A wind farm reactive voltage equalization coordination control system for inhibiting field circulating current
By constructing a dynamic network model and using virtual impedance compensation technology, the problems of reactive power circulation and voltage imbalance in wind farms were solved, enabling efficient and stable operation of wind farms and health management of equipment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI YIKONG ELECTRIC POWER TECH CO LTD
- Filing Date
- 2026-05-19
- Publication Date
- 2026-06-19
Smart Images
Figure CN122246776A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of power system automation technology, specifically relating to a wind farm reactive power voltage balancing and coordinated control system for suppressing in-field circulating currents. Background Technology
[0002] With the transformation of the global energy structure, large-scale wind power integration has become an important component of modern power systems. The stable operation of wind farms is directly related to the safety and reliability of the power system, especially in the context of high grid penetration, where the wind farm's ability to support grid voltage and power quality management become particularly critical. To ensure that wind farms meet grid connection requirements while achieving internal operational optimization, precise control of power parameters in the complex collection system is necessary.
[0003] Among these technologies, reactive power and voltage coordination control in wind farms serves as a core means of ensuring system stability, primarily responsible for the unified scheduling of wind turbine generators and various reactive power compensation devices within the farm. Its basic principle is to ensure that the reactive power output at the wind farm's grid connection point meets the grid dispatch requirements through real-time voltage monitoring and feedback adjustment, while simultaneously optimizing internal voltage distribution to reduce operating losses. This technology plays an irreplaceable role in enhancing the wind farm's ability to cope with grid disturbances and improving resource utilization efficiency.
[0004] Current technologies still face significant challenges in reactive power and voltage control in wind farms. Due to the complex topology of large wind farms and the significant differences in impedance distribution among their collector lines, traditional control methods often focus on global regulation of the grid connection voltage while neglecting the voltage coupling relationships between nodes within the farm. This leads to severe reactive power circulating currents between wind turbines, significantly increasing additional losses in the farm's lines and potentially causing excessive equipment temperature rise or even triggering voltage protection actions, resulting in unplanned grid disconnection. Furthermore, existing coordinated control logic suffers from poor balance between response speed and regulation accuracy when dealing with drastic wind speed fluctuations or dynamic grid response demands. It also lacks proactive intervention mechanisms to suppress circulating currents within the farm, ultimately resulting in severe imbalances in reactive power output within the wind farm, becoming a technical bottleneck restricting the long-term stable operation of wind farms. Summary of the Invention
[0005] The purpose of this invention is to provide a wind farm reactive voltage balancing and coordination control system that suppresses in-field circulation, so as to solve the problems mentioned in the background art.
[0006] The technical solution of this invention includes: a multi-dimensional parameter real-time sensing unit, used to synchronously collect voltage phasors of each node in the wind farm's power collection network, wind turbine outlet current, reactive power compensation equipment status, and physical topology parameters of each power collection line; an impedance network dynamic reconstruction unit, used to identify the equivalent complex impedance of each power collection line online based on the physical topology parameters and power operation data collected by the multi-dimensional parameter real-time sensing unit, and construct a dynamic network model including mutual impedance characteristics; and a circulating potential energy assessment unit, used to calculate the potential difference vector between each wind turbine node using the dynamic network model, extract key indicators characterizing the reactive power circulating current intensity within the field, and identify the core path and influencing factors that generate the circulating current; The balanced voltage coordination calculation unit is used to calculate the reactive power output correction value of each wind turbine and reactive power compensation equipment based on the output results of the circulating potential energy assessment unit, with multiple constraints including minimizing the overall voltage deviation, minimizing the network loss within the field, and minimizing the reactive power circulating current intensity. The virtual impedance compensation control unit is used to adjust the control gain of the wind turbine converter in real time according to the reactive power output correction value, and inject virtual impedance components into the converter control logic to offset the voltage instability effect caused by the difference in physical line impedance. The high-speed cooperative instruction distribution unit is used to time-align the instructions calculated by the balanced voltage coordination calculation unit with the parameters of the virtual impedance compensation control unit and send them to each execution terminal.
[0007] Furthermore, the multi-dimensional parameter real-time sensing unit includes a high-precision synchronous phasor measurement module and a topology automatic identification module. The high-precision synchronous phasor measurement module is configured on the high-voltage side of the outlet transformer of each wind turbine and at the collecting bus of the collection line. It achieves nanosecond-level synchronization of data acquisition through a unified clock signal provided by the global navigation satellite system. The topology automatic identification module is used to monitor the status of circuit breakers and disconnectors in real time. When the position of the switches in the field changes, it automatically updates the connection relationship of the collection network.
[0008] As one embodiment of the present invention, the identification process of the impedance network dynamic reconstruction unit is as follows: obtaining the voltage vector difference and current vector between the two ends of the collector line within 10 consecutive sampling periods; using the least squares iterative algorithm to estimate the resistance and reactance parameters of the line online; combining the historical operating temperature data of the line to correct the resistance value for temperature rise, and simultaneously compensating the reactance value for frequency drift based on the current system frequency deviation, thereby generating a dynamic complex impedance matrix that reflects the real physical environment.
[0009] Furthermore, the circulating potential energy assessment unit identifies the circulating path in the following way: establishing a reactive power exchange sensitivity matrix between each wind turbine node; when a deviation outside a preset range is detected between the reactive power flow vector direction and the voltage gradient direction between two adjacent nodes, the branch is determined as a potential circulating path; the reactive power circulation loss on the path is calculated and defined as the circulating potential energy level.
[0010] As one embodiment of the present invention, the equalization voltage coordination calculation unit adopts distributed collaborative optimization logic; the unit divides the wind farm into multiple unit groups, and each group calculates the local optimal reactive power allocation scheme through a consensus algorithm; then the boundary nodes of each group exchange boundary information and perform global convergence iteration across the entire field; during the iteration process, the voltage safety threshold is set between 0.95 times and 1.05 times the rated voltage, and the reactive power output difference between units is required not to exceed a preset threshold of 15%.
[0011] Furthermore, the control logic of the virtual impedance compensation control unit is as follows: a feedback compensation loop is introduced into the current control loop of the wind turbine converter; the virtual inductance and virtual resistance values to be compensated are calculated based on the complex impedance characteristics of the branch; this value is converted into the voltage compensation vector of the converter modulation wave, so that the equivalent impedance exhibited by the unit externally is consistent with the balanced target impedance in the collector network; this process bypasses the modification of physical hardware and realizes the logical reconstruction of line impedance through the active intervention of the control algorithm.
[0012] As one embodiment of the present invention, the high-speed collaborative instruction distribution unit operates on a dual-redundant industrial Ethernet architecture; the unit prioritizes all control instructions, with instructions involving emergency voltage support having the highest priority; in order to overcome the impact of network communication delay on control accuracy, the unit embeds time stamp information of the expected execution time into the instruction data packet, and each wind turbine unit waits or performs advance compensation according to its local high-precision clock after receiving the instruction, ensuring that the action time error of all units is controlled within 5 milliseconds.
[0013] Furthermore, the system of the present invention also includes a health status self-diagnosis unit, which is used to monitor the deviation between the reactive power output of each node and the command value in real time; when the deviation exceeds the allowable error of 10% for three consecutive times, the system automatically switches to the safety degradation operation mode. At this time, the voltage stability of the grid connection point will be prioritized, the fine-grained circulating current suppression logic will be temporarily suspended, and a fault location alarm will be pushed to the maintenance terminal.
[0014] As one embodiment of the present invention, the sampling frequency of the multi-dimensional parameter real-time sensing unit is set to 12.8 kHz to ensure that the high-frequency voltage fluctuation components in the collector network can be captured; the collected analog signal is processed by a 24-bit high-resolution analog-to-digital converter and then transmitted to the central processing unit through an optical fiber bus.
[0015] Furthermore, the impedance network dynamic reconstruction unit also considers the influence of the transformer tap position on the equivalent impedance during the identification process; by reading the feedback signal of the transformer's on-load tap changer, the change in the transformer's leakage reactance is included in the dynamic network model in real time.
[0016] As one embodiment of the present invention, the equalization voltage coordination calculation unit also considers the remaining reactive power capacity of the wind turbine during the calculation process; by acquiring wind speed data and current active power output value in real time, it calculates the maximum reactive power limit that the unit can provide under the current operating conditions, ensuring that the issued reactive power correction value will not exceed the physical safety boundary of the equipment.
[0017] Furthermore, the virtual impedance compensation control unit also has a harmonic suppression function; when calculating the virtual impedance parameters, it extracts the 3rd, 5th and 7th harmonic components in the voltage signal and adjusts the equivalent impedance characteristics of the converter in this frequency band accordingly, thereby achieving synchronous suppression of harmonic circulating current in the field.
[0018] As one embodiment of the present invention, the high-speed cooperative instruction distribution unit adopts a data frame integrity verification mechanism; during data transmission, a cyclic redundancy check code is used to detect each instruction data packet in real time; if an error is found in the data packet due to electromagnetic interference during transmission, a fast retransmission protocol is immediately triggered.
[0019] Furthermore, the system has a large-capacity historical database in the station-level collaborative control center to store more than 180 days of operational data. Through trend analysis of historical data, the system can automatically optimize the target weight parameters in the equalization voltage coordination calculation unit to better adapt to the wind farm operation characteristics under different seasons and meteorological conditions.
[0020] In one embodiment of the present invention, the dynamic reactive power compensation and control unit is connected to the collector bus and contains a static var generator and a switching capacitor bank. When the reactive power regulation capacity of the wind turbine reaches saturation, the equalization voltage coordination calculation unit allocates the remaining reactive power regulation demand to the dynamic reactive power compensation and control unit. Through the rapid response of the large-capacity equipment, it provides the underlying reactive power support foundation for the entire field.
[0021] Furthermore, when calculating multiple constraint objectives, the balanced voltage coordination calculation unit assigns dynamically changing weights to the circulating current suppression objective and the minimum network loss objective. During periods of severe grid voltage fluctuations, the system automatically increases the weight of the minimum voltage deviation objective to ensure safety. During periods of stable grid voltage, the weights of the circulating current suppression and energy-saving loss reduction objectives are increased to improve operational economy.
[0022] As one embodiment of the present invention, the virtual impedance compensation control unit tracks the grid phase in real time through software phase-locked loop technology; when a phase jump occurs, the unit can quickly adjust the angle of the compensation vector to prevent the virtual impedance from introducing additional phase oscillations and ensure the stability of the system during transient processes.
[0023] Furthermore, the impedance network dynamic reconstruction unit can also identify the ground capacitance parameters of the collector line; especially in long-distance cable collector lines, by calculating the reactive charging power generated by the ground capacitance, the input data of the circulating current potential energy assessment unit is corrected, thereby improving the control accuracy of the voltage rise phenomenon in the field under light load conditions.
[0024] As one embodiment of the present invention, the high-speed collaborative instruction distribution unit has instruction feedback confirmation logic; after the execution end completes the control action, it immediately feeds back the execution result and the status parameters after execution; the central control unit compares the feedback information with the expected target, and if it finds that the adjustment effect does not meet the expectations, it automatically starts the feedback compensation logic in the next control cycle.
[0025] Furthermore, the system of the present invention supports direct connection with the upper-level power grid dispatch center; through an interface conforming to the standard communication protocol, it receives the grid connection point voltage command value or reactive power output command value issued by the dispatch center, and uses it as the global input constraint of the equalization voltage coordination calculation unit.
[0026] As one embodiment of the present invention, the multi-dimensional parameter real-time sensing unit also includes an environmental temperature and humidity sensor for monitoring the microclimate environment inside the trench or tower where the current collector cable is located; based on the environmental parameters, the insulation characteristics and impedance thermal stability of the cable are evaluated to further improve the accuracy of the dynamic network model.
[0027] Furthermore, this system adopts a hierarchical control architecture, dividing the control task into a second-level global optimization layer and a millisecond-level local response layer. The second-level global optimization layer is responsible for the macroscopic allocation of reactive resources across the entire field, while the millisecond-level local response layer is handled by the wind turbine terminal control unit, which is used to quickly handle sudden local voltage flicker and instantaneous circulating current impacts.
[0028] As one embodiment of the present invention, the equalization voltage coordination calculation unit adopts a search strategy based on the Pareto optimality principle when processing multi-objective optimization tasks; by finding a set of non-dominated solutions in the solution space and combining the current wind farm's operating preferences, it selects the unique execution scheme most suitable for the current operating conditions.
[0029] Furthermore, the parameter update frequency of the virtual impedance compensation control unit is matched with the switching frequency of the converter, typically set between 2 kHz and 5 kHz, thereby ensuring that the control system has sufficient bandwidth to suppress rapidly changing field circulating currents.
[0030] As one embodiment of the present invention, the high-speed collaborative instruction distribution unit has self-healing capability; when a physical failure occurs in a communication link, the system can automatically switch to a backup optical fiber link, and the amount of instruction loss generated during the switching process does not exceed one control cycle.
[0031] Furthermore, by sorting the voltage sensitivity of all nodes in the field online, the system of the present invention can accurately identify the key units that contribute the most to the voltage stability of the field; under resource-scarce conditions, it prioritizes the use of the reactive power potential of the key units, so as to achieve the greatest stability improvement with the least adjustment cost.
[0032] As one embodiment of the present invention, the impedance network dynamic reconstruction unit also takes into account the impedance characteristics of the wind turbine outlet filter; the response characteristics of the filter inductor and filter capacitor at different frequencies are integrated into the dynamic network model, so that the circulating current suppression logic can be extended to the harmonic frequency band, thereby improving the comprehensiveness of power quality management.
[0033] Furthermore, this system employs a real-time operating system kernel at the software level to ensure that the scheduling of all tasks has deterministic timing, thus avoiding control logic failures due to software runtime delays.
[0034] Compared with the prior art, the advantages and positive effects of the present invention are as follows: This invention achieves real-time and accurate modeling of the physical characteristics of complex power collection networks by constructing a dynamic impedance network reconfiguration unit, breaking the limitations of traditional methods that rely solely on static parameters. Through online identification technology, the system can sense impedance fluctuations caused by line temperature rise, frequency deviation, and topology changes, providing a solid data foundation for the precise suppression of reactive circulating currents. This effectively reduces additional losses in the field caused by impedance mismatch and improves the overall operating efficiency of the wind farm.
[0035] This invention creatively utilizes the control flexibility of wind turbine converters to reconstruct the logical impedance of the line by introducing a virtual impedance compensation control unit, thereby achieving active intervention in the circulating current within the field. This method eliminates potential imbalance between nodes without the need for expensive physical compensation equipment, and can be achieved solely through software algorithm optimization. It fundamentally solves the technical problem of wind turbines competing for reactive power, significantly reduces the risk of unit temperature rise and unplanned grid disconnection, and extends the service life of power electronic components.
[0036] This invention employs a multi-objective constrained balanced voltage coordination calculation and a high-speed collaborative command distribution mechanism, achieving a high degree of unity between voltage stability, energy efficiency optimization, and circulating current suppression. The decision logic based on the Pareto optimal principle ensures that the system can obtain the global optimal solution under various extreme operating conditions, while the millisecond-level time-scale alignment control technology ensures that hundreds of units across the entire field can coordinate precisely as a whole, greatly enhancing the wind farm's ability to support grid voltage and providing reliable technical support for the grid connection of high proportions of renewable energy.
[0037] The system architecture of this invention has extremely high robustness and scalability. Through distributed collaborative optimization and dual-redundant communication design, it ensures stable operation even under local faults or communication jitter. At the same time, the integration of harmonic suppression and environmental parameter sensing functions makes the system not only a voltage regulation tool, but also a comprehensive power quality management and equipment health monitoring platform, which has significant industry demonstration value for promoting the transformation of wind farms towards intelligent and refined operation and maintenance. Attached Figure Description
[0038] Figure 1 This is a schematic diagram of the overall technical solution architecture proposed in this invention; Figure 2 This is a schematic diagram of the core principle framework of field reactive circulating current suppression based on virtual impedance compensation in this invention. Figure 3 This is a logical flowchart of the dynamic reconstruction of collector network impedance and the assessment of circulating current potential energy in this invention. Figure 4 This is a flowchart illustrating the logic framework of balanced voltage coordination calculation and distributed optimization under multi-objective constraints in this invention. Figure 5 This is a schematic diagram of the multi-level interaction relationship and data flow between virtual impedance compensation control and high-speed instruction distribution in this invention. Detailed Implementation
[0039] Example 1 Please refer to the attached document. Figure 1 This embodiment provides a reactive power and voltage balancing coordinated control system for wind farms to suppress circulating currents within the farm. Deployed in the power collection system control center of a large-scale wind farm, this system achieves precise control of complex reactive power flows within the farm through deep interaction with the converter control systems of hundreds or thousands of wind turbine units and the reactive power compensation equipment. The system's physical architecture relies on a high-bandwidth fiber optic communication network, forming a closed-loop management system from data sensing, model reconstruction, risk assessment to collaborative control.
[0040] The wind farm reactive power and voltage balancing coordinated control system for suppressing intra-field circulating currents in this embodiment includes a multi-dimensional parameter real-time sensing unit. This unit is responsible for the synchronous acquisition and preprocessing of raw data from the entire field. The multi-dimensional parameter real-time sensing unit integrates a high-precision synchronous phasor measurement module and an automatic topology identification module. (See attached...) Figure 1As shown, the high-precision synchronous phasor measurement module is physically configured on the high-voltage side of the outlet transformer of each wind turbine, and at each bus node where the collection lines converge at the step-up substation. This module uses a unified clock reference signal provided by the Global Navigation Satellite System (GNSS) to continuously calibrate the local sampling clock using a second pulse signal, ensuring that all monitoring points distributed within a radius of tens of kilometers can achieve synchronous sampling within a nanosecond-level time deviation range. The automatic topology identification module monitors the auxiliary contact status of all circuit breakers, disconnectors, and grounding switches within the wind farm in real time via hardwiring or communication protocols. When the switch positions change due to fault clearing, maintenance operations, or operation mode switching, this module can instantly identify the change in the physical connectivity of the collection network and push the updated topology to subsequent calculation stages in real time.
[0041] To ensure accurate capture of dynamic processes in the power grid, the sampling frequency of the multi-dimensional parameter real-time sensing unit is set to 12.8 kHz. This means that within each 50 Hz power frequency cycle, the system can acquire 256 sampling points, effectively capturing voltage fluctuations caused by nonlinear loads or high-frequency switching of converters. The acquired analog current and voltage signals are processed by a 24-bit high-resolution analog-to-digital converter. This high-bit-count analog-to-digital converter provides a signal-to-noise ratio of up to 144 dB, ensuring sufficient sensitivity in extracting weak circulating current characteristics. The digitized signal is transmitted in real-time to the central processing unit of the control system via a fiber optic Ethernet bus, using a proprietary real-time data message protocol during transmission.
[0042] The multi-dimensional parameter real-time sensing unit also includes environmental temperature and humidity sensors. These sensors are distributed at key locations within the trench or tower where the current collector cable is located, monitoring the microclimate environment in real time. Environmental parameters are used to correct the cable's insulation characteristics and impedance thermal stability. Because the cable's resistance increases with temperature, especially under heavy-load operation in summer, this impedance drift can cause static errors in the preset control model. By introducing environmental parameters, the system can achieve dynamic correction of the line impedance parameters, further improving control accuracy.
[0043] This embodiment of the system also includes an impedance network dynamic reconstruction unit. Please refer to the attached diagram. Figure 3This unit executes online identification logic based on the physical topology parameters and power operation data provided by the multi-dimensional parameter real-time sensing unit. The identification process of the impedance network dynamic reconstruction unit is as follows: The system acquires the voltage vector difference and current vector between the two ends of each collector line segment within 10 consecutive sampling periods. Using a least squares iterative algorithm, with the measured current as the excitation and the measured voltage drop as the response, the resistance and reactance parameters of the line are estimated online. During the iteration process, the system introduces a temperature rise correction mechanism, dynamically adjusting the resistance value based on the historical operating temperature data fed back by the multi-dimensional parameter real-time sensing unit and the Joule heating effect generated by the current real-time current. Simultaneously, for small fluctuations in the system frequency, the reactance value is compensated for frequency drift based on the current system frequency deviation. This compensation is crucial because the reactance value is proportional to the frequency; when disturbances in the power grid cause frequency deviations, static reactance parameters will lead to model distortion.
[0044] The impedance network dynamic reconstruction unit also considers the impact of transformer tap position on equivalent impedance during the identification process. By reading the feedback signal from the transformer's on-load tap changer in real time, the system obtains the current tap ratio. Since changes in tap position directly alter the value of the transformer's leakage reactance mapped to the system side, the system incorporates this change into the dynamic network model in real time. Furthermore, the impedance network dynamic reconstruction unit can also identify the ground capacitance parameters of the collector lines. In long-distance submarine or buried cables, the charging power generated by ground capacitance is significant. The system calculates the ground susceptance value of each line segment by analyzing the reactive power flow characteristics under light load conditions.
[0045] The impedance network dynamic reconstruction unit further considers the impedance characteristics of the wind turbine's outlet filter. The system integrates the frequency response characteristics of the filter inductor and filter capacitor at different frequencies into the dynamic network model. This allows the subsequent circulating current suppression logic to extend beyond the power frequency range to the harmonic frequency band. Ultimately, this unit generates a dynamic complex impedance matrix that reflects the real physical environment. This matrix includes self-impedance and mutual impedance characteristics, forming the underlying mathematical model for overall voltage coordination.
[0046] This embodiment of the system also includes a circulating potential energy assessment unit. Please refer to the attached diagram. Figure 2 With appendix Figure 3This unit uses a dynamic network model to calculate the potential difference vector between each wind turbine node. The circulating potential energy assessment unit identifies circulating paths by establishing a reactive power exchange sensitivity matrix between each wind turbine node. When the system detects a deviation outside a preset range between the reactive power flow vector direction and the voltage gradient direction between two adjacent nodes, it can determine that there is an abnormal reactive power backflow or circulating phenomenon in that branch. The system calculates the reactive power circulation loss on this path and defines it as the circulating potential energy level. The higher the circulating potential energy level, the more severe the reactive power imbalance on that path, and the greater the threat to the safe operation of the unit and the energy efficiency of the electric field.
[0047] The circulating current potential energy assessment unit extracts key indicators through the following algorithm: The system constructs a voltage node equation for the entire field and calculates the theoretical voltage distribution using the complex power injection values of each node. The theoretical voltage distribution is compared with the measured voltage vector collected by the multi-dimensional parameter real-time sensing unit, and residual analysis is used to identify the core paths and influencing factors that generate the circulating current. For example, if the circulating current on a certain path is mainly caused by inconsistent reactive power output commands between the two generators, this influencing factor is marked as control source induced; if it is mainly caused by line impedance imbalance, it is marked as structural source induced.
[0048] This embodiment of the system also includes a voltage equalization coordination calculation unit. This unit is the command center of the entire system. Please refer to the appendix. Figure 4 The voltage balancing and coordination calculation unit employs distributed collaborative optimization logic. The system divides the wind farm into multiple turbine groups based on geographical location and electrical distance. Within each group, a consensus algorithm calculates the optimal reactive power allocation scheme locally. Boundary nodes of each group exchange boundary information and perform global convergence iteration across the entire farm. During the iteration process, the system sets a voltage safety threshold between 0.95 and 1.05 times the rated voltage, and requires that the reactive power output difference between any two turbines not exceed a preset threshold of 15%.
[0049] The voltage balancing and coordination calculation unit also considers the remaining reactive power capacity of the wind turbine during the calculation process. By acquiring real-time wind speed data and current active power output values, the system calculates the maximum reactive power limit that the unit can provide under the current operating conditions based on the PQ operating limit curve of the wind turbine converter. This ensures that the issued reactive power output correction value is always within the physical safety boundary of the equipment, avoiding converter overcurrent tripping due to excessive reactive power extraction.
[0050] The voltage balancing coordination calculation unit employs a search strategy based on the Pareto optimality principle when solving for multiple constraint objectives. The system's objective function includes three core dimensions: minimizing the overall voltage deviation, minimizing network losses within the field, and minimizing reactive power circulating current intensity.
[0051] In the above formula, To comprehensively optimize the objective function, For all in the power grid Sum the results of each node / branch. Representing the Real-time voltage per unit value of each node. This represents the per-unit value of the grid connection point reference voltage. Representing the Power loss of the line segment Representing the The effective value of the circulating current of each branch. , , These represent dynamic weighting coefficients. The equalization voltage coordination calculation unit adjusts these weights in real time according to the grid operating status. During periods of severe grid voltage fluctuations, the system automatically increases weight α to ensure safety and prioritize ensuring that the voltage does not exceed limits; during periods of stable grid voltage, weights β and γ are increased to improve operational economy and suppress circulating currents.
[0052] By finding a set of non-dominated solutions within the solution space, the system, combined with the current operating preferences of the wind farm, selects the unique execution scheme most suitable for the current operating conditions. The output of this scheme is the reactive power output correction value for each wind turbine and the dynamic reactive power compensation and control unit. If the reactive power regulation capacity of the wind turbine itself reaches saturation, the equalization voltage coordination calculation unit allocates the remaining reactive power regulation demand to the dynamic reactive power compensation and control unit. This dynamic reactive power compensation and control unit is connected to the collector bus and internally contains a static var generator and switching capacitor banks, providing underlying reactive power support for the entire farm through the rapid response of large-capacity equipment.
[0053] This embodiment of the system also includes a virtual impedance compensation control unit. Please refer to the attached document. Figure 2 With appendix Figure 5 This unit adjusts the control gain of the wind turbine converter in real time based on the reactive power output correction value issued by the equalization voltage coordination calculation unit. The control logic of the virtual impedance compensation control unit introduces a feedback compensation loop into the current control loop of the wind turbine converter. Based on the complex impedance characteristics of the branch, the system calculates the virtual inductance and virtual resistance values that need to be compensated.
[0054] In the above formula, The voltage compensation vector value represents the modulated wave of the converter. Represents the unit current reference vector. Represents virtual resistance parameters. Represents virtual reactance parameters. (Virtual reactance) represents the imaginary part, which mainly affects the phase of voltage / current. This value is converted into a voltage bias term in the converter control algorithm, ensuring that the equivalent impedance exhibited by the unit externally is consistent with the target impedance in the collector network. This method bypasses modifications to the physical hardware and achieves logical reconstruction of line impedance entirely through active intervention in the software algorithm, thereby offsetting the voltage instability effect caused by differences in physical line impedance.
[0055] The virtual impedance compensation control unit also features harmonic suppression. When calculating virtual impedance parameters, the system extracts the 3rd, 5th, and 7th harmonic components from the voltage signal using Fast Fourier Transform (FFT) and adjusts the converter's equivalent impedance characteristics in that frequency band accordingly, making it exhibit high impedance to block the path of harmonic circulating currents. The virtual impedance compensation control unit tracks the grid phase in real time using software phase-locked loop (PLL) technology. When a phase transition occurs, the unit can quickly adjust the angle of the compensation vector to prevent the virtual impedance from introducing additional phase oscillations. Its parameter update frequency matches the converter's switching frequency, typically set between 2 kHz and 5 kHz, ensuring the control system has sufficient bandwidth to suppress rapidly changing circulating currents within the field.
[0056] This embodiment of the system also includes a high-speed cooperative instruction distribution unit. This unit operates on a dual-redundant industrial Ethernet architecture and is responsible for time-scale alignment of the instructions calculated by the equalization voltage coordination calculation unit with the parameters of the virtual impedance compensation control unit. Please refer to the appendix. Figure 5 The high-speed collaborative instruction distribution unit prioritizes all control instructions. Instructions involving emergency grid voltage support are given the highest priority, ensuring they can bypass the regular task queue and be quickly transmitted.
[0057] To overcome the impact of network communication latency on control accuracy, the high-speed collaborative command distribution unit embeds time stamp information of the expected execution time into the command data packet. Upon receiving the command, each wind turbine unit waits or compensates for delays based on its local high-precision clock. If the data packet arrives late, the system interpolates the command value based on the delay time. This mechanism ensures that the timing error of hundreds of units distributed across different geographical locations is controlled within 5 milliseconds. The unit employs a data frame integrity verification mechanism, using cyclic redundancy check codes to detect each command data packet in real time. If errors are detected in the data packet due to electromagnetic interference during transmission, a fast retransmission protocol is immediately triggered.
[0058] The high-speed collaborative command distribution unit also features command feedback confirmation logic. After completing the control action, the execution end immediately feeds back the execution result and the voltage and current status parameters after execution. The central control unit compares the feedback information with the expected target; if the adjustment effect does not meet expectations, it automatically initiates feedback compensation logic in the next control cycle. This unit has self-healing capabilities; when a physical break occurs in a fiber optic link, the system can automatically switch to a backup redundant link, with the command loss during the switching process not exceeding one control cycle.
[0059] This embodiment of the system also includes a health status self-diagnosis unit. This unit operates independently of the main control logic and monitors the deviation between the reactive power output of each node and the command value in real time. When the system detects that the deviation between the actual reactive power output of a certain unit and the command value exceeds the allowable error of 10% for three consecutive times, it determines that the node may have a sensor fault or actuator malfunction. At this time, the system automatically switches to a safety-degraded operation mode. In degraded mode, the system prioritizes ensuring the voltage stability of the grid connection point, temporarily suspending fine-grained circulating current suppression logic, sacrificing some operating economy to ensure that the entire field does not disconnect from the grid. At the same time, the system pushes detailed fault location alarms to the maintenance terminal.
[0060] This embodiment of the system also includes a large-capacity historical database. This database stores over 180 days of full-field operational data, including phasor measurement data for each node, topology change records, and control command execution records. Through long-term trend analysis of this historical data, the system uses machine learning algorithms to automatically optimize the target weight parameters in the equalization voltage coordination calculation unit. For example, the system can identify the impedance variation patterns of certain lines under specific wind conditions in a specific season, thereby adjusting the control strategy in advance to better adapt to the wind field operating characteristics under different meteorological conditions.
[0061] This system supports direct connection with the superior power grid dispatch center. Through a northbound interface conforming to international standard communication protocols, the system receives grid connection point voltage reference values or total reactive power output commands from the dispatch center. The voltage balancing coordination calculation unit treats these external commands as global input constraints, and, while meeting the requirements of the superior dispatch center, performs circulating current suppression and voltage balancing optimization within the field.
[0062] This system employs a hierarchical control architecture, dividing control tasks into a second-level global optimization layer and a millisecond-level local response layer. The second-level global optimization layer is handled by the central processing unit, performing complex Pareto optimality searches and global reactive power allocation. The millisecond-level local response layer, on the other hand, is distributed in the terminal control units of each wind turbine, rapidly handling sudden local voltage flicker using a preset virtual impedance slope. This hierarchical architecture balances the trade-off between computational complexity and response speed.
[0063] Because this system employs a real-time operating system kernel at the software level, it ensures that the scheduling of all tasks has deterministic timing. Whether it's data acquisition, impedance calculation, or command issuance, all tasks strictly adhere to preset time slice quotas. This deterministic scheduling mechanism avoids control logic delays caused by resource contention during software operation, providing a reliable operating environment for real-time scenarios like wind farms.
[0064] By ranking the voltage sensitivity of all nodes across the entire field online, this system can accurately identify the key generating units that contribute the most to the field's voltage stability. Under extremely weak grid conditions or when reactive power resources are extremely scarce, the system will prioritize utilizing the reactive power regulation potential of these key generating units. Through precise measures targeting critical nodes, the system can achieve the greatest improvement in overall field stability with minimal regulation costs.
[0065] Example 2 Building upon Example 1, this example is further optimized for the specific operating conditions of large-scale offshore wind farms. Offshore wind farms, due to the use of long submarine cables, exhibit significantly higher ground capacitance and charging power compared to onshore wind farms. Therefore, this example incorporates a distributed parameter model for submarine cables within the impedance network dynamic reconstruction unit.
[0066] In the impedance network dynamic reconstruction unit, the system not only identifies the lumped parameters of the line, but also calculates the propagation constant and characteristic impedance of the submarine cable by measuring the voltage phasors at both ends of the cable and using a hyperbolic function model. This allows the system to more accurately describe the voltage rise phenomenon of long-distance cables under light load conditions, i.e., the Freanti effect. Through accurate modeling, the circulating potential energy assessment unit can identify the risk of reactive power backfeed between units caused by the cable capacitance effect and provide early warning.
[0067] To address the extreme environments of offshore wind farms characterized by high salt spray and high humidity, the environmental sensors in the multi-dimensional parameter real-time sensing unit of this embodiment possess a higher level of protection. The system utilizes real-time humidity data and feedback from the insulation monitoring device to dynamically adjust the damping factor of the virtual impedance compensation control unit. When the ambient humidity exceeds 95% and there is a tendency for insulation degradation, the system actively increases the virtual resistance value to limit potential transient overcurrents, thus providing soft protection.
[0068] In the balanced voltage coordination calculation unit, this embodiment specifically considers the challenges and high costs of maintaining offshore wind turbines. An equipment health weight term is added to the Pareto optimal search objective function. By monitoring the temperature rise curves of the converter power devices in real time, if the system detects that the core temperature of a unit's converter is too high, even if it has a large reactive power regulation margin, the balanced voltage coordination calculation unit will automatically reduce its weight in the optimization objective and transfer its reactive power regulation task to an adjacent unit with a lower temperature. This reactive power allocation mechanism based on equipment health status can significantly reduce the failure rate of offshore wind turbines.
[0069] In this embodiment, the high-speed collaborative command distribution unit employs a more robust ring network redundancy protocol. Given the extreme vulnerability of offshore wind farm fiber optic links to seabed shifts or anchor damage, the system constructs a redundant mesh with multiple physical paths. When the packet loss rate of a backbone link exceeds 5%, the system immediately initiates multi-path concurrent transmission, meaning the same control command is simultaneously transmitted through multiple electrical paths. After receiving multiple copies, each wind turbine terminal retains the first arriving valid data packet through time stamp verification. This mechanism improves the reliability of the communication link to over 99.999%.
[0070] In marine operating conditions, the virtual impedance compensation control unit adds the function of filtering out load fluctuations caused by wave impact. The mechanical load generated by waves is reflected in the generator output power through the transmission chain, which in turn causes low-frequency voltage oscillations. The virtual impedance compensation control unit, by connecting a quasi-proportional resonant controller in series in the control loop, dynamically reconstructs the virtual impedance characteristics according to the main frequency of load fluctuations, thereby achieving online suppression of power quality problems caused by mechanical disturbances.
[0071] The health status self-diagnostic unit adds a function to monitor communication delay jitter in offshore wind power scenarios. Due to the long communication links and numerous relay nodes at sea, delay jitter may disrupt command synchronization. When the diagnostic unit detects that the standard deviation of delay jitter exceeds 2 milliseconds, it will automatically notify the equalization voltage coordination calculation unit to reduce the adjustment step size and adopt a more robust gradual adjustment strategy to prevent self-excited oscillations in the field induced by control asynchrony.
[0072] The system in this embodiment also specifically considers the interaction between the reactive power compensation equipment and the turbine group within the offshore wind farm's booster station. When the static var compensator (SVC) at the booster station performs large-scale reactive power switching, the voltage balancing coordination calculation unit issues a feedforward pre-compensation command to all turbines in the field. The turbine converters actively absorb or release a portion of the reactive power through instantaneous adjustment of virtual impedance, thereby mitigating the voltage surge generated during the switching of large-capacity compensation equipment and achieving efficient coordination between the turbines and the station-end equipment.
[0073] Example 3 Building upon the above embodiments, this embodiment provides a more refined control scheme for wind farms containing multiple reactive power sources. This wind farm includes not only wind turbine converters but also distributed static synchronous compensators deployed at the end of the power collection network.
[0074] In this heterogeneous reactive power source environment, the impedance network dynamic reconfiguration unit incorporates the grid-connected node of each distributed static synchronous compensator into the dynamic modeling. The system constructs a more granular three-dimensional impedance map by identifying the electrical distance and equivalent impedance from the compensator outlet to the collector trunk line. Using this map, the circulating current potential energy assessment unit can accurately calculate the circulating current components between the unit and the compensator, and between compensators themselves.
[0075] For this operating condition, the voltage balancing coordination calculation unit adopts a hierarchical and zoned control strategy. The system defines the compensator as the primary regulating node and the wind turbine converter as the secondary regulating node. In the initial stage of voltage fluctuation, the distributed compensator with faster response speed is given priority to provide transient support using its overload capacity; when the fluctuation enters the steady-state compensation stage, the voltage balancing coordination calculation unit calculates the optimal ratio between the wind turbine and the compensator to achieve a smooth handover of reactive power output.
[0076] In this embodiment, the virtual impedance compensation control unit achieves frequency response decoupling. For the wind turbine converter, its virtual impedance characteristics focus on suppressing the static voltage difference caused by the physical distance difference between units; while for the distributed compensator, its virtual impedance characteristics are set as high-pass filtering characteristics to specifically absorb high-frequency ripple and transient circulating current components in the power collection network. This frequency-division control strategy fully leverages the dynamic characteristics of different types of reactive power sources.
[0077] To address this complex topology, the high-speed collaborative command distribution unit introduces a priority dynamic preemption mechanism. When the system detects an emergency such as a single-phase ground fault in the field, control data packets involving impedance reconfiguration of the faulty branch will directly preempt the physical layer's transmission bandwidth. By configuring hardware-level priority queues in the Ethernet switch, the end-to-end transmission latency of critical commands is reduced to 1.2 milliseconds.
[0078] The health status self-diagnosis unit integrates a predictive maintenance module based on big data analytics. By retrieving tens of thousands of circulating current suppression process records from the historical database, it uses a deep neural network to extract early fault features. For example, if it detects that the virtual impedance compensation component of a unit deviates from the normal distribution range for a long period, accompanied by subtle high-frequency current fluctuations, the system will preemptively determine that the converter filter capacitors of that unit are at risk of aging. This proactive diagnostic mode extends the system's functionality from operation control to asset management.
[0079] The system in this embodiment also considers the linkage with the wind farm energy storage system. In the event of a severe voltage drop leading to the depletion of reactive power resources, the voltage balancing coordination calculation unit can urgently dispatch the energy storage converter to participate in voltage support. At this time, the virtual impedance compensation control unit will calculate a set of special droop control parameters for the energy storage converter, enabling it to provide inertia support and voltage clamping for the power collection network, just like a synchronous generator.
[0080] The multi-dimensional parameter real-time sensing unit adds an online measurement function for the dynamic response delay of each reactive power source in environments with multiple reactive power sources. The system periodically sends out small disturbance commands and observes the response waveforms of each node, thereby determining the actual physical delay from receiving the command to the output change of different types of equipment. This delay parameter is fed back to the equalization voltage coordination calculation unit to correct the time constant in the collaborative optimization process, thus completely eliminating the control oscillation problem caused by inconsistent equipment response speeds.
[0081] This embodiment further enhances the system's adaptability to grid harmonic distortion. The impedance network dynamic reconstruction unit extracts background harmonics from the grid, calculates the system's equivalent impedance at each frequency band, and feeds it back to the virtual impedance compensation control unit. By exhibiting extremely high or extremely low impedance at specific harmonic frequencies, the converter can effectively block background harmonics from being injected into the wind farm or actively compensate for harmonic currents generated within the farm. This not only suppresses reactive circulating currents within the farm but also significantly improves the power quality at the grid connection point.
[0082] The voltage balancing coordination calculation unit of this system introduces a parallel computing architecture when handling large-scale nonlinear optimization problems. Utilizing the computing resources of multi-core processors, the optimization task for the entire wind farm is decomposed into multiple parallel sub-tasks that run synchronously. This allows the global optimization calculation cycle to be stably controlled within 20 milliseconds, even for ultra-large wind farms containing more than 500 wind turbines. This extremely high computational efficiency provides the technical prerequisite for the system to cope with rapid transient processes in the power grid.
[0083] The high-speed collaborative command distribution unit also employs a flow control algorithm based on information entropy. When local congestion occurs in the communication network, the system automatically compresses the transmission frequency of non-core data, prioritizing the data flow of core nodes that contribute the most to suppressing circulation. Through this intelligent bandwidth allocation, the system can maintain basic equalization control functions even in harsh communication environments, demonstrating extremely strong system robustness.
[0084] Through the organic combination of the above-described embodiments, the wind farm reactive power and voltage balancing coordinated control system for suppressing circulating currents within the wind farm, as described in this invention, constructs a highly intelligent, real-time control system with self-healing capabilities through in-depth technological innovation in various aspects such as sensing, modeling, evaluation, optimization, control, and distribution. This system not only solves the industry-wide problems of uneven reactive power distribution and severe circulating currents within wind farms, but also transforms the wind farm from a passive power output unit into an intelligent entity capable of actively supporting the power grid and self-optimizing its operational quality through advanced methods such as virtual impedance reconstruction. This demonstrates extremely significant technological advancement and engineering application value.
[0085] It should be clarified that the above embodiments are merely preferred forms of the technical solution of the present invention and are not intended to limit the scope of protection of the present invention. Any equivalent substitutions or improvements made by those skilled in the art regarding specific parameters, calculation methods, communication protocols, or hardware selections without departing from the core design concept of the present invention should be covered within the scope of the claims of the present invention. In particular, the setting of various weight parameters, the adjustment of sampling frequency, and the evolution of optimization algorithms should all be considered as specific applications of the present invention in different engineering environments. The hierarchical distributed architecture adopted by this system provides ample technical interfaces for future integration of more renewable energy forms and more complex distribution network interaction functions.
Claims
1. A wind farm reactive power and voltage balancing coordinated control system for suppressing circulating currents within the field, characterized in that, include: The multi-dimensional parameter real-time sensing unit is used to synchronously collect the voltage phasors of each node in the wind farm's power collection network, the output current of the wind turbine generator, the status of the reactive power compensation equipment, and the physical topology parameters of each power collection line. The impedance network dynamic reconstruction unit is used to identify the equivalent complex impedance of each collector line online based on the physical topology parameters and power operation data collected by the multi-dimensional parameter real-time sensing unit, and to construct a dynamic network model that includes mutual impedance characteristics. The circulating potential energy assessment unit is used to calculate the potential difference vector between the nodes of each wind turbine using the dynamic network model, extract key indicators characterizing the reactive circulating intensity in the field, and identify the core path and influencing factors that generate the circulating current. The balanced voltage coordination calculation unit is used to calculate the reactive power output correction value of each wind turbine and reactive power compensation equipment based on the output results of the circulating potential energy assessment unit, with multiple constraint objectives including minimizing the overall voltage deviation, minimizing the network loss in the field, and minimizing the reactive circulating current intensity. The virtual impedance compensation control unit is used to adjust the control gain of the wind turbine converter in real time according to the reactive power output correction value, and inject virtual impedance components into the converter control logic to counteract the voltage instability effect caused by the difference in physical line impedance. The high-speed collaborative instruction distribution unit is used to perform time-scale alignment processing between the instructions calculated by the equalization voltage coordination calculation unit and the parameters of the virtual impedance compensation control unit, and then distribute them to each execution terminal.
2. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The multi-dimensional parameter real-time sensing unit includes: A high-precision synchronous phasor measurement module is used to achieve nanosecond-level synchronous data acquisition at the high-voltage side of the outlet transformer and the busbar of the collector line of each wind turbine. The topology auto-identification module is used to monitor the status of circuit breakers and disconnectors in real time, and automatically update the connectivity of the collector network when the switch position changes.
3. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The impedance network dynamic reconstruction unit is used for: Obtain the voltage vector difference and current vector at both ends of the collector line; Based on the voltage vector difference and current vector, an iterative algorithm is used to estimate the resistance and reactance parameters of the line online. By combining ambient temperature and system frequency, temperature rise correction and frequency drift compensation are applied to the resistance and reactance parameters respectively to generate a dynamic complex impedance matrix.
4. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The circulating potential energy assessment unit is used for: Establish the reactive power exchange sensitivity matrix among each wind turbine node; When an abnormal deviation is detected between the reactive flow vector direction and the voltage gradient direction between adjacent nodes, the branch is determined to be a potential circulating current path. Calculate the reactive energy circulation loss on the potential circulation path and define it as the circulation potential energy level.
5. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The equalization voltage coordination calculation unit is used for: The wind farm is divided into multiple turbine groups, and the local optimal reactive power allocation scheme is calculated within each group through a consensus algorithm. Information is exchanged at boundary nodes to achieve global convergence iteration across the entire field; During the iteration process, the voltage across the entire field is constrained to be within a safe threshold range, and the reactive power output difference between any two units is limited to a preset ratio.
6. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The virtual impedance compensation control unit is used for: A feedback compensation circuit is introduced into the current control loop of the wind turbine converter. Calculate the virtual inductance and virtual resistance values to be compensated based on the complex impedance characteristics of the branch. The virtual inductance and virtual resistance are converted into voltage compensation vectors of the converter modulation wave, so that the unit presents the target equivalent impedance to the outside world.
7. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The high-speed collaborative instruction distribution unit is used for: It runs on a dual-redundant industrial Ethernet architecture and prioritizes control commands. Insert time stamp information of the expected execution time into the instruction data packet; Each wind turbine unit is configured to wait or advance compensation based on a local high-precision clock, ensuring that the overall operation time error is controlled within the millisecond range.
8. The wind farm reactive power and voltage balancing coordination control system for suppressing intra-field circulating currents according to claim 1, characterized in that, Also includes: The health status self-diagnosis unit is used to monitor the deviation between the reactive power output of each node and the command value in real time; When the deviation exceeds the allowable error multiple times consecutively, the system automatically switches to a safety-degraded operation mode and sends a fault location alarm to the maintenance terminal.
9. The wind farm reactive power and voltage balancing coordinated control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The equalization voltage coordination calculation unit is also used for: By combining wind speed data with the current active power output, the maximum reactive power limit of the wind turbine under the current operating conditions is calculated. Ensure that the issued reactive power output correction value does not exceed the physical safety boundary of the equipment.
10. The wind farm reactive power and voltage balancing coordinated control system for suppressing intra-field circulating currents according to claim 1, characterized in that, The virtual impedance compensation control unit is also used for: Harmonic components are extracted from the voltage signal, and the equivalent impedance characteristics of the converter are adjusted accordingly in the corresponding frequency band. The software phase-locked loop tracks the grid phase in real time and dynamically adjusts the compensation vector angle when the phase changes, preventing the introduction of additional phase oscillations.