A method and system for protecting the power module of a wind turbine generator set

By analyzing voltage fluctuation characteristics in real time and dynamically adjusting protection parameters, the problem of damage to the power module of the wind turbine generator set caused by voltage transient impact and excitation inrush current during the closing operation is solved, achieving highly reliable and fast power module protection.

CN123121440APending Publication Date: 2026-11-17THREE GORGES NEW ENERGY OFFSHORE WIND POWER OPERATION & MAINTENANCE JIANGSU CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611021521.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-07-09
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In the current technology, the power module cannot be effectively protected from high-amplitude voltage transients and inrush current damage during the closing operation of wind turbine generator sets. Furthermore, the existing protection scheme cannot adaptively cope with changes in the duration of the impact caused by the strength of the power grid and the drift of electrical component parameters.

Method used

The voltage fluctuation characteristics are analyzed in real time using differential operation. The degree of voltage fluctuation is judged by the first and second derivative sequences. The protection parameters are dynamically adjusted by combining the exponential weighted moving average method. The voltage zero-crossing shielding stage and dual redundant switch coordinated control are configured to realize intelligent conduction timing decision.

Benefits of technology

It achieves closed-loop intelligent decision-making on the timing of closing and conduction, adaptively matches the actual transient process length, improves the reliability and speed of the protection system, and avoids damage to the power module caused by delay mismatch.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123121440A_ABST
    Figure CN123121440A_ABST
Patent Text Reader

Abstract

This invention discloses a protection method and system for a wind turbine generator power module. The method involves acquiring a sequence of instantaneous voltage values ​​at the power module input terminal using a voltage acquisition module with a fixed sampling period. The control unit performs first-order and second-order differential operations on this sequence to obtain first-order and second-order derivative sequences characterizing the severity of voltage fluctuations. After a zero-point shielding phase, the control unit continuously checks whether the absolute values ​​of the derivative sequences are less than the first-order impulse slope threshold and the second-order impulse curvature threshold, respectively. When the stability criteria are consistently met, the control unit generates a conduction command, synchronously driving two parallel delayed conduction switches to close, thus connecting the power module's power supply circuit. This invention dynamically updates the reference stabilization time and adaptively adjusts the confirmation time window length for the next protection action, achieving self-learning and adaptation of the protection strategy to changes in grid operating conditions. This fundamentally solves the failure risk of fixed-delay protection in closing transient protection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of electrical protection technology for wind turbine generator sets, specifically relating to a protection method and system for the power module of a wind turbine generator set. Background Technology

[0002] In the operation and maintenance of wind power plants, operators often need to perform power outage and power restoration operations on the collector lines due to site shutdowns, planned maintenance, or troubleshooting. A typical scenario is as follows: the 35kV collector line switchgear is disconnected on-site, but the 35kV ring main unit of the generator is kept closed. After maintenance is completed, the 35kV switchgear is closed again to re-energize the line. Since the wind turbine generator is in a completely offline state at this time, the input terminal of the power module is directly exposed to the electrical stress generated by this closing operation. At the moment of closing, the system phase voltage charges the equivalent capacitance of the collector lines and the box transformer. However, inevitably, inductive components in the line, such as the internal impedance of the power supply, the leakage inductance of the box transformer, and the distributed inductance of the collector lines themselves, form a resonant circuit with the capacitor. This results in the charging process not being a smooth voltage build-up, but rather a violent high-frequency transient oscillation process. The voltage amplitude generated by this transient process far exceeds the power frequency overvoltage during normal operation, theoretically reaching two to three times the peak value of the system phase voltage. After attenuation by line resistance and corona loss, it can still typically reach 2.0 to 2.5 times the steady-state peak value. Simultaneously, at the moment of closing, the transformer core flux needs to build up from zero or residual magnetism to a steady state. If the closing time occurs near the zero-crossing point of the power supply voltage, the voltage change rate is at its maximum, and a DC bias flux will be generated in the core to counteract the steady-state flux, causing the instantaneous total flux to reach nearly twice the steady-state value, resulting in deep core saturation. Core saturation causes a sharp drop in the transformer's magnetizing inductance, which in turn draws a decaying inrush current from the power supply side with an amplitude of 6 to 12 times or even higher than the rated current, containing a large number of second harmonics and aperiodic components, lasting for hundreds of milliseconds to one second.

[0003] These severe voltage and current surges directly affect the input terminals of the power supply modules that convert 400V to 24V to power control and communication loads such as PLCs, switches, contactors, and relays. This constitutes a major cause of power supply module failure and seriously threatens the stable operation of the generator control system. Therefore, the technical problem addressed in this application is: how to effectively protect the power supply modules in wind turbine generator sets from high-amplitude voltage transients and inrush currents during the moment the collector line is energized and closed.

[0004] To address the transient overvoltage and inrush current issues caused by the closing operation, existing technologies typically employ two types of protection methods. The first type is passive protection based on energy absorption and clamping. A typical approach is to connect a metal oxide varistor (MOV) or a transient voltage suppressor diode (TVS) in parallel at the input of the power module. Utilizing the characteristic that the impedance of these devices drops sharply when the terminal voltage exceeds their nominal breakdown voltage, the overvoltage energy is bypassed and discharged as heat, thereby clamping the input voltage to a relatively safe level.

[0005] The second type is active isolation protection based on fixed-delay conduction. A typical solution involves inserting a time-delay relay in series in the input circuit of the power module. The relay coil is energized the instant the power supply circuit becomes live, but its main contacts do not close immediately. Instead, they conduct the power supply circuit of the power module only after a pre-set fixed delay. The design concept of this type of solution is to artificially isolate the power module from the initial transient impact period using the mechanical or electronic delay set by the time relay. By the time the relay contacts close, the preset worst voltage and current surge peaks have passed, and the grid voltage has stabilized. Some solutions also employ a redundant design with two time relays connected in parallel to improve the reliability of a single delay circuit, preventing conduction failure due to contact sticking or coil burnout in a single relay.

[0006] However, the aforementioned existing technologies have significant technical shortcomings when dealing with the specific closing transient process of wind turbine generators: For passive protection schemes using varistors or transient suppression diodes, their protection capability is limited by the current carrying capacity and energy tolerance limits of the devices. In applications such as wind turbine generators with high voltage levels and large line capacitance, the energy of a single closing surge may far exceed the single absorption capacity of ordinary surge protection devices, causing device performance degradation or even irreversible thermal collapse.

[0007] Furthermore, these devices respond to voltage amplitude and have almost no ability to suppress the heating and stress impact on the rectifier bridge and filter capacitors inside the power module caused by the inrush current. The core drawback of active isolation protection schemes using fixed-delay conduction is that the set delay is fixed and open-loop. This fixed delay value is a constant set by the designer based on experience or a rough estimate of the worst-case impact duration. On the one hand, if the set delay is insufficient and fails to fully cover the actual transient process, the power module will still be subjected to the tail voltage fluctuation of the impact when the relay closes. On the other hand, if an excessively long delay is set to ensure safety, it will cause the control system to be delayed in starting due to prolonged inability to power on, which is unacceptable in applications such as offshore wind power where response speed is crucial. More importantly, the duration of the impact transient changes dynamically with the strength of the power grid, the parameter drift of electrical components, and the complex electromagnetic coupling relationships between multiple units. A fixed, non-adaptively adjustable delay value cannot achieve an optimal balance between safety and timeliness. Therefore, a protection method and system for wind turbine power modules needs to be designed to solve these problems. Summary of the Invention

[0008] The purpose of this invention is to provide a protection method for the power module of a wind turbine generator set, which aims to overcome the problem that existing fixed-delay protection methods cannot adaptively identify the safe conduction time while taking into account the dynamic changes in the transient intensity and duration of the closing impact. It has the characteristics of achieving a leap from passive fixed waiting to active intelligent redundancy protection by performing real-time gradient analysis of voltage fluctuation characteristics and self-learning conduction timing decision.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: A protection method for a wind turbine generator power module, applied to a protection system comprising a control unit, a voltage acquisition module, and at least one time-delayed on-off switch, the method comprising the following steps: The voltage acquisition module acquires the instantaneous voltage value sequence at the input terminal of the power module at a preset sampling period, denoted as v(k), where k is the sampling sequence number and k is a natural number; By performing first-order and second-order difference operations on the instantaneous voltage value sequence through the control unit, a first-order derivative sequence characterizing the degree of voltage fluctuation is obtained. and second derivative sequence ; The control unit determines whether the absolute value of the first derivative sequence is continuously less than the first derivative impact slope threshold and whether the absolute value of the second derivative sequence is continuously less than the second derivative impact curvature threshold within a continuous confirmation time window. When the determination is correct, the control unit generates a conduction command, drives the delayed conduction switch to close, and turns on the power supply circuit of the power module.

[0010] Preferably, the time-domain fluctuation characteristic parameters include a sequence of the first derivatives of the instantaneous voltage values. and second derivative sequence ; Calculate the time-domain fluctuation characteristic parameters that characterize the severity of voltage transient fluctuations, specifically including: By performing difference operations on the instantaneous voltage value sequence, the first-order derivative sequence and the second-order derivative sequence are obtained, and the calculation formulas are as follows: ; ; Where v(k) represents the instantaneous voltage value at the k-th sampling time, v(k-1) represents the instantaneous voltage value at the (k-1)-th sampling time, and v(k-2) represents the instantaneous voltage value at the (k-2)-th sampling time. This indicates the preset sampling period.

[0011] Preferably, the stability criterion is as follows: Within the continuous confirmation time window, the absolute value of the first derivative sequence All are less than the preset first derivative impact slope threshold And the absolute value of the second derivative sequence All are less than the preset second derivative impact curvature threshold. The first derivative impact slope threshold and second derivative impact curvature threshold All are based on the rated phase voltage peak value of the wind turbine generator set. Set the frequency f with the power grid frequency and satisfy the following: ; ; in, This represents the rated peak phase voltage at the input of the power module, where f represents the mains frequency. and It is a tuning coefficient greater than 1.

[0012] Preferably, the method further includes: Record the time period from the end of the shielding phase to the moment when the control unit generates the conduction command each time the conduction is successfully turned on, and use it as the effective stabilization time t(n) for the nth time; The initial reference stabilization time preset during the first run. Substituting the effective settling time t(1) of the first recorded value into the exponentially weighted moving average formula, we obtain the baseline settling time after the first processing. ; After each successful conduction, the baseline settling time is iteratively updated using an exponentially weighted moving average method. The update formula is as follows: ; in, This represents the baseline settling time after the nth processing. Let t(n) represent the baseline settling time after the (n-1)th processing, and t(n) represent the effective settling time of the nth record. Represents the smoothing factor and ; Based on the benchmark settling time The length of the confirmation time window for the next protection startup is dynamically updated, specifically including: utilizing the reference settling time. and safety factor Calculate the updated confirmation time window length. The calculation formula is: ; in, The preset safety factor is greater than 1.

[0013] Furthermore, before calculating the time-domain fluctuation characteristic parameters, a shielding phase is included: the control unit continuously detects the zero-crossing points of the instantaneous voltage value v(k), and the shielding phase continues until the number of detected zero-crossing points reaches a preset number of shielding cycles. Then, the calculation of the first-order derivative sequence and the second-order derivative sequence is initiated, where N0 is a positive integer.

[0014] Preferably, a wind turbine generator power module protection system includes: The voltage acquisition module is used to acquire the instantaneous voltage value sequence v(k) at the input terminal of the power supply module in real time. The control unit is connected to the voltage acquisition module. The control unit is equipped with a memory, a processor, and a program stored in the memory and executed by the processor. When the program is executed, it calculates the time-domain fluctuation characteristic parameters, judges the stability criteria, and generates the turn-on command. The first delay switch and the second delay switch have their input terminals connected in parallel to the positive terminal of the power supply circuit, their output terminals connected in parallel to the positive input terminal of the power module, and their control terminals connected to the first output port and the second output port of the control unit, respectively. The control unit sends conduction commands to the first delayed conduction switch and the second delayed conduction switch through the first output port and the second output port respectively, so as to synchronously drive them to close. A manual bypass switch is provided, with one end connected to the positive terminal of the power supply circuit and the other end connected to the positive input terminal of the power module. The manual bypass switch is configured to be manually closed by the operator to forcibly connect the power supply circuit of the power module when the control unit fails.

[0015] Furthermore, both the first delayed-on switch and the second delayed-on switch are energized delayed-closing time relays; the voltage acquisition module is an isolated Hall voltage sensor or a voltage acquisition circuit coupled with an isolation transformer.

[0016] Preferably, the control unit is further configured with a zero-crossing detection module, which is used to detect the zero-crossing point of the instantaneous voltage value and generate a zero-crossing count. The control unit is used to start the calculation of the time-domain fluctuation characteristic parameter after the zero-crossing count reaches a preset number of shielding cycles N0.

[0017] Furthermore, the control unit is also equipped with a communication interface, which is used to record the instantaneous voltage value sequence v(k), the effective settling time t(n), and the generated reference settling time for each closing process. The recorded data is then uploaded to the host computer monitoring system via a communication interface.

[0018] Preferably, the program of the control unit further includes a threshold adaptive adjustment module, which is used to adjust the threshold adaptive adjustment module based on the first derivative sequence of historical records. The maximum value and the second derivative sequence The maximum value of the first derivative impacts the slope threshold within a preset tuning period. and the second derivative impact curvature threshold Make corrections.

[0019] Preferably, the slope threshold of the first derivative is used. and second derivative impact curvature threshold The corrections will be made in the following manner: The peak value of the absolute value of the first derivative sequence before the end of the confirmation time window during the most recent M closing processes is denoted as . And the peak value of the absolute value of the second derivative sequence, denoted as ,in ; Calculate the sequence respectively and The exponentially weighted moving average yields the average first-order peak value. and average second-order peak ; After processing each set of M peaks, the average first-order peak value is... and the average second-order peak value Perform iterative updates: ; ; Where j is the iteration number, The preset threshold smoothing factor and , and These are the average first-order peak value and the average second-order peak value after the previous round of processing, respectively. Impact the slope threshold with the first derivative Updated to The second derivative impacts the curvature threshold. Updated to ,in and These are preset margin coefficients, all of which are greater than 1.

[0020] Furthermore, the preset initial reference settling time Based on the rated capacity of the wind turbine generator transformer and the equivalent length of the collector line, a fixed value greater than 0.5 seconds and less than 2 seconds is set by referring to a table.

[0021] Preferably, the device further includes a computer device, comprising a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the wind turbine power module protection method.

[0022] Preferably, the method further includes a computer-readable storage medium storing computer instructions for causing a computer to execute the wind turbine power module protection method.

[0023] The beneficial effects of the wind turbine generator power module protection method provided by this invention are as follows: 1. This invention achieves closed-loop intelligent decision-making for the timing of circuit breaker activation by employing differential computation to analyze voltage fluctuation characteristics in real time and setting dynamic stability criteria. This solves the defect in existing technologies where fixed-delay protection cannot adaptively match the actual transient process length. Unlike existing technologies that pre-set a blind and fixed waiting time, this solution utilizes the control unit to continuously collect instantaneous voltage values ​​and calculates online the first and second derivative sequences characterizing the intensity of the waveform. By continuously comparing the absolute values ​​of these two derivatives with the impact slope and impact curvature thresholds set according to system physical parameters, a activation command is only generated when the slope and curvature of the transient oscillation simultaneously drop below the thresholds, confirming that the voltage fluctuation has indeed subsided. This judgment logic, based on real-time signal dynamic characteristics rather than a preset time, ensures that the power module's activation time precisely corresponds to the actual end of the transient process, fundamentally eliminating the technical problems of missed protection windows due to excessively short delays or ineffective system waiting due to excessively long delays.

[0024] 2. This invention introduces an adaptive learning and correction mechanism for turn-on timing based on an exponentially weighted moving average method, endowing the protection system with the ability to autonomously adapt to changes in operating conditions. This solves the problem in existing technologies where protection parameter settings rely on manual experience and fail after system parameter drift. The fixed delay values ​​in existing solutions cannot be changed once written. This solution records the actual effective stabilization time during each successful turn-on process and uses this as a sample. Through an exponentially weighted moving average method that assigns greater weight to recent samples, a benchmark stabilization time reflecting the actual transient characteristics of the current system is iteratively updated. Then, this benchmark stabilization time is used to dynamically adjust the confirmation time window length for the next protection judgment. This allows the protection system to continuously track changes in grid impedance due to seasonal changes, parameter drift caused by equipment aging, and even differences in the electrical environment between different wind farms. This ensures that the confirmation time window length is always within the optimal range that balances reliability and speed, achieving an evolutionary protection performance that becomes more accurate with operation.

[0025] 3. This invention, through the configuration of a voltage zero-crossing shielding stage and a control architecture with dual redundant switches, comprehensively avoids the combined damage of voltage and current surges, and significantly improves the reliability and safety of protection actions. Addressing the problem that existing technologies cannot effectively handle transformer inrush current, this invention sets up a shielding stage based on voltage zero-crossing counts before initiating fluctuation characteristic calculations. This stage allows skipping the time interval where the initial inrush current is most intense due to core saturation after closing, but the voltage waveform is distorted due to deep saturation, thus avoiding interference from complex distorted waveforms on subsequent voltage fluctuation characteristic calculation algorithms. Simultaneously, after the control unit calculates the conduction decision, it drives two parallel-connected delayed conduction switches to close synchronously through independent ports. This architecture, combining an intelligent decision-making core with redundant execution terminals, overcomes the blindness of simple hardware redundancy in strategy and avoids the reliability shortcomings that pure electronic solutions may have when driving a single contact, resulting in extremely high overall stability of the entire link from transient detection to reliable power supply. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the system structure of the present invention; Figure 3 This is a box plot comparing the startup delay distribution of three protection schemes in an embodiment of the present invention; Figure 4 This is a schematic diagram of the adaptive convergence trend of the startup delay of the method in this embodiment of the invention and the band error of the 95% confidence interval; Figure 5 This is a schematic diagram of the structure of a computer device in an embodiment of the present invention. Detailed Implementation

[0027] Example 1: like Figure 1 As shown, a protection method for a wind turbine generator power module is applied to a protection system comprising a control unit, a voltage acquisition module, and at least one time-delayed on-off switch. The method includes the following steps: The voltage acquisition module acquires the instantaneous voltage value sequence at the input terminal of the power module at a preset sampling period, denoted as v(k), where k is the sampling sequence number and k is a natural number; By performing first-order and second-order difference operations on the instantaneous voltage value sequence through the control unit, a first-order derivative sequence characterizing the degree of voltage fluctuation is obtained. and second derivative sequence ; The control unit determines whether the absolute value of the first derivative sequence is continuously less than the first derivative impact slope threshold and whether the absolute value of the second derivative sequence is continuously less than the second derivative impact curvature threshold within a continuous confirmation time window. When the determination is correct, the control unit generates a conduction command, drives the delayed conduction switch to close, and turns on the power supply circuit of the power module.

[0028] Preferably, the time-domain fluctuation characteristic parameters include a sequence of the first derivatives of the instantaneous voltage values. and second derivative sequence ; Calculate the time-domain fluctuation characteristic parameters that characterize the severity of voltage transient fluctuations, specifically including: By performing difference operations on the instantaneous voltage value sequence, the first-order derivative sequence and the second-order derivative sequence are obtained, and the calculation formulas are as follows: ; ; Where v(k) represents the instantaneous voltage value at the k-th sampling time, v(k-1) represents the instantaneous voltage value at the (k-1)-th sampling time, and v(k-2) represents the instantaneous voltage value at the (k-2)-th sampling time. This indicates the preset sampling period.

[0029] Preferably, the stability criterion is as follows: Within the continuous confirmation time window, the absolute value of the first derivative sequence All are less than the preset first derivative impact slope threshold And the absolute value of the second derivative sequence All are less than the preset second derivative impact curvature threshold. The first derivative impact slope threshold and second derivative impact curvature threshold All are based on the rated phase voltage peak value of the wind turbine generator set. Set the frequency f with the power grid frequency and satisfy the following: ; ; in, This represents the rated peak phase voltage at the input of the power module, where f represents the mains frequency. and It is a tuning coefficient greater than 1.

[0030] Preferably, the method further includes: Record the time period from the end of the shielding phase to the moment when the control unit generates the conduction command each time the conduction is successfully turned on, and use it as the effective stabilization time t(n) for the nth time; The initial reference stabilization time preset during the first run. Substituting the effective settling time t(1) of the first recorded value into the exponentially weighted moving average formula, we obtain the baseline settling time after the first processing. ; After each successful conduction, the baseline settling time is iteratively updated using an exponentially weighted moving average method. The update formula is as follows: ; in, This represents the baseline settling time after the nth processing. Let t(n) represent the baseline settling time after the (n-1)th processing, and t(n) represent the effective settling time of the nth record. Represents the smoothing factor and ; Based on the benchmark settling time The length of the confirmation time window for the next protection startup is dynamically updated, specifically including: utilizing the reference settling time. and safety factor Calculate the updated confirmation time window length. The calculation formula is: ; in, The preset safety factor is greater than 1.

[0031] Furthermore, before calculating the time-domain fluctuation characteristic parameters, a shielding phase is included: the control unit continuously detects the zero-crossing points of the instantaneous voltage value v(k), and the shielding phase continues until the number of detected zero-crossing points reaches a preset number of shielding cycles. Then, the calculation of the first-order derivative sequence and the second-order derivative sequence is initiated, where N0 is a positive integer.

[0032] like Figure 2 As shown, a wind turbine generator power module protection system includes: The voltage acquisition module is used to acquire the instantaneous voltage value sequence v(k) at the input terminal of the power supply module in real time. The control unit is connected to the voltage acquisition module. The control unit is equipped with a memory, a processor, and a program stored in the memory and executed by the processor. When the program is executed, it calculates the time-domain fluctuation characteristic parameters, judges the stability criteria, and generates the turn-on command. The first delay switch and the second delay switch have their input terminals connected in parallel to the positive terminal of the power supply circuit, their output terminals connected in parallel to the positive input terminal of the power module, and their control terminals connected to the first output port and the second output port of the control unit, respectively. The control unit sends conduction commands to the first delayed conduction switch and the second delayed conduction switch through the first output port and the second output port respectively, so as to synchronously drive them to close. A manual bypass switch is provided, with one end connected to the positive terminal of the power supply circuit and the other end connected to the positive input terminal of the power module. The manual bypass switch is configured to be manually closed by the operator to forcibly connect the power supply circuit of the power module when the control unit fails.

[0033] Furthermore, both the first delayed-on switch and the second delayed-on switch are energized delayed-closing time relays; the voltage acquisition module is an isolated Hall voltage sensor or a voltage acquisition circuit coupled with an isolation transformer.

[0034] Preferably, the control unit is further configured with a zero-crossing detection module, which is used to detect the zero-crossing point of the instantaneous voltage value and generate a zero-crossing count. The control unit is used to start the calculation of the time-domain fluctuation characteristic parameter after the zero-crossing count reaches a preset number of shielding cycles N0.

[0035] Furthermore, the control unit is also equipped with a communication interface, which is used to record the instantaneous voltage value sequence v(k), the effective settling time t(n), and the generated reference settling time for each closing process. The recorded data is then uploaded to the host computer monitoring system via a communication interface.

[0036] Preferably, the program of the control unit further includes a threshold adaptive adjustment module, which is used to adjust the threshold adaptive adjustment module based on the first derivative sequence of historical records. The maximum value and the second derivative sequence The maximum value of the first derivative impacts the slope threshold within a preset tuning period. and the second derivative impact curvature threshold Make corrections.

[0037] Preferably, the slope threshold of the first derivative is used. and second derivative impact curvature threshold The corrections will be made in the following manner: The peak value of the absolute value of the first derivative sequence before the end of the confirmation time window during the most recent M closing processes is denoted as . And the peak value of the absolute value of the second derivative sequence, denoted as ,in ; Calculate the sequence respectively and The exponentially weighted moving average yields the average first-order peak value. and average second-order peak ; After processing each set of M peaks, the average first-order peak value is... and the average second-order peak value Perform iterative updates: ; ; Where j is the iteration number, The preset threshold smoothing factor and , and These are the average first-order peak value and the average second-order peak value after the previous round of processing, respectively. Impact the slope threshold with the first derivative Updated to The second derivative impacts the curvature threshold. Updated to ,in and These are preset margin coefficients, all of which are greater than 1.

[0038] Furthermore, the preset initial reference settling time Based on the rated capacity of the wind turbine generator transformer and the equivalent length of the collector line, a fixed value greater than 0.5 seconds and less than 2 seconds is set by referring to a table.

[0039] Preferably, the device further includes a computer device, comprising a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the wind turbine power module protection method.

[0040] Preferably, the method further includes a computer-readable storage medium storing computer instructions for causing a computer to execute the wind turbine power module protection method.

[0041] Example 2: This embodiment applies the wind turbine generator power module protection method and system proposed in this invention to an offshore wind farm, which is responsible for supplying power to the generator's main control PLC, switch, and various contactors and relays.

[0042] Offshore wind farms operate in high-salt-spray, high-humidity marine environments, and face maintenance windows limited by typhoons and sea ice each summer and winter, respectively. This results in frequent outages and power restoration operations for the collection lines. Before adopting this invention, after each maintenance, 8 to 10 power module failures occurred annually when the 35kV switchgear was switched back on after maintenance due to the aforementioned voltage transients and transformer inrush currents, severely impacting power generation availability.

[0043] To verify the effectiveness of the present invention, three adjacent turbines with similar operating conditions were selected as experimental subjects in an offshore wind farm. The first turbine was equipped with the adaptive intelligent protection system described in this invention; the second turbine retained its original protection scheme, which used two parallel fixed-delay time relays with a delay set to 1.0 second, serving as comparative example A; the third turbine adopted another common protection scheme in the industry, which used a metal oxide varistor with a large current capacity connected in parallel at the power module input terminal, serving as comparative example B.

[0044] The protection system of this invention installed in the first unit has the following specific configuration: The voltage acquisition module employs an isolation transformer coupling circuit with a turns ratio of 400:1, coupled with a 12-bit resolution analog-to-digital converter with a sampling frequency of 2kHz, resulting in a preset sampling period Ts of 500 microseconds. The control unit utilizes an embedded processor based on an ARM Cortex-M7 core, integrating a zero-crossing detection module. First derivative impact slope threshold. and second derivative impact curvature threshold The settings are calculated and set using preset adjustment coefficients and unit rated parameters, where the rated phase voltage peak value Vp is taken as... The power grid frequency is f=50Hz, and the setting factor is... Take 3.0, Take 1.5; substitute into the formula to calculate... Approximately 533.2 V / s, Approximately 83.8 kV / s²; the number of shielding cycles N0 is set to 3 power frequency cycles, or approximately 60 milliseconds. Smoothing factor. Take 0.2 as the safety factor. Set the initial reference settling time to 1.5. The preset time is 0.8 seconds based on the transformer capacity and the length of the collector line.

[0045] To obtain sufficient data samples, a 60-day special test was conducted. During the test, the collector lines of the three experimental units were periodically disconnected and energized, with a total of 30 closing tests performed. Before each closing test, all three units were in a shutdown state, the ring main unit remained closed, and the operators closed the 35kV switchgear, recording the voltage waveform at the input terminal of each unit's power module, the timing of protection actions, and whether the power module was damaged. The protection effects of each protection scheme under closing impact are compared in Table 1 below, and the control system startup delay time under each protection scheme is compared in Table 2 below. Each round included two consecutive closing tests, for a total of 15 rounds and 30 tests. The startup delay was defined as the time period from the moment the 35kV switchgear was closed until the power module normally outputs 24V. The parallel varistor scheme has no switching device isolation, and the delay is only caused by the charging of the rectifier bridge filter capacitor.

[0046] Table 1: Cumulative number of power module failures;

[0047] Table 2: Comparison of control system startup delay time under various protection schemes;

[0048] As can be clearly seen from the cumulative damage data in Table 1, after 30 rigorous closing impact tests, the power supply modules of the first unit using the adaptive intelligent protection method of this invention showed no damage, while the second unit using a fixed time-delay relay experienced a total of 4 failures, and the third unit using a parallel varistor experienced a total of 5 failures. The fundamental reason for this result is that the method of this invention no longer relies on a blind fixed waiting time. Although the 1.0-second delay set by the fixed time-delay relay can avoid transient impacts in most conventional operating conditions, in a few operating conditions where the equivalent impedance of the power grid decreases due to changes in system operation, resulting in slower attenuation of transient oscillations, the voltage fluctuation amplitude at the end of the 1.0-second wait still exceeds the tolerance range of the power supply module, causing the second unit to experience protection failure in the 7th to 8th and 13th to 14th test rounds. The parallel varistor solution is completely unsuitable for adapting to the duration of transients. Although its clamping voltage is set at around 800V, it cannot filter out the high-frequency components and DC bus overshoot caused by inrush current during the impact process. As a result, the rectifier bridge and filter capacitor of the power module of the third unit are continuously subjected to current thermal stress impact and are gradually damaged.

[0049] As shown in Table 2, the average startup delay of the method of this invention is only 0.603 seconds, significantly better than the fixed 1.0 second delay of the fixed delay relay. The control system can complete startup on average about 0.4 seconds earlier after each closing. This speed stems from the fact that this invention does not preset a fixed wait time, but instead accurately captures the true moment of transient end by real-time detection of the derivative value of voltage fluctuations after the shielding phase. Once the stability criterion is met, a conduction command is immediately issued. Although the parallel varistor scheme has a startup delay of only about 0.042 seconds, almost zero wait time, this comes at the cost of sacrificing protection effectiveness—its power module input is always exposed to the surge. As shown in Table 1, this instantaneous conduction is not safe. Therefore, this invention achieves an optimal balance between protection reliability and rapid startup that traditional technologies cannot achieve.

[0050] like Figure 3As shown in the figure, the statistical distribution of the start-up delay time from the moment the 35kV switchgear is closed to the normal output of 24V voltage from the power module is displayed for the first unit using the method of this invention, the second unit using a fixed time-delay relay, and the third unit using a parallel varistor. It can be clearly seen from the figure that the delay value of the fixed time-delay relay scheme is constant at 1.0 second, without any fluctuation, which precisely reflects its rigid control. The parallel varistor scheme has almost no delay, with a median of only about 0.041 seconds, but at the cost of a high failure rate as shown in Table 1. The start-up delay of the method of this invention is approximately 0.59 seconds, significantly lower than the fixed time-delay scheme, creating conditions for the rapid commissioning of the control system. More importantly, the enclosure height of the method of this invention is moderate, the range of wires is reasonable, and there are no outliers, indicating that its decision-making under 30 different transient conditions maintains consistent stability and reliability, demonstrating both adaptive adjustment capability for different impact durations and avoiding premature conduction due to misjudgment.

[0051] like Figure 4 As shown, the start-up delay data from 30 experiments on the first unit were processed using a sliding window with a window width of 5. The trajectory curve of the moving average was plotted, and the upper and lower bounds of the 95% confidence interval were calculated based on the t-distribution, represented by the light blue band. The red horizontal line in the figure represents the constant 1.0-second delay of the fixed-delay scheme as a comparison benchmark. The trend of the moving average curve clearly reveals the self-learning characteristics of this invention: in the early stages of the experiment, the moving average fluctuated around 0.63 seconds, and the confidence interval was relatively wide, reflecting that the system was conservative due to the initial preset benchmark stabilization time. As the number of successful turn-on operations increased, the exponentially weighted moving average mechanism continuously incorporated historical effective stabilization times into the calculation, optimizing the benchmark stabilization time. The moving average curve showed a significant downward trend and gradually converged to around 0.575 seconds, while the confidence interval also narrowed significantly. This indicates that the uncertainty of the protection system's prediction of the current power grid transient characteristics rapidly decreased, and the decision-making accuracy and stability continuously improved, enabling it to autonomously evolve to a better turn-on timing. The fixed-delay scheme, on the other hand, completely lacks this adaptive optimization capability.

[0052] Example 3: like Figure 5 As shown, embodiments of the present invention also provide a computer device. Figure 5Taking a single processor 10 as an example, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other using different buses and can be mounted on a common motherboard or otherwise installed as needed. The processor can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices, such as display devices coupled to the interface. In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations, for example, as a server array, a group of blade servers, or a multiprocessor system.

[0053] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0054] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0055] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0056] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0057] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0058] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

Claims

1. A protection method for a wind turbine generator power module, applied in a protection system comprising a control unit, a voltage acquisition module, and at least one time-delayed on-off switch, characterized in that, The method includes the following steps: The voltage acquisition module acquires the instantaneous voltage value sequence at the input terminal of the power module at a preset sampling period, denoted as v(k), where k is the sampling sequence number and k is a natural number; By performing first-order and second-order difference operations on the instantaneous voltage value sequence through the control unit, a first-order derivative sequence characterizing the degree of voltage fluctuation is obtained. and second derivative sequence ; The control unit determines whether the absolute value of the first derivative sequence is continuously less than the first derivative impact slope threshold and whether the absolute value of the second derivative sequence is continuously less than the second derivative impact curvature threshold within a continuous confirmation time window. When the determination is correct, the control unit generates a conduction command, drives the delayed conduction switch to close, and turns on the power supply circuit of the power module.

2. The method for protecting the power module of a wind turbine generator set according to claim 1, characterized in that, Time-domain fluctuation characteristic parameters include the sequence of first derivatives of the instantaneous voltage values. and second derivative sequence ; The time-domain fluctuation characteristic parameters characterizing the severity of voltage transient fluctuations are calculated, specifically including: By performing difference operations on the instantaneous voltage value sequence, the first-order derivative sequence and the second-order derivative sequence are obtained, and the calculation formulas are as follows: ; ; Where v(k) represents the instantaneous voltage value at the k-th sampling time, v(k-1) represents the instantaneous voltage value at the (k-1)-th sampling time, and v(k-2) represents the instantaneous voltage value at the (k-2)-th sampling time. Indicates the preset sampling period.

3. The method for protecting the power module of a wind turbine generator set according to claim 2, characterized in that, The stability criterion is specifically as follows: Within the continuous confirmation time window, the absolute value of the first derivative sequence All are less than the preset first derivative impact slope threshold. And the absolute value of the second derivative sequence All are less than the preset second derivative impact curvature threshold. The first derivative impact slope threshold and second derivative impact curvature threshold All are based on the rated phase voltage peak value of the wind turbine generator set. Set the frequency f with the power grid frequency and satisfy the following: ; ; in, This represents the rated peak phase voltage at the input of the power module, where f represents the mains frequency. and It is a tuning coefficient greater than 1.

4. The method for protecting the power module of a wind turbine generator set according to claim 3, characterized in that, Also includes: Record the time period from the end of the shielding phase to the moment when the control unit generates the conduction command each time the conduction is successfully turned on, and use it as the effective stabilization time t(n) for the nth time; The initial reference stabilization time preset during the first run. Substituting the effective settling time t(1) of the first recorded value into the exponentially weighted moving average formula, we obtain the baseline settling time after the first processing. ; After each successful conduction, the baseline settling time is iteratively updated using an exponentially weighted moving average method. The update formula is as follows: ; in, This represents the baseline settling time after the nth processing. Let t(n) represent the baseline settling time after the (n-1)th processing, and t(n) represent the effective settling time of the nth record. Represents the smoothing factor and ; Based on the benchmark settling time The length of the confirmation time window for the next protection startup is dynamically updated, specifically including: utilizing the reference settling time. and safety factor Calculate the updated confirmation time window length. The calculation formula is: ; in, The preset safety factor is greater than 1.

5. A wind turbine generator power module protection system, used to execute the wind turbine generator power module protection method according to any one of claims 1 to 4, characterized in that, include: The voltage acquisition module is used to acquire the instantaneous voltage value sequence v(k) at the input terminal of the power supply module in real time. The control unit is connected to the voltage acquisition module. The control unit is equipped with a memory, a processor, and a program stored in the memory and executed by the processor. When the program is executed, it calculates the time-domain fluctuation characteristic parameters, judges the stability criteria, and generates the turn-on command. The first delay switch and the second delay switch have their input terminals connected in parallel to the positive terminal of the power supply circuit, their output terminals connected in parallel to the positive input terminal of the power module, and their control terminals connected to the first output port and the second output port of the control unit, respectively. The control unit sends conduction commands to the first delayed conduction switch and the second delayed conduction switch through the first output port and the second output port respectively, so as to synchronously drive them to close. A manual bypass switch is provided, with one end connected to the positive terminal of the power supply circuit and the other end connected to the positive input terminal of the power module. The manual bypass switch is configured to be manually closed by the operator to forcibly connect the power supply circuit of the power module when the control unit fails.

6. The wind turbine generator power module protection system according to claim 5, characterized in that, The control unit is also equipped with a zero-crossing detection module, which is used to detect the zero-crossing point of the instantaneous voltage value and generate a zero-crossing count. The control unit is used to start the calculation of the time-domain fluctuation characteristic parameter after the zero-crossing count reaches a preset number of shielding cycles N0.

7. A wind turbine generator power module protection system according to claim 5, characterized in that, The control unit's program also includes a threshold adaptive adjustment module, which is used to adjust the first derivative sequence based on historical data. The maximum value and the second derivative sequence The maximum value of the first derivative impacts the slope threshold within a preset tuning period. and the second derivative impact curvature threshold Make corrections.

8. A wind turbine generator power module protection system according to claim 7, characterized in that, Impact slope threshold of the first derivative and second derivative impact curvature threshold The corrections will be made in the following manner: The peak value of the absolute value of the first derivative sequence before the end of the confirmation time window during the most recent M closing processes is denoted as . And the peak value of the absolute value of the second derivative sequence, denoted as ,in ; Calculate the sequence respectively and The exponentially weighted moving average yields the average first-order peak value. and average second-order peak ; After processing each set of M peaks, the average first-order peak value is... and the average second-order peak value Perform iterative updates: ; ; Where j is the iteration number, The preset threshold smoothing factor and , and These are the average first-order peak value and the average second-order peak value after the previous round of processing, respectively. Impact the slope threshold with the first derivative Updated to The second derivative impacts the curvature threshold. Updated to ,in and These are preset margin coefficients, all of which are greater than 1.

9. A computer device, characterized in that, It includes a memory and a processor, which are interconnected and communicate with each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the wind turbine generator power module protection method according to any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the wind turbine generator power module protection method according to any one of claims 1 to 4.