiple parameters. Control systems contain multiple interrelated parameters, and the changes in these parameters need to maintain a certain degree of coordination to preserve system stability and consistency. The system employs a group smoothing strategy, organizing related parameters into parameter groups to ensure that the changes in parameters within each group maintain consistent time characteristics, thus avoiding control anomalies caused by asynchronous parameter changes.

[0171] For special control parameters, such as mode switching signals, the system employs a more complex smoothing strategy. Mode switching typically involves discrete state changes, and direct switching can lead to control discontinuity. The system uses a soft-switching mechanism, mixing the control outputs of two modes during the switching process, with the proportional coefficient changing smoothly over time to achieve a smooth transition between modes.

[0172] Parameter smoothing also considers the dynamic characteristics of the system. Different rate-of-change limits are set for different parameters to ensure that the rate of change is within an acceptable range for the system. For example, the rate of change of current loop parameters can be relatively high because the current loop has a fast response speed; while the rate of change of voltage loop parameters needs to be lower to maintain the stability of the outer loop control.

[0173] After smoothing, the system outputs the smoothed control parameters, including the continuous parameter trajectory over the entire control time window. These parameters will serve as input for the next step of parameter verification, ensuring they remain within valid ranges.

[0174] Parameter smoothing is a crucial aspect of ensuring the stable operation of a control system. By employing appropriate interpolation algorithms, the system can maintain the target parameters while avoiding oscillations and instabilities caused by abrupt parameter changes, thereby improving the stability and reliability of the control.

[0175] Step S5.3: Using the smoothed control parameters as input, execute the parameter verification algorithm to check whether they are within the rated parameter range, and output valid control parameters; After parameter smoothing, the system needs to perform parameter verification to ensure that all parameters are within a safe and effective range. This step executes a parameter verification algorithm to check whether the smoothed control parameters meet the equipment's operating requirements, adjusts parameters that are out of range, and outputs the final effective control parameters.

[0176] The first step in parameter verification is range checking. The system sets a safe operating range for each control parameter, including minimum and maximum values. These ranges are determined based on the physical characteristics of the equipment, the stability requirements of control theory, and engineering experience. For example, for the proportional gain Kp of a PI controller, the minimum value is typically set as the lower limit to ensure basic control performance, while the maximum value considers the upper limit of system stability. For virtual impedance parameters, the range setting takes into account the impact of impedance magnitude on system damping and resonance. The system checks whether each smoothed parameter is within the specified range; if it exceeds the range, the parameter is trimmed to the range boundary.

[0177] The second step in parameter verification is consistency checking. Multiple parameters in a control system are interdependent, and these relationships need to be kept consistent to ensure normal system operation. For example, the bandwidths of the current loop and voltage loop need to maintain a certain proportional relationship; typically, the current loop bandwidth should be 3-5 times higher than the voltage loop bandwidth. Parameters for different phase sequence controls also need to meet specific proportional relationships to ensure control balance. The system checks the relationships between parameters through a series of consistency rules. If inconsistencies are found, the relevant parameters are adjusted according to priority to maintain overall system consistency.

[0178] Parameter verification also includes stability checks. The system employs small-signal stability analysis to assess the system's stability margin under the current parameter settings. Specifically, the system calculates the eigenvalues ​​or Bode plot of the closed-loop control system and checks whether the gain margin and phase margin meet the requirements. If insufficient stability margin is found, the system will automatically adjust relevant parameters to increase system stability, typically by reducing the control gain or increasing the damping factor.

[0179] For certain key parameters, the system also undergoes functional testing. A simplified system model is used to quickly simulate the system's response characteristics under the current parameter settings, checking for issues such as overshoot, oscillation, or excessive steady-state error. If any functional abnormalities are detected, the system will make targeted adjustments to the parameters to ensure the normal operation of basic control functions.

[0180] After completing all verifications, the system outputs the final valid control parameters. These parameters meet the range, consistency, and stability requirements and can be safely and effectively used for practical control. Simultaneously, the system records adjustments made during the parameter verification process as reference information for subsequent optimization.

[0181] Parameter verification is a crucial guarantee for the safe operation of control systems. Through a rigorous verification process, the system can prevent control anomalies and equipment damage caused by unreasonable parameter settings, thereby improving the reliability and safety of system operation. This is particularly important for offshore wind power systems, as the offshore environment presents significant maintenance challenges and demands even higher equipment reliability.

[0182] Step S5.4: Based on the control parameters as input, execute the control mode selection algorithm to select either the positive sequence voltage support control mode or the negative sequence voltage suppression control mode, and output the control mode signal; After determining the effective control parameters, the system needs to select an appropriate control mode based on the fault type and control requirements. Different types of power grid faults require different control strategies, especially since the handling methods for symmetrical and asymmetrical faults differ significantly. This step executes a control mode selection algorithm to choose the most suitable control mode based on the fault characteristics and control objectives.

[0183] The control mode selection is first based on fault type determination. The system analyzes the positive and negative sequence components of the grid voltage to determine whether the fault is symmetrical or asymmetrical. Symmetrical faults (such as three-phase short circuits) are mainly characterized by a uniform decrease in positive sequence voltage with a very small negative sequence component; asymmetrical faults (such as single-phase grounding, two-phase short circuits, etc.) are characterized by a decrease in positive sequence voltage accompanied by a significant negative sequence component. When the detected negative sequence voltage component exceeds 5% of the baseline value, the system determines it to be an asymmetrical fault; otherwise, it is determined to be a symmetrical fault.

[0184] For symmetrical faults, the system selects the positive-sequence voltage support control mode. The main objective of this mode is to provide reactive power support to help restore grid voltage. The control strategy calculates the required reactive current injection based on the deviation of the positive-sequence voltage, while considering the coordination between the system's active power balance and voltage support capability. Specifically, the system calculates the required reactive current injection according to the depth of voltage drop and a specified voltage-reactive power characteristic curve (usually a piecewise linear or quadratic curve), and adjusts the active current accordingly to ensure that the total current does not exceed the limit.

[0185] For asymmetrical faults, the system selects a composite control mode, simultaneously achieving both positive-sequence voltage support and negative-sequence voltage suppression. In this mode, the system needs to inject both positive-sequence and negative-sequence currents to suppress the negative-sequence voltage component in the grid and improve voltage imbalance. The control strategy is based on the decomposition of positive-sequence and negative-sequence voltages, calculating the required positive-sequence and negative-sequence current injection amounts separately. Since the converter's total current capacity is limited, the system needs to rationally allocate between positive-sequence and negative-sequence control, typically prioritizing positive-sequence voltage support and using the remaining capacity for negative-sequence voltage suppression.

[0186] In addition to basic voltage support and negative sequence suppression, the system also considers frequency support requirements. When the system frequency deviates from the rated value, especially when the frequency drop exceeds a preset threshold, the system will activate the frequency support mode, providing frequency support by adjusting the active power output. Frequency support and voltage support can occur simultaneously, and the system dynamically adjusts the distribution ratio of active and reactive power based on the voltage and frequency deviations.

[0187] The selection of control mode also takes into account the system's operating state and constraints. For example, when the system approaches the overcurrent limit, basic voltage support functions will be prioritized, and negative sequence suppression will be reduced or eliminated; when the converter approaches the modulation limit, the priority of the control target will be adjusted to ensure that the system can provide optimal support under constraints.

[0188] Finally, the system outputs control mode signals, including mode selection flags (such as positive sequence support, negative sequence suppression, frequency support, etc.) and parameter settings for each mode (such as support strength, response speed, etc.). These signals will serve as inputs for subsequent voltage transient control, guiding the execution of specific control strategies.

[0189] The appropriate selection of control mode is crucial for achieving effective transient control. By choosing the most suitable control mode for different fault types, the system can provide optimal grid support under limited resource conditions, thereby improving the overall fault response capability.

[0190] Step S5.5: Based on the control mode signal as input, execute the voltage transient control algorithm and output voltage control data; After determining the control mode, the system executes a voltage transient control algorithm to support and stabilize the grid voltage. Voltage transient control is a core function of offshore wind power systems under fault conditions, directly affecting the system's grid friendliness and fault ride-through capability. This step executes the corresponding control strategy based on the control mode signal and outputs the operational data required for voltage control.

[0191] For the positive-sequence voltage support mode, the system adopts a control strategy based on reactive current injection. First, the system calculates the degree of voltage drop in the grid, i.e., the deviation between the actual voltage and the rated voltage. Then, based on a preset voltage-reactive current characteristic curve, the required reactive current injection amount is calculated. A typical characteristic curve uses a piecewise linear form; for example, injecting 20% ​​of the rated reactive current when the voltage drops by 10%, injecting 40% of the rated reactive current when the voltage drops by 20%, and so on, until the upper limit of the reactive current is reached. The direction of reactive current injection is opposite to the direction of voltage drop; that is, the system injects inductive reactive current (lagging current) to support the voltage drop.

[0192] While calculating reactive current, the system also needs to consider active current adjustment. Since the converter's total current capacity is limited, increasing reactive current injection may reduce active current output. The system employs a priority strategy, prioritizing reactive power support during faults and then maintaining active current output as much as possible within the remaining capacity. Coordination between active and reactive power is achieved through real-time calculation and limiting of the current vector, ensuring that the total current does not exceed safe limits.

[0193] For composite modes involving negative-sequence voltage suppression, the system employs a more complex control strategy. First, the system decomposes the three-phase voltage into positive-sequence and negative-sequence components, calculating the positive-sequence voltage drop and negative-sequence voltage amplitude separately. Then, the system simultaneously calculates the required positive-sequence reactive current and negative-sequence current injection. The positive-sequence current is used to support the positive-sequence voltage, while the negative-sequence current is used to suppress the negative-sequence voltage. The phase relationship between the two needs to be precisely controlled to achieve the desired support effect.

[0194] The key to suppressing negative-sequence voltage lies in the calculation and control of negative-sequence current. The amplitude of the negative-sequence current is proportional to the negative-sequence voltage, but its phase requires special design. A typical method is to make the negative-sequence current conjugate with the negative-sequence voltage, i.e.: Ineg = k * Vneg* Where I_neg is the negative sequence current, Vneg* is the conjugate of the negative sequence voltage, and k is the proportionality coefficient. This design allows the negative sequence current to suppress the negative sequence voltage most effectively, improving the three-phase balance of the system.

[0195] When performing voltage transient control, the system also needs to consider dynamic response characteristics. To avoid oscillations and overshoots during the control process, the system employs dynamic limiting and smooth transition strategies. The rate of change of the current reference value is limited to ensure a smooth system response; simultaneously, the gain of the control loop is dynamically adjusted according to the system state, improving response speed in the early stages of a fault and ensuring stability during the recovery phase.

[0196] Ultimately, the system outputs voltage control data, including reference values ​​for positive and negative sequence currents, control gain parameters, and mode switching signals. This data will serve as input for subsequent frequency transient control, collectively forming a complete transient control strategy.

[0197] Voltage transient control is a direct reflection of the grid-friendliness of offshore wind power systems. Through precise voltage control, the system can provide effective grid support under fault conditions, improve grid voltage quality, and enhance the overall system stability and reliability.

[0198] Step S5.6: Execute the frequency transient control algorithm based on the voltage control data as input, output frequency control data, and dynamically adjust the control parameters based on the frequency control data.

[0199] After completing voltage transient control, the system executes a frequency transient control algorithm to respond to grid frequency disturbances and provide frequency support. Frequency transient control is a crucial function for offshore wind power systems to participate in grid frequency regulation, and it is of great significance for maintaining grid power balance and frequency stability. This step executes a frequency support strategy based on voltage control data and system frequency status, and dynamically adjusts control parameters based on the control effect.

[0200] Frequency transient control first requires detecting the deviation and rate of change of the grid frequency. The system uses phase-locked loop (PLL) technology to measure the grid frequency in real time and calculate the frequency deviation (the difference between the actual frequency and the rated frequency) and the rate of change of frequency (ROCOF). These two parameters are key indicators for evaluating the system's frequency status and determining support strategies. When the frequency deviation exceeds a preset threshold (typically ±0.2Hz) or the ROCOF exceeds a warning value (typically ±0.5Hz / s), the system triggers the frequency support function.

[0201] The core of frequency support is adjusting active power output in response to frequency changes. The system's control strategy comprises two parts: inertial response and frequency regulation. The inertial response simulates the rotor inertial characteristics of a synchronous generator, adjusting power output according to the rate of frequency change.

[0202] in, This refers to the inertial response power, Kinertia is the inertia coefficient, and df / dt is the rate of frequency change. The inertial response can provide rapid power support in the early stages of frequency changes and suppress rapid frequency changes.

[0203] Frequency regulation adjusts power output based on frequency deviation, similar to the speed governor function of a traditional generator. in, It refers to frequency regulation power, and Kdroop is the droop coefficient. It's a frequency deviation. Frequency regulation provides continuous power adjustment to help the system frequency return to its rated value.

[0204] After calculating the required power adjustment, the system needs to consider actual power constraints and coordination issues. Firstly, there's the limitation on the power adjustment rate. Considering the mechanical characteristics of the wind turbine, the power change rate should not be too large to avoid excessive mechanical stress. The system sets a power change rate limit, typically 10-20% / s of the rated power, to ensure smooth power adjustment without causing mechanical resonance.

[0205] Secondly, there is the management of power margin. Frequency support requires a certain power margin, especially when power output needs to be increased during frequency drops. The system creates power margin by actively reducing load (i.e., not operating at maximum power under normal conditions), or by temporarily increasing power output (such as utilizing the kinetic energy of the wind turbine and drive system) to provide short-term frequency support. For frequency rises, the system can reduce power output, which is usually easier to achieve.

[0206] Frequency support also needs to be coordinated with voltage support. In some cases, such as severe low-voltage faults, voltage support may need to be prioritized, in which case frequency support capacity may be limited. The system dynamically adjusts the active and reactive power allocation ratio according to the grid status and the priority of control objectives, providing frequency support as much as possible while meeting voltage support requirements.

[0207] Based on the effect of frequency control, the system dynamically adjusts control parameters to improve the support effect. Adjusted parameters include the inertia coefficient, droop coefficient, and power change rate limit. The system analyzes the effect of the frequency response, such as the frequency recovery speed and steady-state deviation, to evaluate the effectiveness of the current parameters and increase or decrease the support strength as needed. For example, if the frequency recovery is too slow, the droop coefficient may be increased; if a system oscillation trend is detected, the inertia coefficient may be decreased to increase damping characteristics.

[0208] Ultimately, the system outputs frequency control data, including power adjustment commands, updated control parameters, and operating mode information. Simultaneously, this data is fed back to the entire control system for optimization in the next cycle, forming a closed-loop adaptive control system.

[0209] Through frequency transient control, offshore wind power systems can effectively participate in grid frequency regulation, improving system frequency stability and grid friendliness. Especially in grids with a high proportion of wind power integration, the frequency support capability of wind power is crucial for the stable operation of the entire grid. The frequency transient control strategy proposed in this invention, by comprehensively considering mechanical and electrical characteristics, achieves safe and effective frequency support, making a significant contribution to improving the grid support capability of wind power systems.

[0210] In this embodiment, the execution of control commands first requires parameter configuration. Based on the received control command sequence, the system configures the converter's control parameters, including PI parameters for the current loop and voltage loop, and virtual impedance parameters. The parameter configuration process employs a piecewise linear interpolation method to ensure smooth parameter changes and avoid system oscillations caused by abrupt parameter changes. Simultaneously, the system performs parameter validity checks to ensure that the configured parameters are within the device's allowable range.

[0211] The transient control strategy employs a hierarchical control structure. The outer layer control is responsible for power distribution and operating mode switching, while the inner layer control implements specific voltage and current regulation. During a fault, the system automatically switches control modes based on the fault type. For example, in symmetrical fault conditions, positive-sequence voltage support control is used; in asymmetrical fault conditions, both positive-sequence voltage support and negative-sequence voltage suppression control are activated simultaneously. The control strategy executes at a frequency of 10kHz, which meets the requirements for rapid transient response.

[0212] Real-time monitoring is a crucial element in ensuring effective control. The monitoring system employs high-speed data acquisition devices to collect electrical and mechanical quantities of the fan in real time. The acquired data includes key parameters such as voltage, current, power, speed, and torque. The data acquisition frequency is 20kHz, meeting the requirements for transient process analysis. The monitoring system also includes fault diagnosis capabilities, enabling timely detection of control anomalies and triggering protective measures.

[0213] The dynamic adjustment based on monitoring results employs an adaptive control mechanism. The system dynamically adjusts control parameters by comparing the actual response with the expected response. The adjustment process uses fuzzy control rules to determine the parameter adjustment amount based on the magnitude and trend of the deviation. For example, when a slow voltage recovery rate is detected, the system appropriately increases the proportional gain of the voltage control loop; when excessive mechanical stress is detected, the response speed of the power support is reduced. This adaptive mechanism ensures that the control system maintains good performance under different operating conditions. Example

[0214] The present invention also provides a transient control system for offshore wind power, comprising: Data acquisition module 10 is used to acquire voltage, current and frequency data of offshore wind power system, filter and synchronize timestamp data to obtain offshore wind power system operating status information; The fault analysis module 20 is used to determine whether the offshore wind power system has a fault based on the operating status information of the offshore wind power system, using a sliding time window detection algorithm, and to obtain fault characteristic parameters by analyzing the fault type through symmetric component decomposition. The constraint modeling module 30 is used to establish equipment overcurrent capacity constraint and converter modulation constraint models based on the fault characteristic parameters, construct the discrete-time offshore wind power system robust predictive control barrier function, and obtain the constraint model framework. The control strategy module 40 is used to construct a dynamic model of the wind turbine that considers the mechanical-electrical coupling characteristics based on the constraint model framework, execute an iterative learning predictive control algorithm, and generate a sequence of control commands. The control execution module 50 is used to configure control parameters according to the control instruction sequence, execute transient control strategies, and dynamically adjust the control parameters based on real-time monitoring results.

[0215] In this embodiment, the data acquisition module 10 includes multiple distributed measurement units deployed at key nodes of the offshore wind farm to collect data such as voltage, current, and frequency in real time. The collected data is processed by a filtering algorithm to eliminate noise interference, and then time consistency is ensured through synchronization timestamp technology. The processed data forms system operating status information for use by the fault analysis module.

[0216] The fault analysis module 20 employs a sliding time window-based fault detection algorithm to monitor the system's operating status in real time and identify potential faults. Using symmetrical component theory, it decomposes the three-phase voltage and current into positive-sequence, negative-sequence, and zero-sequence components to determine the fault type. Simultaneously, it calculates characteristic parameters such as fault depth and duration, providing a basis for subsequent control strategy decisions.

[0217] The constraint modeling module 30 is responsible for establishing the constraint condition model for system operation. Based on the equipment's overcurrent capability, a current constraint model is established; considering the converter's modulation characteristics, a modulation constraint model is constructed. Using robust predictive control theory, a control barrier function is designed to ensure the system operates safely under the constraints. All constraints are integrated into a unified constraint model framework, serving as input for control strategy design.

[0218] The control strategy module 40 first constructs a dynamic model of the wind turbine, which comprehensively considers the mechanical and electrical coupling characteristics. Based on this model, an iterative learning predictive control algorithm is designed to improve control performance through continuous learning and optimization. The control strategy comprehensively considers multiple control objectives such as voltage support and frequency support, and generates the optimal control command sequence by solving a multi-objective optimization problem.

[0219] The control execution module 50 receives a sequence of control commands and configures the converter's control parameters. The parameter configuration process ensures a smooth transition to avoid system oscillations. Based on the fault type, the system automatically selects an appropriate control mode and executes a transient control strategy. Simultaneously, by monitoring the control effect in real time and dynamically adjusting the control parameters, adaptive control is achieved, improving the system's robustness and adaptability.

[0220] This invention discloses a method and system for transient control of offshore wind power. Through five key steps—system state monitoring, fault characteristic analysis, constraint modeling, multi-objective control strategy design, and transient control execution—it achieves optimized transient control of offshore wind power systems in the face of various grid faults. This method considers equipment constraints and mechanical-electrical coupling characteristics, significantly improving the system's grid support capability and operational reliability, and is of great significance for promoting the large-scale application of offshore wind power.

[0221] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall fall within the protection scope of the present invention.

[0222] This disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of a method for a transient control strategy for offshore wind power as described in the above-described method embodiments. The storage medium can be a volatile or non-volatile computer-readable storage medium.

[0223] Furthermore, this disclosure also provides a computer program product storing a computer program. When the computer program is run by a processor, it executes the steps of a method for a transient control strategy for offshore wind power provided in any of the above embodiments of this disclosure. For details, please refer to the above method embodiments, which will not be repeated here.

[0224] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium, which can be a volatile or non-volatile computer-readable storage medium. In another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0225] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices and apparatuses described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. In the several embodiments provided in this disclosure, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the displayed or discussed mutual coupling or direct coupling or communication connection may be through some communication interfaces; the indirect coupling or communication connection of devices or units may be electrical, mechanical, or other forms.

[0226] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0227] In addition, the functional units in the various embodiments of this disclosure can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0228] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this disclosure, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this disclosure. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0229] Finally, it should be noted that the above-described embodiments are merely specific implementations of this disclosure, used to illustrate the technical solutions of this disclosure, and not to limit it. The protection scope of this disclosure is not limited thereto. Although this disclosure has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some of the technical features, within the scope of the technology disclosed in this disclosure; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this disclosure, and should all be covered within the protection scope of this disclosure. Therefore, the protection scope of this disclosure should be determined by the protection scope of the claims.

Claims

1. A transient control method for offshore wind power, characterized in that, include: The voltage, current, and frequency data of the offshore wind power system are collected, and the voltage, current, and frequency data are filtered and time-stamped to obtain the operating status information of the offshore wind power system. Based on the operating status information of the offshore wind power system, a sliding time window detection algorithm is used to determine whether the offshore wind power system has a fault. If a fault occurs, the fault type is analyzed by symmetric component decomposition to obtain fault characteristic parameters. Based on the fault characteristic parameters, constraints on equipment overcurrent capacity and converter modulation are established, a robust predictive control barrier function for discrete-time offshore wind power system is constructed, and a constraint model framework is generated. Based on the aforementioned constraint model framework, a wind turbine dynamic model considering mechanical-electrical coupling characteristics is constructed. The wind turbine dynamic model is then used to execute an iterative learning predictive control algorithm to generate a sequence of control commands. Configure control parameters according to the control command sequence, execute transient control strategies, and dynamically adjust the control parameters based on real-time monitoring results.

2. The method according to claim 1, characterized in that, Voltage, current, and frequency data of the offshore wind power system are collected. This data is then filtered and time-stamped to obtain the operating status information of the offshore wind power system, including: The voltage, current and frequency data of the offshore wind power system are collected as input, the voltage deviation rate is calculated, and voltage assessment data is obtained. The frequency deviation rate is calculated based on the voltage evaluation data as input, and the frequency evaluation data is obtained. Based on the voltage assessment data and the frequency assessment data as input, a weighted average algorithm is executed to perform data aggregation analysis and output the operating status information of the offshore wind power system.

3. The method according to claim 1, characterized in that, Based on the operational status information of the offshore wind power system, a sliding time window detection algorithm is used to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault type is analyzed through symmetric component decomposition to obtain fault characteristic parameters, including: The system receives the operating status information of the offshore wind power system as input, compares the deviation between the voltage of the offshore wind power system and the voltage threshold, and outputs the voltage fault detection result. The offshore wind power system's operating status information is used as input, and the deviation between the offshore wind power system's current and the rated current value is compared to output the current fault detection result. Based on the voltage fault detection results and the current fault detection results as input, a fault determination matrix algorithm is executed for comprehensive analysis to determine whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault characteristic parameters are output, including the fault type, fault severity, and fault duration.

4. The method according to claim 3, characterized in that, Based on a comprehensive analysis of the voltage fault detection results and the current fault detection results, it is determined whether a fault has occurred in the offshore wind power system. If a fault has occurred, the fault characteristic parameters are obtained, including: The voltage and current of the offshore wind power system are used as inputs to perform positive-sequence component, negative-sequence component and zero-sequence component decomposition, and the component decomposition data is output. Using the component decomposition data as input, the ratio of the negative-order component to the base value is calculated, and asymmetric fault judgment data is output. The fault type is determined by executing a fault type identification algorithm based on the asymmetric fault judgment data as input, the fault feature parameters are updated, and the results are output for subsequent constraint model construction.

5. The method according to claim 1, characterized in that, By establishing constraints on equipment current carrying capacity and converter modulation, a robust predictive control barrier function for the discrete-time offshore wind power system is constructed, generating a constraint model framework, including: Based on the fault characteristic parameters as input, the current of the offshore wind power system is limited to no more than 1.2 times the rated current, and an overcurrent constraint condition is output. Based on the overcurrent constraint condition as input, the converter modulation ratio is limited within the linear modulation region, and the modulation constraint condition is output. Based on the overcurrent constraint and the modulation constraint as input, the control barrier function construction algorithm is executed to output the constraint model framework.

6. The method according to claim 1, characterized in that, Based on the aforementioned constraint model framework, a dynamic model of the wind turbine considering the mechanical-electrical coupling characteristics is constructed, including: Using the aforementioned constraint model framework as input, the wind turbine, drive shaft, and generator rotor are constructed as a multi-mass block structure model, and the inertia distribution data is output. Based on the inertia distribution data as input, the spring-damping relationship construction algorithm is executed to establish the spring-damping connection relationship and output mechanical dynamic characteristic data. Based on the mechanical dynamic characteristic data and electromagnetic characteristic parameters as input, a coupled model construction algorithm is executed to output the dynamic response characteristics of the wind turbine dynamic model.

7. The method according to claim 6, characterized in that, The method further includes: executing an iterative learning prediction control algorithm based on the dynamic response characteristics, including: Using the aforementioned dynamic response characteristics as input, the dynamic response of the offshore wind power system is predicted based on the state-space equation, and the predicted response data is output. The deviation between the predicted response data and the expected response is calculated as input, a parameter optimization algorithm is executed, and the updated control parameter values ​​are output. The control parameters are optimized by an iterative optimization algorithm based on the updated control parameter values ​​as input, and the control command sequence is output.

8. The method according to claim 1, characterized in that, Configure control parameters according to the control command sequence, including: The system receives the control command sequence as input, configures the current loop and voltage loop parameters of the converter, and outputs the initial control parameters. The initial control parameters are used as input to perform a piecewise linear interpolation algorithm, and the output is the smoothed control parameters. Using the smoothed control parameters as input, the parameter verification algorithm is executed to check whether they are within the rated parameter range, and outputs valid control parameters.

9. The method according to claim 1, characterized in that, Implement transient control strategies, including: Based on the control parameters as input, the control mode selection algorithm is executed to select either the positive sequence voltage support control mode or the negative sequence voltage suppression control mode, and the control mode signal is output. Based on the control mode signal as input, a voltage transient control algorithm is executed, and voltage control data is output. The frequency transient control algorithm is executed based on the voltage control data as input, outputs frequency control data, and dynamically adjusts the control parameters based on the frequency control data.

10. A transient control system for offshore wind power, characterized in that, include: The data acquisition module is used to collect voltage, current and frequency data of the offshore wind power system, filter the data and synchronize the timestamp to obtain the operating status information of the offshore wind power system. The fault analysis module is used to determine whether the offshore wind power system has a fault based on the operating status information of the offshore wind power system, using a sliding time window detection algorithm, and to obtain fault characteristic parameters by analyzing the fault type through symmetric component decomposition. The constraint modeling module is used to establish equipment overcurrent capacity constraint and converter modulation constraint models based on the fault characteristic parameters, construct the robust predictive control barrier function of the discrete-time offshore wind power system, and obtain the constraint model framework. The control strategy module is used to construct a dynamic model of the wind turbine that considers the mechanical-electrical coupling characteristics based on the constraint model framework, execute an iterative learning predictive control algorithm, and generate a sequence of control commands. The control execution module is used to configure control parameters according to the control instruction sequence, execute transient control strategies, and dynamically adjust the control parameters based on real-time monitoring results.

Images