A multi-stage suppression device for high voltage ride-through transient overvoltage in direct-drive wind turbine units

Abstract

This utility model belongs to the field of wind power generation technology, specifically a multi-stage collaborative suppression device for high voltage ride-through (HVRT) transient overvoltage in direct-drive permanent magnet wind turbines. It includes a three-stage hardware topology in series: a first-stage crowbar circuit comprising an anti-parallel thyristor module connected in series with a current-limiting inductor, and paralleled between the positive and negative terminals of the DC bus; a second-stage energy-dissipating branch comprising IGBT switching devices driving a chromium-aluminum alloy resistor array mounted on an air-cooled heat dissipation substrate; a third-stage unloading circuit comprising a mechanical contactor controlling a water-cooled resistor box, with a built-in copper water channel within the resistor box, the resistor plates inside the resistor box being tightly fitted to the copper water channel, and heat dissipation through an external circulating water cooling system; and an FPGA controller that compares the DC bus voltage with the third-stage threshold in real time, generates a trigger signal, and adjusts the duty cycle of the second-stage PWM. This device solves the problem of DC bus overvoltage caused by grid faults.

Description

Technical Field

[0001] This utility model belongs to the field of wind power generation technology, specifically a multi-stage coordinated suppression device for high voltage ride-through (HVRT) transient overvoltage in direct-drive permanent magnet wind turbines. Background Technology

[0002] Direct-drive permanent magnet generator sets use a full-power converter (generator-side converter + grid-side converter) to connect the generator and the grid. This is the core of energy conversion and grid interaction, and also the most direct bearer of transient high voltage. Specific impacts include: DC bus overvoltage damage. The generator-side converter of the direct-drive unit rectifies the AC output from the generator into DC (DC bus), and then inverts it into AC synchronized with the grid through the grid-side converter. When the grid voltage transiently increases, the grid-side converter's ability to deliver power to the grid is limited (power is difficult to deliver when the grid voltage is higher than the DC bus voltage). However, because the permanent magnet generator's speed changes slowly (due to the large inertia of the wind turbine), it continues to inject energy into the DC bus through the generator-side converter, causing a sudden rise in the DC bus voltage. If the transient voltage is too high (e.g., exceeding 1.2-1.5 times the rated withstand voltage of the DC bus capacitor), it will directly lead to: DC bus capacitor (energy storage core) breakdown (electrolyte ionization, plate breakdown) or a drastically shortened lifespan (accelerated aging of the insulation medium); excessive reverse voltage on power devices such as IGBTs on the generator side / grid side, exceeding their avalanche breakdown voltage (e.g., a 1700V device subjected to over 2000V), resulting in permanent damage to the devices (short circuit or open circuit); excessively high voltage will damage electrical equipment on the DC bus, even causing severe damage; excessively high voltage will increase the current on the DC bus, thereby increasing the line load and placing a huge burden on the power system. When the DC bus voltage exceeds the rated value, the operating state of the power system will become unstable, and may even lead to system collapse. Therefore, we must pay attention to the impact of grid faults on DC bus overvoltage and take corresponding measures to prevent it from happening.

[0003] However, there are two major technical bottlenecks in the protection measures for direct-drive wind turbines when the grid voltage suddenly rises:

[0004] 1. Insufficient single-stage protection: After the traditional crowbar circuit is activated, there is still a 20-30% residual overvoltage (>1300V).

[0005] 2. Response delay: IGBT turn-off delay causes overvoltage suppression hysteresis >5ms, threatening converter safety.

[0006] Prior art document CN112636385A discloses a microgrid control method with multi-energy flow complementary control, including: S1, local consumption of distributed energy resources in the microgrid; S2, peak shaving and valley filling of the AC / DC microgrid, and charge / discharge control according to the planned curve mode, using the charging and discharging periods of the planned curve and the corresponding charging and discharging power values ​​for each charging and discharging period; S3, free switching between grid-connected and off-grid states of the AC / DC microgrid. This utility model realizes coordinated control of "source, grid, load, and storage" in substations, data centers, and large-capacity integrated energy stations, utilizes big data mining technology to adjust operating strategies in real time, fully leverages the advantages of source-grid-load-storage complementarity, and improves microgrid reliability and overall grid benefits. However, its unloading circuit only consumes energy at a single stage and cannot cope with continuous transient overvoltage; there are unresolved timing coordination issues in multi-level modules. Utility Model Content

[0007] The purpose of this invention is to provide a multi-stage coordinated suppression device for high voltage ride-through transient overvoltage in direct-drive permanent magnet wind turbines, which addresses the problems existing in the prior art and solves the problem of DC bus overvoltage caused by grid faults.

[0008] The technical solution of this utility model is:

[0009] A multi-stage suppression device for high-voltage ride-through transient overvoltage in direct-drive wind turbine generators, comprising:

[0010] Serial three-level hardware topology:

[0011] The first-stage crowbar circuit consists of an anti-parallel thyristor module connected in series with a current-limiting inductor, and connected in parallel between the positive and negative terminals of the DC bus.

[0012] Secondary power consumption branch: includes switching device IGBT, driving chromium-aluminum alloy resistor array, which is mounted on air-cooled heat dissipation substrate;

[0013] Three-stage unloading circuit: including mechanical contactor control of water-cooled resistor box, the resistor box has built-in copper pipe water channel, the resistors inside the resistor box are tightly attached to the copper pipe water channel, and heat is dissipated through an external circulating water cooling system.

[0014] FPGA controller: compares the DC bus voltage with the three-level threshold in real time, generates a trigger signal, and adjusts the duty cycle of the two-level PWM.

[0015] Specifically, the trigger threshold of the first-stage crowbar circuit is 1150V±2%, the thyristor module adopts a 1700V / 200A double anti-parallel structure, and the current-limiting inductor value is 2mH±5%.

[0016] The PWM duty cycle of the secondary energy-consuming branch increases linearly with voltage.

[0017] Specifically, the IGBT model is SKM300GB176D, and the chromium-aluminum alloy resistor array is a resistor array composed of 6 chromium-aluminum alloy resistors, each with a resistance of 20Ω / 5kW, connected in parallel.

[0018] Specifically, the trigger threshold of the three-stage unloading circuit is 1350V±2%, the rated current of the mechanical contactor is 400A, the resistance of the water-cooled resistor box is 0.5Ω, and the copper pipe waterway interface is a G3 / 4 quick-connect type.

[0019] Specifically, the FPGA controller implements multi-level timing control: Level 1 trigger delay < 500μs; Level 2 action interval 1ms ± 0.1ms; Level 3 action interval 1ms ± 0.1ms.

[0020] Specifically, the device is integrated between the positive and negative poles of the DC bus of the grid-side converter of the direct-drive wind turbine, and the series three-level hardware topology is connected to the bus through a fast-acting fuse.

[0021] The beneficial effects of this invention are as follows: This device overcomes the limitations of traditional overvoltage suppression devices through an innovative combination of a three-stage progressive hardware structure and FPGA high-precision timing control: a three-stage coordinated system of crowbar (rapid current discharge), energy-consuming branch (precise control), and unloading (backup protection); FPGA achieves nanosecond-level response, with duty cycle linearly adjusted with voltage; and a combination of air cooling and water cooling solves the problem of high-energy dissipation temperature rise. In the application of the Guodian Longyuan 2MW unit, the peak voltage of the DC bus is controlled within 1280V under a 130% voltage surge in the grid, reducing the converter failure rate by 90%. It can suppress transient overvoltages of direct-drive wind turbine units within a safe range within 200ms, reducing the overvoltage peak by 40% compared to the traditional single-stage crowbar solution, and controlling the temperature rise of resistive elements within 80K, significantly improving the unit's high-voltage ride-through capability. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the structure of this utility model. Detailed Implementation

[0023] The technical solution of this utility model will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0024] Example 1

[0025] like Figure 1 The diagram shown is a structural schematic of a multi-stage suppression device for high-voltage ride-through transient overvoltage in a direct-drive wind turbine provided in this embodiment, including...

[0026] Serial three-level hardware topology:

[0027] The first-stage crowbar circuit includes a series connection of an anti-parallel thyristor module 1011 and a current-limiting inductor 1012, which are connected in parallel between the positive and negative terminals of the DC bus.

[0028] Secondary power consumption branch: includes switching device IGBT1021, driving chromium-aluminum alloy resistor array 1022, chromium-aluminum alloy resistor array mounted on air-cooled heat dissipation substrate 1023;

[0029] Three-stage unloading circuit: including mechanical contactor 1031 controlling water-cooled resistor box 1032, resistor box with built-in copper pipe water channel 1033, resistor sheet inside resistor box is tightly attached to copper pipe water channel, heat dissipation through external circulating water cooling system.

[0030] FPGA controller: compares the DC bus voltage with the three-level threshold in real time, generates a trigger signal, and adjusts the duty cycle of the two-level PWM.

[0031] Core components and function of the primary crowbar circuit

[0032] 1. The anti-parallel thyristor module 1011 adopts a 1700V / 200A double anti-parallel structure (two thyristors connected in opposite parallel directions). Its function is: the double anti-parallel connection enables the thyristors to carry bidirectional current (suitable for AC scenarios or DC scenarios with bidirectional voltage), ensuring that it can conduct regardless of the current direction; the 1700V rated withstand voltage is higher than the trigger threshold (1150V), ensuring that the thyristors can withstand overvoltage before conducting in case of a fault and will not be broken down; the 200A rated current meets the current carrying requirements in case of a fault.

[0033] 2. The current-limiting inductor 1012 with an inductance of 2mH±5% is connected in series in the thyristor module. Its function is to limit the peak current and rate of change of current (di / dt) at the moment the thyristor is turned on, so as to avoid excessive inrush current from damaging the thyristor (the thyristor is sensitive to di / dt, and exceeding the rated value will cause mis-conduction or burnout); and stabilize the fault current to make the energy release process smoother.

[0034] 3. The trigger threshold is 1150V±2%. This is the action criterion for the crowbar circuit. When the voltage at the monitoring point (such as the DC bus or AC side) exceeds 1150V±2%, the thyristor is triggered to conduct. This threshold is higher than the normal operating voltage of the system (such as the normal voltage of the DC bus of a wind power converter, which is usually 800-1000V), ensuring that the crowbar does not malfunction during normal operation.

[0035] Working principle

[0036] 1. Normal state

[0037] The thyristor module is in the off state (no trigger signal), the crowbar circuit is not connected to the main circuit, and it does not affect the normal operation of the system.

[0038] 2. Fault Trigger (Overvoltage)

[0039] When a system fault occurs (such as a short circuit in the power grid causing DC side overvoltage, or a converter runaway causing a sudden voltage rise), and the voltage at the monitoring point exceeds 1150V±2%, the trigger circuit immediately sends a trigger pulse to the thyristor module to turn on the anti-parallel thyristors (regardless of the voltage direction, one thyristor is always forward biased and turned on).

[0040] 3. Energy Release and Protection

[0041] After the thyristor is turned on, the crowbar circuit is connected to the main circuit, forming an energy release path (usually in conjunction with energy storage components in the main circuit, such as capacitors and inductors):

[0042] The current-limiting inductor (2mH) limits the peak current at the moment of conduction (to avoid exceeding the thyristor's rated current of 200A) and di / dt (to protect the thyristor from impact damage).

[0043] The conducting thyristor releases excess energy (such as the electrical energy stored in the capacitor) through a preset path (e.g., by connecting to a power-consuming resistor or guiding it to the grounding loop), quickly reducing the main circuit voltage to a safe range and protecting the core components in the main circuit (such as IGBTs, electrolytic capacitors, etc.) from being overvoltage-damaged.

[0044] 4. After the fault is cleared

[0045] When the system voltage drops to a safe value (below the trigger threshold), the trigger signal is withdrawn, the thyristor is turned off, and the crowbar circuit returns to standby mode.

[0046] The first-stage crowbar circuit provided by this utility model uses the logic of "voltage monitoring - threshold triggering - thyristor conduction - inductor current limiting - energy release" to act quickly in the event of a system overvoltage fault, which is equivalent to the "safety valve" of the main circuit. It protects the core components by sacrificing its own path (short-term conduction).

[0047] In this embodiment, the secondary energy-consuming branch is a key circuit used for the precise consumption of excess energy. Working in conjunction with the primary crowbar circuit, it forms a "tiered energy consumption" mechanism—the primary branch handles sudden large energy surges, while the secondary branch is responsible for the continuous and controllable consumption of residual energy or small amounts of excess energy, preventing the system from frequently entering emergency protection states. Based on its composition, the specific principle is as follows:

[0048] Core components and functions

[0049] 1. The switching device is IGBT1021. IGBT (Insulated Gate Bipolar Transistor) is the core control switch of the branch. Compared with the thyristor in the first-stage crowbar, its advantages are: high-frequency controllability: it can be quickly turned on / off by the gate voltage signal (response time in microseconds), supports pulse width modulation (PWM) control, and can accurately adjust the conduction time and duty cycle to achieve flexible adjustment of power consumption; self-turn-off capability: it can turn off without relying on the current crossing zero, which is suitable for handling continuous or non-periodic energy surplus scenarios and avoids the limitation of the first-stage crowbar that "once it is turned on, it must wait for the energy to be exhausted before it can be turned off".

[0050] 2. Chromium-aluminum alloy resistor array 1022

[0051] A resistor array, composed of multiple chromium-aluminum alloy resistors, is the core component for energy consumption.

[0052] Material Characteristics: Chromium-aluminum alloys (such as CrAlSi) are characterized by high temperature resistance (long-term operating temperature can reach 200-300℃), high power density, and small temperature coefficient of resistance (resistance value changes little with temperature), making them suitable as high-power energy-consuming resistors. Resistor Array Design: Multiple resistors are connected in series / parallel to form an "array," which allows for flexible adjustment of the total resistance value and rated power (for example, when consuming 10kW of energy, the power can be shared by multiple resistors to avoid overloading a single resistor), ensuring stable resistance and uniform heat generation during energy consumption. The air-cooled heat dissipation substrate 1023 serves as the heat dissipation carrier for the resistor array. Its functions are: Heat Conduction: The substrate is usually made of a metal with high thermal conductivity (such as aluminum alloy). The resistor array is tightly mounted on the surface of the substrate, and the heat generated by the resistors is transferred to the substrate through thermal conduction; Forced Heat Dissipation: The substrate, in conjunction with an air-cooling system (such as a fan), dissipates heat into the environment through air convection, preventing the resistors from burning out due to long-term heating (electrical energy is converted into heat energy during energy consumption) exceeding the tolerance temperature (such as the maximum allowable operating temperature of chromium-aluminum alloy resistors is usually 250℃), thus ensuring the long-term stable operation of the branch.

[0053] Working principle:

[0054] 1. Normal state

[0055] When the IGBT is in the off state (no gate drive signal), the secondary power-consuming branch is disconnected from the main circuit, does not participate in system operation, and does not consume energy.

[0056] 2. Triggering conditions (when energy consumption is required)

[0057] The control unit (such as MCU or DSP) will trigger the secondary energy-consuming branch to work when the system is in the following situations: after the primary crowbar circuit is activated, the main circuit still has some residual energy (which needs to be further consumed to restore normal voltage); the system has a small amount of excess power (such as grid fluctuations causing the input power to be greater than the output power, and the DC side voltage to rise slowly but not reach the primary crowbar trigger threshold); the feedback energy generated by sudden load changes (such as motor deceleration) needs to be consumed.

[0058] 3. Controllable energy consumption process

[0059] The control unit sends a drive signal (applying a positive voltage to the gate) to the IGBT, turning it on. At this time, excess energy in the main circuit (such as the electrical energy stored in the DC bus capacitor or the rectified AC feedback energy) flows through the IGBT into the chromium-aluminum alloy resistor array. The resistor array converts electrical energy into heat energy (following Joule's law: Q=I). 2 Rt); The control unit can precisely control the power consumption by adjusting the duty cycle of the IGBT (such as PWM control) (the higher the duty cycle, the more energy is consumed per unit time), to achieve "on-demand power consumption" and avoid excessive energy consumption that could cause a sudden drop in system voltage.

[0060] 4. Heat dissipation guarantee

[0061] The heat generated by the resistor array is quickly conducted through the air-cooled heat dissipation substrate and dissipated by the fan, maintaining the resistor temperature within a safe range (the heat dissipation capacity is usually designed to cover the maximum power dissipation of the resistor array), ensuring the long-term stable operation of the branch.

[0062] 5. Stop consuming energy.

[0063] When the system energy is restored to balance (e.g., the voltage drops to the normal range), the control unit cancels the IGBT drive signal, the IGBT is turned off, the branch stops working, and the system returns to normal operation.

[0064] The secondary energy-consuming branch is a "supplement and refinement" of the primary crowbar: the primary crowbar (thyristor) has a fast response and can withstand large currents, making it suitable for handling millisecond-level sudden overvoltage / large energy surges (such as power grid short circuits), but its energy consumption is uncontrollable (once it conducts, it needs to release all excess energy); the secondary branch (IGBT + resistor array) has a more refined response and controllable energy consumption, making it suitable for handling continuous or small-amplitude excess energy, avoiding system fluctuations caused by frequent movements of the primary crowbar, and improving overall operational stability.

[0065] The secondary energy-consuming branch achieves refined and controllable consumption of excess energy through the combination of "IGBT controllable switch + high-power resistor array + forced heat dissipation". It is the core link of "flexible protection" in power electronic system. It forms a hierarchical collaboration with the primary crowbar, which not only ensures system safety but also improves operating efficiency.

[0066] In this embodiment, the three-stage unloading circuit serves as the "ultimate protection line" against severe and sustained overvoltage. Its trigger threshold (1350V±2%) is higher than that of the first-stage crowbar (1150V±2%) and the second-stage energy-dissipating branch. It is primarily used when the first two stages of protection have not completely eliminated the overvoltage, or when the system experiences extreme energy excess (such as a prolonged grid failure preventing energy transmission). Through high-power unloading, it prevents damage to the main circuit (such as the DC bus and capacitors) due to overvoltage. The specific principle, based on its composition, is as follows:

[0067] Core components and functions

[0068] 1. The mechanical contactor 1031, acting as a "switch" in a three-stage unloading circuit, employs a mechanical contact structure with a rated current of 400A. Its core functions are: High current carrying capacity: The contacts of the mechanical contactor (usually made of silver alloy) can withstand short-term high current surges (far exceeding the rated 400A, with peak currents reaching over 1000A), making it suitable for conducting high-current unloading under severe overvoltage conditions; Long-term stable disconnection / closing: Under normal conditions, the contacts reliably disconnect (high insulation resistance), preventing accidental connection with the main circuit; Upon triggering, the electromagnetic coil is energized, driving the contacts to close and connecting the unloading circuit, with low contact resistance after closure (usually <10mΩ), reducing its own power consumption; Handling continuous operating scenarios: Compared to single-stage thyristors (which require zero-current turn-off after conduction) and two-stage IGBTs (high-frequency switches are prone to overheating), the mechanical contactor is suitable for long-term conduction (provided heat dissipation allows), capable of handling energy unloading for several seconds to several minutes (such as when the power grid has not been restored).

[0069] 2. Water-cooled resistance box 1032

[0070] The enclosure, composed of high-power resistive elements, is the "energy consumption core" of the three-stage unloading system. With a resistance of 0.5Ω, its key features are: high power unloading capability: According to Ohm's law, when the trigger voltage is 1350V, the current through the resistor is I=U / R=1350V / 0.5Ω=2700A (short-time peak value), corresponding to an unloading power P=U 2 / R≈3.65MW (instantaneous power), which is much higher than that of the secondary energy-consuming branch (usually in the kW range), and can quickly release extreme excess energy; resistance characteristics: the internal resistors are usually made of nickel-chromium alloy or iron-chromium-aluminum alloy (high temperature resistant and oxidation resistant), and through series / parallel combination design, the resistance value is ensured to be stable under high current (avoiding sudden change in resistance value due to heat generation), while dispersing heat (the power of a single resistor is limited, and overload is avoided by sharing the power of multiple resistors).

[0071] 3. The copper water pipe 1033 serves as the "heat dissipation blood vessels" of the water-cooled resistance box, employing a G3 / 4 quick-connect interface. Its function is: a highly efficient heat dissipation carrier: when the resistance box is operating (electrical energy → heat energy), it generates a huge amount of heat (3.65MW instantaneous power corresponds to approximately 3.65×10⁻⁶ heat generation per second). 6The heat dissipation capacity far exceeds that of air cooling. It requires circulating coolant (such as deionized water + ethylene glycol) through copper pipes and water channels to quickly absorb heat by utilizing the high specific heat capacity of the liquid (the specific heat capacity of water is more than 4 times that of air). Installation and circulation guarantee: The high thermal conductivity of copper pipes (about 380W / (m·K)) ensures that the heat of the resistor is quickly transferred to the coolant. The G3 / 4 quick-connect interface (nominal diameter 26.9mm) has a large flow rate and can be connected to an external water cooling system (such as a cooling tower or water pump) to achieve coolant circulation and heat dissipation, and prevent the resistor from burning out due to excessive temperature (above 500℃).

[0072] Working principle

[0073] 1. Normal state

[0074] When the coil of the mechanical contactor (1031) is de-energized, the contacts open, and the three-stage unloading circuit is completely isolated from the main circuit (such as the DC bus), without consuming energy or affecting the normal operation of the system.

[0075] 2. Triggering conditions (when strong unloading is required)

[0076] When the system encounters the following extreme situations, the control unit (such as PLC, DSP) triggers the three-stage unloading circuit: After the first-stage crowbar and the second-stage energy-consuming branch are activated, the DC side voltage continues to rise, exceeding 1350V±2% (i.e., 1323-1377V); the power grid experiences a prolonged fault (such as a power outage exceeding 10 seconds), the main circuit cannot supply energy to the grid, while new energy power generation (such as photovoltaic, wind power) or energy storage batteries continue to input energy, causing the voltage to rise continuously; the main circuit capacitor stores excessive energy (exceeding the safe withstand voltage value, such as when the capacitor's rated withstand voltage is 1500V, it needs to be unloaded in advance at 1350V to avoid breakdown).

[0077] 3. High-power unloading process

[0078] The control unit supplies power to the mechanical contactor coil, and the electromagnetic force drives the contacts to close. At this time, excess electrical energy in the main circuit (such as the electrical energy stored in the DC bus capacitor and the continuously input new energy) flows into the water-cooled resistance tank through the contactor. The resistance tank converts the electrical energy into heat energy (Joule's law: Q=U). 2 / R·t), quickly reduce the main circuit voltage; due to the extremely large unloading power (nearly 4MW instantaneously), the heat generated by the resistor box is quickly carried away by the circulating coolant in the copper pipe water channel (the temperature difference of the coolant inlet is usually designed to be 5-10℃, and the flow rate needs to match the heat dissipation requirements, such as tens of tons per hour), ensuring that the resistor temperature is controlled below 300℃ (far below the melting point of alloy resistors).

[0079] 4. Continuous unloading and resetting

[0080] If the fault persists (e.g., the power grid has not been restored), the mechanical contactor remains closed, and the resistor box continues to unload until: the system voltage drops to a safe range (e.g., below 1200V), the control unit cuts off the power supply to the contactor coil, the contacts open, and unloading stops; the fault is cleared (e.g., the power grid is restored), and the system re-enters normal operating mode.

[0081] The three-stage unloading circuit is the final link in the "graded protection" system. Its coordination with the first two stages reflects the "gradient response" principle: Stage 1 crowbar (1150V trigger): fastest response (microsecond level), handling sudden short-term overvoltages (such as power grid surges), rapidly dissipating peak energy through thyristors and inductors, but cannot continuously unload (the thyristor needs to exhaust its energy before turning off after being turned on); Stage 2 energy dissipation branch (no fixed threshold, triggered according to system state): fine response (millisecond level), handling continuous small-amplitude overvoltages (such as power fluctuations) through IGBTs and air-cooled resistors, and can precisely adjust energy consumption power through PWM control, but has limited current carrying capacity (usually <200A).

[0082] Three-stage unloading (1350V trigger): The response is slightly slower (tens of milliseconds due to the delay in mechanical contactor action), but it can carry ultra-high current (400A rated, 2700A for short time) and handle megawatt-level continuous overvoltage, serving as the system's "last safety barrier".

[0083] The three-stage unloading circuit uses a combination of "mechanical contactor (high current switching) + water-cooled resistor box (high power consumption) + copper pipe water channel (high efficiency heat dissipation)" to specifically deal with extreme and continuous overvoltage scenarios that the first two stages of protection cannot handle. Its core design is "high load capacity, strong heat dissipation, and long duration" to ensure the safety of key components (capacitors, power devices) in the main circuit by forcibly unloading when the system faces catastrophic risks.

[0084] Example 2

[0085] This embodiment describes the detailed implementation steps using overvoltage suppression of a 2MW direct-drive permanent magnet wind turbine as an example.

[0086] I. Preliminary Preparations and Hardware Selection

[0087] 1. Selection of core components

[0088] DC bus core parameters: rated voltage 900V, capacitor bank withstand voltage ≥1500V (with redundancy), total capacity is designed according to the unit power (2MW units are usually equipped with 1000-1500μF).

[0089] Three-stage circuit components: Stage 1: Anti-parallel thyristor module (1700V / 200A, such as Infineon T200Z17KOF) + current-limiting inductor (2mH±5%, ferrite core); Stage 2: IGBT (1700V / 300A, such as Semikron SKM300GB176D) + chromium-aluminum alloy resistor array (total power ≥50kW, single resistor 20Ω / 5kW in parallel) + air-cooled heat sink (heat dissipation capacity ≥60W / ℃); Stage 3: Mechanical contactor (400A / AC250V, such as Schneider LC1D400) + water-cooled resistor box (0.5Ω±5%, power ≥4MW) + copper pipe water channel (G3 / 4 quick-connect type, suitable for water cooling systems with a flow rate ≥50L / min). Sensing and Control: Hall voltage sensor (±1% accuracy, such as LEM LV25-P, range 0-1600V), FPGA controller (such as Xilinx Artix-7, main frequency ≥100MHz, supports high-speed AD sampling).

[0090] II. Hardware Connection and Installation

[0091] 1. Three-stage circuit connected to DC bus

[0092] Parallel connection: Connect the positive and negative terminals of the main circuit inputs of the first stage (thyristor + inductor), second stage (IGBT + resistor array), and third stage (contactor + water-cooled resistor box) in parallel to the positive and negative terminals of the DC bus (close to the capacitor bank output to reduce the influence of line impedance). Protection measures: Connect fast-acting fuses in series for each stage of the circuit (250A for the first stage, 350A for the second stage, and 500A for the third stage) to prevent damage to the bus in case of a short circuit.

[0093] 2. Sensing and control link connection

[0094] Voltage signal acquisition: The Hall sensor is connected in series between the negative terminal of the DC bus and ground (or in parallel between the positive and negative terminals), and outputs an analog signal (e.g., 0-10V corresponds to 0-1600V) which is connected to the 16-bit AD sampling channel of the FPGA (sampling rate ≥100kHz to ensure capture of rapid voltage changes).

[0095] Drive signal connections: The digital output of the FPGA is connected to the trigger board of the first-stage thyristor via optocoupler isolation (such as HCPL-3120) (providing 10V / 100mA trigger pulse); the PWM output of the FPGA is connected to the gate of the second-stage IGBT via an IGBT driver chip (such as 2SC0435T) (controlling turn-on / turn-off); the relay output of the FPGA is connected to the coil of the third-stage mechanical contactor (controlling contact engagement / disengagement).

[0096] 3. Installation of the cooling system

[0097] The secondary air-cooled base plate is fixed inside the unit's heat dissipation duct, and the fan power supply is linked with the IGBT drive signal (the fan starts when the IGBT is working); the copper tube water channel of the tertiary water-cooled resistor box is connected to the unit's water-cooling circulation system through a G3 / 4 quick connector (a flow meter is installed at the inlet to ensure a flow rate ≥ 50L / min).

[0098] III. Sensor Calibration and Signal Acquisition Configuration

[0099] 1. Hall sensor calibration

[0100] Calibration tools: DC regulated power supply (0-1600V, accuracy 0.1%), multimeter (accuracy 0.05%).

[0101] Calibration steps:

[0102] 1. Disconnect the sensor from the busbar and connect it to the output terminal of the regulated power supply;

[0103] 2. Input standard voltages of 500V, 900V, and 1350V respectively, and record the analog voltage values ​​output by the sensor (e.g., 900V corresponds to 5.625V, since 0-1600V corresponds to 0-10V).

[0104] 3. Write a calibration algorithm (linear fitting) in the FPGA to convert the AD sampled values ​​(digital quantity) into the actual voltage (e.g., sampled value = 5.625V × 4096 / 10V = 2304 → corresponding to 900V), ensuring that the error is ≤ ±1%.

[0105] 2. FPGA sampling configuration: Configure the AD sampling channel to continuous sampling mode, collect voltage data every 10μs (100kHz sampling rate), and buffer the most recent 100 data (1ms window) for filtering (removing high-frequency noise).

[0106] IV. Control Logic Programming and Parameter Setting

[0107] The control logic is written based on FPGA to ensure that each stage of the circuit operates according to the designed thresholds and timing sequence. The process is as follows:

[0108] 1. Level 1 crowbar trigger logic

[0109] Threshold setting: Set the first-level trigger threshold in the FPGA to 1150V (corresponding AD sampling value = 1150V × 10V / 1600V × 4096 = 1150 × 2.56 = 2944).

[0110] Triggering condition: When the filtered bus voltage exceeds 1150V for three consecutive sampling points (30μs), a trigger pulse (100μs width) is immediately output to the thyristor module to turn it on.

[0111] 2. PWM control of the secondary energy-consuming branch

[0112] Duty cycle calculation: Linear relationship between set voltage and duty cycle:

[0113] Voltage ≤1250V: Duty cycle = 0 (no operation);

[0114] 1250V < Voltage < 1300V: Duty cycle = 2 × (Voltage - 1250V) (e.g., duty cycle = 50% at 1275V);

[0115] Voltage ≥1300V: Duty cycle = 100% (full power consumption).

[0116] Startup conditions: After the first stage is triggered, if the voltage still rises above 1250V, or if the voltage is directly detected to be ≥1250V (when the first stage is not triggered), the FPGA generates a PWM signal with a corresponding duty cycle (frequency 10kHz) to drive the IGBT to turn on.

[0117] 3. Level 3 unloading backup trigger

[0118] Threshold setting: Level 3 trigger threshold 1350V (corresponding AD sampling value = 1350 × 2.56 = 3456).

[0119] Triggering logic: When the voltage exceeds 1350V, or the voltage continues to rise to 1350V after the first or second stage operation, the FPGA outputs a high-level signal to the contactor coil (continuously energized until the voltage drops below 1200V), causing the contacts to close and connecting to the water-cooled resistor box.

[0120] 4. Collaboration Logic at Each Level

[0121] Priority: Level 3 > Level 2 > Level 1 (to prevent higher-level circuits from intervening in time when lower-level circuits cannot suppress the voltage); Turn-off condition: When the voltage drops below 950V (105% of the rated voltage), turn off Level 3 in sequence (contactor de-energized) → Level 2 (duty cycle 0) → Level 1 (stop trigger pulse, thyristor turns off naturally).

[0122] V. Fault Simulation Test

[0123] 1. Test Preparation

[0124] Equipment connection: Connect the unit to the power grid simulator (such as a regenerative power grid simulator with a capacity of ≥3MW) and set the rated voltage to 690V (corresponding to a DC bus rated voltage of 900V).

[0125] Monitoring instruments: Oscilloscope (bandwidth ≥ 100MHz, probes withstand voltage ≥ 2000V) connected to DC bus to record voltage waveform; data logger synchronously records trigger signals (first-level / second-level / third-level action time).

[0126] 2. Fault Scenario Setup

[0127] Grid simulator parameters: The voltage suddenly rises from the rated value to 130% (i.e., 690V×1.3=897V AC, corresponding to the theoretical value of DC bus ≈897×1.414≈1268V), lasting for 500ms, and then returns to normal.

[0128] 3. Testing Process

[0129] 1. Power on the unit and start it up to normal operating conditions (DC bus stabilizes at 900V).

[0130] 2. Start the power grid simulator and execute the voltage surge command;

[0131] 3. The oscilloscope and data logger synchronously record the bus voltage and the operation time of each stage of the circuit from 0ms to 500ms;

[0132] 4. Repeat the test 3 times and take the average value as the result.

[0133] VI. Data Collection and Effect Analysis

[0134] 1. Key Indicator Extraction

[0135] Peak voltage: Read the highest voltage value from the oscilloscope waveform (e.g., 1280V in the example);

[0136] Energy dissipation: The total energy consumption of each stage of the circuit is calculated using the formula (Stage 1: ∫U). 2 / R1×t, Level 2: ∫U 2 / R2×D×t, Level 3: ∫U 2 / R3×t, where R is the value of each resistor and D is the duty cycle);

[0137] Action response time: the time from when the voltage exceeds the threshold to when the circuit turns on (e.g., 5ms to 7ms for a voltage increase from 1200V to 1150V, with a response time of 2ms).

[0138] 2. Comparison with traditional solutions

[0139] Traditional single-stage solutions (only one crowbar) have a higher peak voltage (1380V), lower energy consumption (78kJ), and slower response (5.2ms) due to the lack of secondary coordination. This solution, through three-stage coordination, reduces the peak voltage by 100V, increases energy consumption (rapid energy release), and speeds up the response to 1.8ms, verifying the overvoltage suppression effect.

[0140] By following the above steps, the overvoltage suppression scheme for the 2MW direct-drive unit can be effectively implemented, ensuring that the DC bus voltage is controlled within a safe range during high-voltage grid ride-through, and the unit maintains grid-connected operation.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of this utility model and not to limit it; although the utility model has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications can still be made to the specific implementation of this utility model or equivalent substitutions can be made to some technical features without departing from the spirit of the technical solution of this utility model, and all such modifications and substitutions should be covered within the scope of the technical solution claimed by this utility model.

Claims

1. A multi-stage suppression device for high voltage ride through transient overvoltage of a direct drive wind turbine generator unit, characterized in that, The hardware topology consists of three levels connected in series: Level 1 crowbar circuit: consisting of anti-parallel thyristor modules (1011) and current-limiting inductors (1012) connected in series and parallel between the positive and negative terminals of the DC bus; Level 2 power dissipation branch: consisting of switching devices IGBTs (1021) driving a chromium-aluminum alloy resistor array (1022), which is mounted on an air-cooled heat dissipation substrate (1023); Level 3 unloading circuit: consisting of a mechanical contactor (1031) controlling a water-cooled resistor box (1032), with a copper water channel (1033) built into the resistor box, and the resistors inside the resistor box closely fitting the copper water channel, and heat dissipation through an external circulating water cooling system; FPGA controller: comparing the DC bus voltage with the level 3 threshold in real time, generating a trigger signal and adjusting the level 2 PWM duty cycle.

2. The direct drive wind turbine high voltage ride through transient overvoltage multi-stage suppression device according to claim 1, characterized in that, The trigger threshold of the first-stage crowbar circuit (101) is 1150V±2%, the thyristor module (1011) adopts a 1700V / 200A double anti-parallel structure, and the current limiting inductor (1012) has an inductance value of 2mH±5%.

3. The multi-stage suppression device for transient overvoltage during high voltage ride through of a direct drive wind turbine generator according to claim 1, characterized in that, The PWM duty cycle of the secondary power-consuming branch increases linearly with voltage.

4. The multi-stage suppression device for high-voltage ride-through transient overvoltage of direct-drive wind turbine generators according to claim 1, characterized in that, The IGBT (1021) is model SKM300GB176D, and the chromium-aluminum alloy resistor array (1022) is a resistor array composed of 6 chromium-aluminum alloy resistors, each with a resistance of 20Ω / 5kW, connected in parallel.

5. The direct drive wind turbine high voltage ride through transient overvoltage multi-stage suppression device according to claim 1, characterized in that, The trigger threshold of the three-stage unloading circuit (103) is 1350V±2%, the rated current of the mechanical contactor (1031) is 400A, the resistance of the water-cooled resistor box (1032) is 0.5Ω, and the interface of the copper pipe water channel (1033) is G3 / 4 quick plug type.

6. The multi-stage suppression device for transient overvoltage during high voltage ride through of a direct drive wind turbine generator unit according to claim 1, characterized in that, The FPGA controller (200) implements multi-level timing control: Level 1 trigger delay < 500μs; Level 2 action interval 1ms ± 0.1ms; Level 3 action interval 1ms ± 0.1ms.

7. The multi-stage suppression device for transient overvoltage during high voltage ride through of a direct drive wind turbine generator unit according to claim 1, characterized in that, The device is integrated between the positive and negative poles of the DC bus of the grid-side converter of the direct-drive wind turbine, and the series three-level hardware topology is connected to the bus through a fast-acting fuse.

Images