Intelligent voltage regulation and control method and system for offshore wind power booster stations

Abstract

This invention provides a method and system for intelligent voltage regulation and control of offshore wind power booster stations, relating to the field of wind power generation technology. The method includes: decomposing the output power of wind turbine generators using wavelet transform to determine a fuzzy rule base and calculate reactive power reserve capacity; dividing the booster station into multiple sub-regions, calculating voltage coupling degree, and forming a joint control group; designing an adaptive coordinated control strategy using a particle swarm optimization algorithm; and executing corresponding controls for high-frequency and low-frequency fluctuations when voltage fluctuations are detected. This method achieves intelligent regulation of the voltage at the offshore wind power grid connection point, improves system operational stability, and extends equipment lifespan.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, and in particular to a method and system for intelligent voltage regulation and control of offshore wind power booster stations. Background Technology

[0002] With the rapid development of the offshore wind power industry and the continuous expansion of offshore wind farms, power quality issues, especially voltage stability, have become increasingly prominent. Offshore wind power operates in a more complex environment than onshore wind power. Offshore wind resources are highly volatile and unpredictable. Furthermore, offshore wind farms are typically located far from load centers, transmitting power via submarine cables, resulting in long transmission lines. These factors make offshore wind power highly susceptible to grid voltage fluctuations and flicker after grid connection. As the interface between the offshore wind farm and the power grid, the voltage stability of the offshore wind farm's booster substation directly affects the safe operation and power quality of the wind farm.

[0003] Currently, voltage control in offshore wind power booster stations mainly employs a combination of on-load tap changer and reactive power compensation devices. Traditional voltage control methods have the following shortcomings:

[0004] Existing control methods lack in-depth analysis and utilization of wind power output characteristics, and cannot effectively cope with the high-frequency disturbances and low-frequency fluctuations of wind power, resulting in inaccurate control response and problems such as untimely or excessive voltage regulation when wind speed changes drastically.

[0005] Traditional voltage control strategies typically control the substation as a whole, ignoring the electrical coupling and voltage correlation between different areas within the substation. This makes it impossible to achieve fine-grained zoned coordinated control, and can easily lead to control conflicts or resource waste in complex topologies.

[0006] Existing control methods mostly adopt fixed parameter control strategies, lacking adaptability and intelligent optimization capabilities. They are difficult to dynamically adjust control objectives and strategies according to the system's operating status, and cannot achieve the optimal balance between voltage qualification rate and equipment lifespan. In particular, they have poor adaptability under extreme weather conditions. Summary of the Invention

[0007] This invention provides a method and system for intelligent voltage regulation and control of offshore wind power booster stations, which can solve the problems in the prior art.

[0008] A first aspect of the present invention provides a method for intelligent voltage regulation and control of offshore wind power booster stations, comprising:

[0009] The collected offshore wind turbine output power is decomposed into high-frequency disturbance components and low-frequency fluctuation components through wavelet transform. The fuzzy rule base of the adaptive fuzzy controller is determined based on the high-frequency disturbance components, and the required reserve capacity of the reactive power compensation device is calculated based on the changing trend of the low-frequency fluctuation components.

[0010] Offshore wind power booster stations are divided into multiple voltage control sub-regions according to their electrical connections. The voltage correlation between each sub-region is calculated based on real-time collected voltage and operational data to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions is greater than the voltage coupling degree threshold, the sub-region and its adjacent sub-regions are classified into a joint control group.

[0011] An adaptive coordinated control strategy is designed for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with minimizing voltage deviation and minimizing the number of equipment switching. The search capability of the particle swarm is adjusted by adaptive inertia weight, and the weight coefficients of the optimization objectives are dynamically updated so that the control strategy can be adaptively adjusted with the system operating state.

[0012] When voltage fluctuations are detected, the adaptive fuzzy controller quickly generates switching commands for reactive power compensation devices for high-frequency disturbance components and puts standby reactive power compensation devices into operation according to the predicted capacity demand for low-frequency fluctuation components; at the same time, the optimal control scheme obtained by the adaptive coordinated control strategy is executed.

[0013] The collected offshore wind turbine output power is decomposed into high-frequency disturbance components and low-frequency fluctuation components using wavelet transform. The fuzzy rule base for the adaptive fuzzy controller is determined based on the high-frequency disturbance components. Based on the changing trend of the low-frequency fluctuation components, the required reserve capacity for the reactive power compensation device is calculated, including:

[0014] Wavelet coefficients are obtained by high-pass filtering and low-pass filtering of the power output signal using wavelet basis functions, and then high-frequency disturbance components and low-frequency fluctuation components are obtained by orthogonal transformation and reconstruction of the wavelet coefficients.

[0015] The disturbance amplitude, disturbance frequency, and disturbance duration are calculated based on the high-frequency disturbance components. The disturbance amplitude, disturbance frequency, and disturbance duration are then classified into levels. A trapezoidal membership function is used to map the disturbance amplitude, disturbance frequency, and disturbance duration to membership values ​​between zero and one. An IF-THEN form fuzzy rule table is generated based on the levels and the membership values ​​to obtain the fuzzy rule base of the adaptive fuzzy controller.

[0016] Extract the data sequence of the low-frequency fluctuation component within a preset time window, and calculate the fluctuation mean, fluctuation standard deviation, and trend slope; use the fluctuation mean, fluctuation standard deviation, and trend slope to perform an autoregressive moving average prediction to obtain the predicted capacity, and take the maximum value of the product of the predicted capacity and the fluctuation standard deviation, and the product of the rated capacity and the preset ratio as the reserve capacity.

[0017] Offshore wind power booster stations are divided into multiple voltage control sub-regions based on their electrical connections. Voltage correlations between sub-regions are calculated based on real-time collected voltage and operational data, generating a voltage coupling matrix. When the voltage coupling between any sub-region and its adjacent sub-regions exceeds a voltage coupling threshold, that sub-region and its adjacent sub-regions are grouped into a joint control group, including:

[0018] A topological correlation matrix is ​​established based on the bus connection relationship of the offshore wind power booster station. The direct connection relationship between corresponding sub-regions is mapped to the element values ​​of the topological correlation matrix. Based on the topological correlation matrix, the offshore wind power booster station is divided into multiple voltage control sub-regions.

[0019] Voltage and operating data of each voltage control sub-region are collected in real time. The voltage correlation coefficient is obtained by weighting the ratio of voltage fluctuation amplitude and phase difference between adjacent voltage control sub-regions within a preset time window. The power transmission coefficient is obtained by calculating the ratio of the maximum active power transmitted between adjacent voltage control sub-regions to the rated capacity of the sub-region. The impedance coupling coefficient is obtained by multiplying the reciprocal of the line impedance between adjacent voltage control sub-regions by the line voltage level. The voltage correlation coefficient, the power transmission coefficient, and the impedance coupling coefficient are weighted according to preset weights to generate a voltage coupling degree matrix.

[0020] The voltage coupling degree threshold is calculated based on the non-zero elements in the voltage coupling degree matrix. It is then determined whether the voltage coupling degree between any voltage control sub-region and its adjacent voltage control sub-region is greater than the voltage coupling degree threshold. If so, the voltage control sub-region and its adjacent voltage control sub-region are merged into a joint control group.

[0021] An adaptive coordinated control strategy is designed for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with minimizing voltage deviation and minimizing the number of equipment switching operations. The search capability of the particle swarm is adjusted through adaptive inertia weights, and the weight coefficients of the optimization objectives are dynamically updated, enabling the control strategy to adaptively adjust with the system operating state. This includes:

[0022] The voltage fluctuation trend is obtained by calculating the change in voltage deviation between the current time and the previous control time of each node in the joint control group, and the equipment adjustment frequency is obtained by counting the cumulative number of actions of transformer taps and reactive power compensation devices within the preset time window.

[0023] The voltage control priority coefficient is dynamically updated based on the voltage fluctuation trend, and the switching constraint coefficient is dynamically updated based on the equipment adjustment frequency. The product of the mean of the sum of squares of voltage deviations of each node in the joint control group and the voltage control priority coefficient is used as the voltage control sub-objective. The result of adding the change in transformer tap position and the change in switching status of reactive power compensation device at adjacent control times and multiplying it with the switching constraint coefficient is used as the switching count sub-objective. The voltage control sub-objective and the switching count sub-objective are added together to obtain the adaptive optimization objective.

[0024] The particle swarm search iteration count is initialized to zero. A maximum iteration count threshold is set, and the ratio of the iteration count to the maximum iteration count threshold is calculated. The ratio is then substituted into the exponential decay function to obtain the adaptive search weight. Multiple sets of particle position vectors composed of transformer tap positions and reactive power compensation device switching states are generated.

[0025] The adaptive search weights are combined with the historical optimal solution and the global optimal solution of the particle position vector to update the particle position vector. The adaptive optimization target value corresponding to the updated particle position vector is calculated. The number of iterations is increased to obtain the optimal transformer tap position and reactive power compensation device switching state.

[0026] When voltage fluctuations are detected, the adaptive fuzzy controller quickly generates switching commands for the reactive power compensation device, targeting high-frequency disturbance components, including:

[0027] The voltage deviation fuzzy sub-interval is divided based on the fluctuation amplitude of the high-frequency component, and the voltage deviation change rate fuzzy sub-interval is divided based on the fluctuation frequency of the high-frequency component.

[0028] The voltage deviation and voltage deviation rate of change are input into the Gaussian membership function, respectively. Based on the fuzzy sub-intervals of the voltage deviation and voltage deviation rate of change, the membership degree of each sub-interval is calculated. The sub-interval corresponding to the highest membership degree is selected as the membership value of the voltage deviation and the voltage deviation rate of change. The fuzzy membership degree is then multiplied by the membership value of the voltage deviation and the voltage deviation rate of change to obtain the comprehensive fuzzy membership value.

[0029] The reference gain coefficient is adaptively adjusted according to the real-time fluctuation amplitude of the high-frequency component. The product of the voltage fluctuation amplitude and the reference gain coefficient is used as the dynamic control gain. The fuzzy membership degree comprehensive value is multiplied by the dynamic control gain to obtain the switching amount of the reactive power compensation device.

[0030] When the voltage fluctuation amplitude of the high-frequency component is detected to be greater than the high-frequency fluctuation threshold, the switching order and switching time of each group of reactive power compensation devices are calculated based on the switching amount of the reactive power compensation device. The switching operation of the reactive power compensation devices is performed one group at a time according to the switching order. After each group is switched, the voltage recovery is detected. When the voltage fluctuation amplitude is reduced to below the high-frequency fluctuation threshold, the switching is stopped.

[0031] To address low-frequency fluctuations, the standby reactive power compensation devices will be put into operation according to the predicted capacity demand, including:

[0032] The reactive capacity increment sequence is obtained by calculating the reactive capacity difference between adjacent sampling times within a preset time window, and the voltage deviation increment sequence is obtained by calculating the voltage deviation difference between adjacent sampling times within the preset time window. The first prediction weight is adaptively adjusted according to the changing trend of the reactive capacity increment sequence, and the second prediction weight is adaptively adjusted according to the changing trend of the voltage deviation increment sequence.

[0033] The reactive capacity prediction component is obtained by multiplying the reactive capacity increment sequence with the first prediction weight, and the voltage deviation prediction component is obtained by multiplying the voltage deviation increment sequence with the second prediction weight. The reactive capacity prediction component and the voltage deviation prediction component are added together to obtain the reactive capacity prediction value. Based on the reactive capacity prediction value, the activation capacity and activation order of the standby reactive power compensation device are determined in ascending order of capacity, and the hierarchical activation instruction of the reactive power compensation device is generated.

[0034] When the duration of voltage fluctuation of the low-frequency component exceeds the low-frequency fluctuation time limit, the standby reactive power compensation device is activated step by step according to the tiered activation command. After each activation, the voltage recovery status is detected until the voltage fluctuation amplitude is reduced to the steady-state range.

[0035] A second aspect of the present invention provides an intelligent voltage regulation and control system for offshore wind power booster stations, comprising:

[0036] The first unit is used to decompose the collected offshore wind turbine output power into high-frequency disturbance components and low-frequency fluctuation components through wavelet transform, determine the fuzzy rule base of the adaptive fuzzy controller based on the high-frequency disturbance components, and calculate the required reserve capacity of the reactive power compensation device based on the changing trend of the low-frequency fluctuation components.

[0037] The second unit is used to divide the offshore wind power booster station into multiple voltage control sub-regions according to the electrical connection relationship. Based on the real-time collected voltage data and operation data, the voltage correlation between each sub-region is calculated to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions is greater than the voltage coupling degree threshold, the sub-region and its adjacent sub-regions are classified into a joint control group.

[0038] The third unit is used to design an adaptive coordinated control strategy for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with the minimization of voltage deviation and the minimization of the number of equipment switching. The search capability of the particle swarm is adjusted by adaptive inertia weight, and the weight coefficients of the optimization objectives are dynamically updated so that the control strategy is adaptively adjusted with the system operating state.

[0039] The fourth unit is used to quickly generate switching commands for reactive power compensation devices when voltage fluctuations are detected, for high-frequency disturbance components, by an adaptive fuzzy controller; for low-frequency fluctuation components, it puts standby reactive power compensation devices into operation according to the predicted capacity demand; and at the same time executes the optimal control scheme obtained by the adaptive coordinated control strategy.

[0040] A third aspect of the present invention,

[0041] An electronic device is provided, comprising:

[0042] processor;

[0043] Memory used to store processor-executable instructions;

[0044] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0045] Fourth aspect of the embodiments of the present invention,

[0046] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0047] The beneficial effects of this invention are as follows:

[0048] This invention decomposes the output power of offshore wind turbines using wavelet transform, enabling separate processing of high-frequency disturbances and low-frequency fluctuations. This improves the accuracy and response speed of voltage control, effectively reduces grid voltage fluctuations, and enhances system stability.

[0049] By dividing the voltage control sub-regions and constructing a voltage coupling matrix, joint control of areas with strong electrical correlations is achieved, avoiding the problem of mutual interference between adjacent control areas, solving the voltage oscillation phenomenon caused by traditional independent control strategies, and improving the overall control coordination.

[0050] An adaptive coordinated control strategy is designed using particle swarm optimization algorithm, and the control target weights are dynamically adjusted through adaptive inertial weights. This enables the control strategy to intelligently adjust according to the system operating status, minimizing the number of equipment switching operations while ensuring voltage quality, thus extending equipment lifespan and reducing system maintenance costs. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating the intelligent voltage regulation and control method for offshore wind power booster stations according to an embodiment of the present invention.

[0052] Figure 2 Heatmap of voltage coupling matrix for offshore wind power booster stations;

[0053] Figure 3 A schematic diagram showing the comparison of the voltage deviation improvement effect of the adaptive coordinated control strategy. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0056] Figure 1 This is a flowchart illustrating the intelligent voltage regulation and control method for offshore wind power booster stations according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0057] The collected offshore wind turbine output power is decomposed into high-frequency disturbance components and low-frequency fluctuation components through wavelet transform. The fuzzy rule base of the adaptive fuzzy controller is determined based on the high-frequency disturbance components, and the required reserve capacity of the reactive power compensation device is calculated based on the changing trend of the low-frequency fluctuation components.

[0058] Offshore wind power booster stations are divided into multiple voltage control sub-regions according to their electrical connections. The voltage correlation between each sub-region is calculated based on real-time collected voltage and operational data to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions is greater than the voltage coupling degree threshold, the sub-region and its adjacent sub-regions are classified into a joint control group.

[0059] An adaptive coordinated control strategy is designed for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with minimizing voltage deviation and minimizing the number of equipment switching. The search capability of the particle swarm is adjusted by adaptive inertia weight, and the weight coefficients of the optimization objectives are dynamically updated so that the control strategy can be adaptively adjusted with the system operating state.

[0060] When voltage fluctuations are detected, the adaptive fuzzy controller quickly generates switching commands for reactive power compensation devices for high-frequency disturbance components and puts standby reactive power compensation devices into operation according to the predicted capacity demand for low-frequency fluctuation components; at the same time, the optimal control scheme obtained by the adaptive coordinated control strategy is executed.

[0061] In one optional implementation, the collected offshore wind turbine output power is decomposed into high-frequency disturbance components and low-frequency fluctuation components using wavelet transform. The fuzzy rule base of the adaptive fuzzy controller is determined based on the high-frequency disturbance components, and the required reserve capacity of the reactive power compensation device is calculated based on the changing trend of the low-frequency fluctuation components.

[0062] Wavelet coefficients are obtained by high-pass filtering and low-pass filtering of the power output signal using wavelet basis functions, and then high-frequency disturbance components and low-frequency fluctuation components are obtained by orthogonal transformation and reconstruction of the wavelet coefficients.

[0063] The disturbance amplitude, disturbance frequency, and disturbance duration are calculated based on the high-frequency disturbance components. The disturbance amplitude, disturbance frequency, and disturbance duration are then classified into levels. A trapezoidal membership function is used to map the disturbance amplitude, disturbance frequency, and disturbance duration to membership values ​​between zero and one. An IF-THEN form fuzzy rule table is generated based on the levels and the membership values ​​to obtain the fuzzy rule base of the adaptive fuzzy controller.

[0064] Extract the data sequence of the low-frequency fluctuation component within a preset time window, and calculate the fluctuation mean, fluctuation standard deviation, and trend slope; use the fluctuation mean, fluctuation standard deviation, and trend slope to perform an autoregressive moving average prediction to obtain the predicted capacity, and take the maximum value of the product of the predicted capacity and the fluctuation standard deviation, and the product of the rated capacity and the preset ratio as the reserve capacity.

[0065] In an embodiment of the present invention, the power output data of the offshore wind turbine is first collected. This data is typically recorded in time series form, for example, power data is continuously collected for 24 hours at a sampling interval of 0.1 seconds, forming a time series of 864,000 data points.

[0066] The collected power output data was decomposed using wavelet transform. An appropriate wavelet basis function, such as the db4 wavelet, was selected to perform multi-scale decomposition of the power signal. During the decomposition process, the number of decomposition levels was set to 5. High-pass and low-pass filters in the wavelet transform were used to extract the detail coefficients and approximation coefficients of the signal, respectively. The wavelet coefficients obtained by high-pass filtering represent the high-frequency components of the signal, while the wavelet coefficients obtained by low-pass filtering represent the low-frequency components.

[0067] The obtained wavelet coefficients are subjected to orthogonal transformation to ensure that the decomposed signal components are mutually orthogonal and to eliminate redundant information. Signal reconstruction is then performed, reconstructing the high-frequency wavelet coefficients into high-frequency disturbance components and the low-frequency wavelet coefficients into low-frequency fluctuation components. For example, for a 5MW wind turbine, the high-frequency disturbance component typically fluctuates within ±0.5MW, while the low-frequency fluctuation component exhibits a relatively smooth trend, with an amplitude range between 0-4.5MW.

[0068] Based on the extracted high-frequency disturbance components, disturbance characteristic parameters are calculated. The disturbance amplitude is calculated, which is the difference between the maximum and minimum values ​​of the high-frequency components; for example, the disturbance amplitude calculated at a certain moment is 0.32 MW. The disturbance frequency is calculated by performing time-frequency analysis on the high-frequency components to determine the frequency of the main disturbance component, for example, 1.5 Hz. The disturbance duration is calculated, which is the duration for which continuous disturbances exceed a preset threshold (e.g., 0.1 MW), for example, 0.8 seconds.

[0069] The calculated disturbance amplitude, frequency, and duration were categorized into five levels: extremely small (0-0.1 MW), relatively small (0.1-0.3 MW), moderate (0.3-0.5 MW), relatively large (0.5-0.8 MW), and extremely large (>0.8 MW). The disturbance frequency was categorized into five levels: extremely low (0-0.5 Hz), relatively low (0.5-1.0 Hz), moderate (1.0-2.0 Hz), relatively high (2.0-3.0 Hz), and extremely high (>3.0 Hz). The disturbance duration was categorized into five levels: instantaneous (0-0.5 s), brief (0.5-1.0 s), moderate (1.0-2.0 s), relatively long (2.0-5.0 s), and persistent (>5.0 s).

[0070] The perturbation characteristic parameters are mapped to membership values ​​between 0 and 1 using a trapezoidal membership function. For example, for a perturbation amplitude of 0.32 MW, the membership is 0.9 in the "small" category and 0.1 in the "medium" category. For a perturbation frequency of 1.5 Hz, the membership is 0.75 in the "medium" category and 0.25 in the "low" category. For a perturbation duration of 0.8 seconds, the membership is 0.6 in the "short" category and 0.4 in the "medium" category.

[0071] Based on the calculated rank and membership values, an IF-THEN fuzzy rule table is generated. For example: "IF Disturbance Amplitude = Medium AND Disturbance Frequency = Medium AND Disturbance Duration = Short THEN Control Gain = Medium". By combining different input conditions and output responses, a complete fuzzy rule base is constructed, containing approximately 125 (5×5×5) rules. These rules will serve as the decision basis for the adaptive fuzzy controller.

[0072] Simultaneously, for low-frequency fluctuation components, data sequences within a preset time window are extracted for analysis. The time window is set to 10 minutes, with a sliding step of 1 minute. Fluctuation characteristic parameters are calculated for the data within the window. The fluctuation mean is calculated, i.e., the average value of the low-frequency components within the window, for example, 2.8 MW. The fluctuation standard deviation is calculated, reflecting the fluctuation amplitude, for example, 0.45 MW. The trend slope is calculated, representing the power change trend, for example, 0.02 MW / min, indicating a slow upward trend in power.

[0073] The mean of fluctuation, standard deviation of fluctuation, and trend slope are input into an autoregressive moving average prediction model to predict future wind power changes. This model predicts future values ​​using a weighted average based on historical data. For example, based on current fluctuation characteristics, the predicted capacity for the next 10 minutes is 3.1 MW, with a predicted standard deviation of 0.5 MW.

[0074] Calculate the required reserve capacity for the reactive power compensation system. Calculate a reference value by multiplying the predicted capacity by the standard deviation of the fluctuation, for example, 3.1MW × 0.5 = 1.55Mvar. Simultaneously, consider another reference value as the product of the rated capacity and the preset ratio; for example, if the wind turbine's rated capacity is 5MW and the preset ratio is 0.4, then the reference value is 5MW × 0.4 = 2Mvar. Take the larger of the two reference values ​​as the final reserve capacity, which is 2Mvar in this example.

[0075] Based on the determined reserve capacity, the reactive power compensation device is activated and deactivated to ensure the voltage stability of the wind turbine generators. When the system detects voltage fluctuations, the reactive power controller dynamically adjusts the control parameters according to the rules in the adaptive fuzzy rule base and the current disturbance characteristics, and outputs corresponding control signals to enable the reactive power compensation device to provide or absorb an appropriate amount of reactive power, thereby maintaining the grid voltage within the allowable range.

[0076] The method provided by this invention enables offshore wind turbines to intelligently adjust reactive power control strategies based on the real-time characteristics of wind power output, thereby improving grid voltage stability and rationally configuring the capacity of reactive power compensation equipment to avoid resource waste.

[0077] In one optional implementation, the offshore wind power booster station is divided into multiple voltage control sub-regions according to electrical connections. Voltage correlation between sub-regions is calculated based on real-time collected voltage and operational data to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions exceeds a voltage coupling threshold, that sub-region and its adjacent sub-regions are grouped into a joint control group, including:

[0078] A topological correlation matrix is ​​established based on the bus connection relationship of the offshore wind power booster station. The direct connection relationship between corresponding sub-regions is mapped to the element values ​​of the topological correlation matrix. Based on the topological correlation matrix, the offshore wind power booster station is divided into multiple voltage control sub-regions.

[0079] Voltage and operating data of each voltage control sub-region are collected in real time. The voltage correlation coefficient is obtained by weighting the ratio of voltage fluctuation amplitude and phase difference between adjacent voltage control sub-regions within a preset time window. The power transmission coefficient is obtained by calculating the ratio of the maximum active power transmitted between adjacent voltage control sub-regions to the rated capacity of the sub-region. The impedance coupling coefficient is obtained by multiplying the reciprocal of the line impedance between adjacent voltage control sub-regions by the line voltage level. The voltage correlation coefficient, the power transmission coefficient, and the impedance coupling coefficient are weighted according to preset weights to generate a voltage coupling degree matrix.

[0080] The voltage coupling degree threshold is calculated based on the non-zero elements in the voltage coupling degree matrix. It is then determined whether the voltage coupling degree between any voltage control sub-region and its adjacent voltage control sub-region is greater than the voltage coupling degree threshold. If so, the voltage control sub-region and its adjacent voltage control sub-region are merged into a joint control group.

[0081] In this embodiment, a topological correlation matrix is ​​first established based on the busbar connection relationships of the offshore wind power booster station. Taking a typical offshore wind farm as an example, the wind farm includes one 220kV main busbar and three 35kV sub-busbars, denoted as M0, M1, M2, and M3, respectively. The connection relationships between the buses can be expressed as follows: M0 is directly connected to M1, M2, and M3; M1, M2, and M3 are not directly connected. Based on this, a 4×4 topological correlation matrix T is established. When busbar i is directly connected to busbar j, T(i,j)=1; otherwise, T(i,j)=0. In the resulting matrix, T(0,1)=T(1,0)=T(0,2)=T(2,0)=T(0,3)=T(3,0)=1, and all other elements are 0. Based on this topological correlation matrix, the offshore wind power booster station can be divided into four voltage control sub-regions A0, A1, A2 and A3, which correspond to the M0, M1, M2 and M3 buses and their connecting equipment, respectively.

[0082] Furthermore, voltage and operational data are collected in real time for the divided voltage control sub-regions. Specifically, several measurement points are set in each sub-region, with a collection period of 100 milliseconds. The collected data includes voltage amplitude, phase angle, active power, reactive power, and equipment operating status. In practical applications, the data collected at a certain moment is as follows: the average voltage of sub-region A0 is 220.5kV, with a phase of 0 degrees; the average voltage of A1 is 35.2kV, with a phase of -1.2 degrees; the average voltage of A2 is 35.1kV, with a phase of -1.5 degrees; and the average voltage of A3 is 34.9kV, with a phase of -1.8 degrees.

[0083] Next, the voltage correlation coefficients of adjacent voltage-controlled sub-regions are calculated. A 10-minute time window is selected, with a total of 6000 sampling points. The voltage fluctuation of each adjacent sub-region within this window is calculated. For example, within this time window, the voltage fluctuation amplitude of sub-regions A0 and A1 is ±0.5kV, and that of A1 is ±0.3kV, with a fluctuation amplitude ratio of 0.6. The phase difference weighting value is 0.8 (calculated based on the cosine of the phase difference). The combined voltage correlation coefficient between A0 and A1 is 0.68. Similarly, the voltage correlation coefficient between A0 and A2 is calculated to be 0.65, and the voltage correlation coefficient between A0 and A3 is 0.62.

[0084] Simultaneously, the power transfer coefficient between adjacent voltage-controlled sub-regions is calculated. Through data analysis, the maximum active power transfer from A0 to A1 is 50MW, and the rated capacity of sub-region A1 is 100MW; therefore, the power transfer coefficient is 0.5. Similarly, the power transfer coefficient from A0 to A2 is 0.45, and the power transfer coefficient from A0 to A3 is 0.4.

[0085] In addition, the impedance coupling coefficient between adjacent voltage-controlled sub-regions is calculated. Taking the connection line between A0 and A1 as an example, with a line impedance of 0.2 ohms and a line voltage level of 220kV, the calculated impedance coupling coefficient is 220 / 0.2 = 1100. Similarly, the impedance coupling coefficient between A0 and A2 is 1000, and the impedance coupling coefficient between A0 and A3 is 900.

[0086] The three coefficients mentioned above are weighted according to preset weights to generate a voltage coupling matrix C. Let the weights of the voltage correlation coefficient, power transmission coefficient, and impedance coupling coefficient be 0.4, 0.3, and 0.3, respectively. For A0 and A1, the voltage coupling C(0,1) = C(1,0) = 0.4 × 0.68 + 0.3 × 0.5 + 0.3 × (1100 / 1200) = 0.272 + 0.15 + 0.275 = 0.697. Similarly, C(0,2) = C(2,0) = 0.646, C(0,3) = C(3,0) = 0.601, and all other elements are 0.

[0087] The voltage coupling threshold is calculated based on the non-zero elements in the voltage coupling matrix. Specifically, the average value of all non-zero coupling values ​​is multiplied by an adjustment coefficient of 0.9, i.e., threshold = (0.697 + 0.646 + 0.601) / 3 × 0.9 = 0.583.

[0088] By comparing the calculated voltage coupling values ​​with the threshold, it was determined whether each sub-region needed to form a joint control group. The comparison results showed that the voltage coupling between A0 and A1 (0.697) was greater than the threshold (0.583), the voltage coupling between A0 and A2 (0.646) was greater than the threshold (0.583), and the voltage coupling between A0 and A3 (0.601) was greater than the threshold (0.583). Therefore, A0, A1, A2, and A3 were all merged into a single joint control group.

[0089] In practical applications, the division of joint control groups can be dynamically adjusted based on changes in voltage coupling. For example, when line impedance changes or power transmission weakens, recalculating the voltage coupling can cause some sub-regions to no longer belong to the same joint control group. For instance, after a certain operational adjustment, if the voltage coupling between A0 and A3 drops to 0.55, below the threshold of 0.583, then A3 can form an independent control group, while A0, A1, and A2 remain within the same joint control group.

[0090] By dividing offshore wind power booster stations into different joint control groups using the above method, targeted voltage control strategies can be implemented for each joint control group, improving the accuracy and response speed of voltage regulation. For example, a coordinated control strategy can be adopted for joint control groups with high coupling, while a local autonomous control strategy can be used for independent sub-regions. This control group division method based on voltage coupling effectively solves the voltage coordination control problem in the group control process of offshore wind farms, improving overall voltage stability and control efficiency.

[0091] Figure 2This is a heatmap of the voltage coupling degree matrix for an offshore wind power booster station, illustrating the coupling strength distribution among the eight voltage control sub-regions proposed in this invention. The horizontal and vertical axes represent the target and source sub-regions, respectively, including A0 (220kV main bus), A1-A3 (35kV transformers 1# to 3#), B0 (220kV standby bus), B1-B2 (35kV transformers 4# to 5#), and C0 (110kV interconnection bus). The color intensity indicates the magnitude of the voltage coupling degree; the higher the value, the darker the color. The graph clearly shows that the coupling degree between A0 and B0 reaches 0.752, between A0 and A1 is 0.697, and between B0 and C0 is 0.678. These values ​​all exceed the coupling degree threshold of 0.583 (marked with an asterisk in the graph), indicating the need for a joint control group for coordinated control. In contrast, the coupling between some sub-regions is relatively low, such as only 0.189 between A1 and A3, and 0.385 between A0 and B2, indicating that the voltage influence between these regions is small. This heatmap effectively verifies the scientific validity of the present invention's method of calculating voltage coupling by comprehensively considering voltage correlation coefficient, power transmission coefficient, and impedance coupling coefficient. It provides a visualized decision-making basis for the dynamic zoning control of offshore wind power booster stations, demonstrating the algorithm's innovative ability to accurately identify voltage coupling relationships under complex power grid topologies.

[0092] In one optional implementation, an adaptive coordinated control strategy is designed for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with minimizing voltage deviation and minimizing the number of equipment switching operations. The search capability of the particle swarm is adjusted through adaptive inertia weights, and the weight coefficients of the optimization objectives are dynamically updated, so that the control strategy adaptively adjusts with the system operating state, including:

[0093] The voltage fluctuation trend is obtained by calculating the change in voltage deviation between the current time and the previous control time of each node in the joint control group, and the equipment adjustment frequency is obtained by counting the cumulative number of actions of transformer taps and reactive power compensation devices within the preset time window.

[0094] The voltage control priority coefficient is dynamically updated based on the voltage fluctuation trend, and the switching constraint coefficient is dynamically updated based on the equipment adjustment frequency. The product of the mean of the sum of squares of voltage deviations of each node in the joint control group and the voltage control priority coefficient is used as the voltage control sub-objective. The result of adding the change in transformer tap position and the change in switching status of reactive power compensation device at adjacent control times and multiplying it with the switching constraint coefficient is used as the switching count sub-objective. The voltage control sub-objective and the switching count sub-objective are added together to obtain the adaptive optimization objective.

[0095] The particle swarm search iteration count is initialized to zero. A maximum iteration count threshold is set, and the ratio of the iteration count to the maximum iteration count threshold is calculated. The ratio is then substituted into the exponential decay function to obtain the adaptive search weight. Multiple sets of particle position vectors composed of transformer tap positions and reactive power compensation device switching states are generated.

[0096] The adaptive search weights are combined with the historical optimal solution and the global optimal solution of the particle position vector to update the particle position vector. The adaptive optimization target value corresponding to the updated particle position vector is calculated. The number of iterations is increased to obtain the optimal transformer tap position and reactive power compensation device switching state.

[0097] The distribution network control system first collects voltage data from each node within the joint control group. It records the voltage value V(t) at the current time t and the voltage value V(t-1) at the previous control time t-1, and calculates the voltage deviation change ΔV = V(t) - V(t-1). If ΔV is positive, it indicates that the voltage at that node is trending upwards; if ΔV is negative, it indicates that the voltage is trending downwards. For example, if the voltage at a distribution network node was 10.2kV at the previous control time and is currently 10.5kV, then the voltage deviation change is 0.3kV, showing an upward trend.

[0098] At the same time, a preset time window T (usually 24 hours) is set, and the number of transformer tap adjustments N within the window is counted. tap and the number of times the reactive power compensation device is switched on and off (N) cap If a transformer has had its tap adjusted 5 times in the past 24 hours, and a capacitor bank has been switched on and off 4 times, then the adjustment frequency data N is obtained. tap =5 and N cap =4.

[0099] The voltage control priority coefficient α is dynamically updated based on voltage fluctuation trends. When the voltage fluctuation trend is significant (large voltage deviation), the value of α is increased; conversely, it is decreased. For example, when the voltage rise exceeds 5% of the rated voltage, the value of α can be set to 0.8; when the voltage is stable, the value of α can be set to 0.6. The switching constraint coefficient β is dynamically updated based on the equipment adjustment frequency. When the cumulative number of equipment actions approaches a preset threshold (e.g., the daily adjustment frequency of a transformer approaches 10 times), the value of β is increased to limit excessive equipment action; conversely, the value of β is decreased. For example, when the transformer adjustment frequency reaches 8 times, the value of β can be set to 0.7; when the adjustment frequency is only 2 times, the value of β can be set to 0.3.

[0100] Assuming there are n nodes in the joint control group, calculate the voltage deviation δV of each node. i (i=1,2,...,n), which is the difference between the actual voltage and the reference voltage. For example, if the actual voltage at a node is 10.5kV and the reference voltage is 10.0kV, then δV i=0.5kV. Calculate the mean value J of the sum of squares of voltage deviations. v =((δV1) 2 +(δV2) 2 +...+(δV n ) 2 Multiplying this by the voltage control priority coefficient α, we obtain the voltage control sub-target α·J. v .

[0101] Calculate the changes in equipment state at adjacent control moments: the change in transformer tap position |Tap(t)-Tap(t-1)| and the change in reactive power compensation device switching state |Cap(t)-Cap(t-1)|. Multiply their sum by the switching constraint coefficient β to obtain the sub-objective J of the number of switches. s =β·(|Tap(t)-Tap(t-1)|+|Cap(t)-Cap(t-1)|). For example, if a transformer tap is adjusted from 5 to 7, and a capacitor bank is switched from the connected state to the disconnected state, and the switching constraint coefficient β is 0.5, then J s =0.5·(|7-5|+|0-1|)=1.5.

[0102] Adding the voltage control sub-objective and the switching count sub-objective, we obtain the adaptive optimization objective J = α·J v +β·(|Tap(t)-Tap(t-1)|+|Cap(t)-Cap(t-1)|).

[0103] The execution process of the particle swarm optimization algorithm begins with the initial iteration count iter = 0, and a maximum iteration count maxIter is set (usually 50~100). The iteration progress ratio r = iter / maxIter is calculated, and r is substituted into the exponential decay function w = w max -r·(w max -w min Calculate the adaptive search weight w, where w max and w min These are the upper and lower limits of the weight (usually 0.9 and 0.4). For example, if the current iteration number is 20 and the maximum iteration number is 100, then r = 0.2, and the adaptive search weight w = 0.9 - 0.2 * (0.9 - 0.4) = 0.8.

[0104] m groups of particles are randomly generated (usually m is 20~30). Each particle represents a control scheme, including the transformer tap position vector Tap and the reactive power compensation device switching state vector Cap. For example, a particle is represented by the position vector X=[5,1], which means that the transformer tap position is 5 and the capacitor bank is in the working state (value is 1).

[0105] In each iteration, the velocity vector V of each particle is updated. i and position vector X i V i =w·V i +c1·r1·(P i -X i )+c2·r2·(P g -X i ), X i =X i +V i , where P i P represents the optimal position in the particle's history. g For the globally optimal position, c1 and c2 are learning factors (usually both set to 2.0), and r1 and r2 are random numbers between 0 and 1. The updated particle position vector needs to be converted to integers and limited to the adjustment range allowed by the device.

[0106] Calculate the corresponding adaptive optimization objective value J based on the updated particle position vector, and update the historical best position P of each particle. i and the global optimal position P g Then, increment the iteration count by 1 (iter = iter + 1) and continue the search process until the maximum number of iterations is reached.

[0107] After completing the iteration, output the globally optimal position P. g As the final control strategy, this includes the optimal tap position of the transformer. opt And the optimal switching state of the reactive power compensation device Cap opt The control system then issues control commands accordingly. In practical applications, if the voltage of a feeder is too high, the algorithm provides the optimal solution to adjust the transformer tap from level 3 to level 2, while simultaneously disconnecting part of the capacitor bank, in order to achieve a balance between voltage control targets and equipment lifespan.

[0108] Through the method of this invention, the power distribution network can adaptively adjust the control strategy according to the real-time operating status, effectively balance the relationship between voltage quality and equipment life, and improve the operating efficiency and reliability of the power distribution network.

[0109] Figure 3This diagram illustrates the comparison of voltage deviation improvement effects of the adaptive coordinated control strategy proposed in this invention with three existing technologies. The horizontal axis represents control time (0-50 hours), and the vertical axis represents the average voltage deviation percentage (0-8%). The four curves represent the performance of this invention, traditional voltage reactive power control, genetic algorithm optimization, and fuzzy logic control methods, respectively. The diagram clearly shows that the proposed method (circles) initially exhibits a voltage deviation of 6.6%, which decreases rapidly over time, eventually stabilizing at a low level of 0.7%. In contrast, the traditional voltage reactive power control method (rectangles) maintains a consistently high voltage deviation, decreasing only slightly from an initial 7.3% to 4.5%. The genetic algorithm optimization method (triangles) performs moderately, stabilizing at 1.7%. The fuzzy logic control method (diamonds) shows relatively limited improvement, with a final voltage deviation of 3.0%. The comparison results fully demonstrate that the present invention, through innovative mechanisms such as dynamically adjusting weight coefficients and adaptive inertial weights, can significantly improve the voltage control accuracy of the distribution network, achieve the optimal balance between voltage quality and equipment protection, and reflect the superiority and practical value of the algorithm.

[0110] In one optional implementation, when voltage fluctuations are detected, the adaptive fuzzy controller rapidly generates switching commands for the reactive power compensation device for high-frequency disturbance components, including:

[0111] The voltage deviation fuzzy sub-interval is divided based on the fluctuation amplitude of the high-frequency component, and the voltage deviation change rate fuzzy sub-interval is divided based on the fluctuation frequency of the high-frequency component.

[0112] The voltage deviation and voltage deviation rate of change are input into the Gaussian membership function, respectively. Based on the fuzzy sub-intervals of the voltage deviation and voltage deviation rate of change, the membership degree of each sub-interval is calculated. The sub-interval corresponding to the highest membership degree is selected as the membership value of the voltage deviation and the voltage deviation rate of change. The fuzzy membership degree is then multiplied by the membership value of the voltage deviation and the voltage deviation rate of change to obtain the comprehensive fuzzy membership value.

[0113] The reference gain coefficient is adaptively adjusted according to the real-time fluctuation amplitude of the high-frequency component. The product of the voltage fluctuation amplitude and the reference gain coefficient is used as the dynamic control gain. The fuzzy membership degree comprehensive value is multiplied by the dynamic control gain to obtain the switching amount of the reactive power compensation device.

[0114] When the voltage fluctuation amplitude of the high-frequency component is detected to be greater than the high-frequency fluctuation threshold, the switching order and switching time of each group of reactive power compensation devices are calculated based on the switching amount of the reactive power compensation device. The switching operation of the reactive power compensation devices is performed one group at a time according to the switching order. After each group is switched, the voltage recovery is detected. When the voltage fluctuation amplitude is reduced to below the high-frequency fluctuation threshold, the switching is stopped.

[0115] This invention provides a technical solution for rapidly generating reactive power compensation device switching commands in response to high-frequency disturbance components when voltage fluctuations are detected. In power systems, voltage fluctuations can adversely affect the stability of electrical equipment and the system, especially high-frequency disturbance components which require rapid response and compensation. Therefore, this invention employs an adaptive fuzzy controller to achieve rapid response and precise control of high-frequency disturbances.

[0116] In actual power system operation, voltage fluctuations can be divided into low-frequency and high-frequency components. This invention focuses on the high-frequency disturbance component and achieves rapid compensation control through adaptive fuzzy control technology. In implementation, the high-frequency disturbance component in the voltage signal is first extracted using methods such as wavelet transform or bandpass filter, decomposing the voltage signal into components of different frequency bands, and extracting the high-frequency fluctuation component, which is usually above 10Hz.

[0117] To address the extracted high-frequency disturbance components, this invention constructs an adaptive fuzzy control system. Two input variables are designed: voltage deviation and voltage deviation rate of change. Voltage deviation is defined as the difference between the actual voltage value and the rated voltage value, while the voltage deviation rate of change is the rate of change of voltage deviation over time. Based on the fluctuation amplitude of the high-frequency components, the voltage deviation is divided into seven fuzzy sub-intervals: negative large (NB), negative medium (NM), negative small (NS), zero (ZO), positive small (PS), positive medium (PM), and positive large (PB). For example, for a 10kV system, the voltage deviation can be divided into: NB [-0.6kV, -0.4kV], NM [-0.4kV, -0.2kV], NS [-0.2kV, 0kV], ZO [-0.05kV, 0.05kV], PS [0kV, 0.2kV], PM [0.2kV, 0.4kV], and PB [0.4kV, 0.6kV].

[0118] Similarly, based on the fluctuation frequency of the high-frequency component, the voltage deviation change rate is also divided into seven corresponding fuzzy sub-intervals. For example, for a system with a change rate in the range of ±10kV / s, it can be divided into: NB[-10kV / s,-7kV / s], NM[-7kV / s,-4kV / s], NS[-4kV / s,0kV / s], ZO[-1kV / s,1kV / s], PS[0kV / s,4kV / s], PM[4kV / s,7kV / s], PB[7kV / s,10kV / s].

[0119] The membership degree of voltage deviation and the rate of change of voltage deviation is calculated using a Gaussian membership function. For a voltage deviation value x, its membership degree μA(x) in the fuzzy sub-interval A is calculated using a Gaussian function. For example, for a voltage deviation of -0.3kV, the calculated membership degree is 0.85 in the NM interval, 0.25 in the NS interval, and close to 0 in other intervals. Therefore, the maximum membership degree of 0.85 is selected, and the membership degree value of the voltage deviation is determined to be 0.85, belonging to the NM sub-interval. Similarly, for the rate of change of voltage deviation of -5kV / s, the calculated membership degree is 0.7 in the NM interval, and the membership degree value of the rate of change of voltage deviation is 0.7, belonging to the NM sub-interval.

[0120] Multiplying the membership value of the voltage deviation by the membership value of the rate of change of voltage deviation yields the fuzzy membership composite value. In the example above, the fuzzy membership composite value is 0.85 × 0.7 = 0.595.

[0121] The core innovation of this invention lies in the adaptive adjustment of the reference gain coefficient. The reference gain coefficient K0 is automatically adjusted based on the real-time fluctuation amplitude of the high-frequency component. When the fluctuation amplitude increases, K0 increases accordingly; when the fluctuation amplitude decreases, K0 decreases accordingly. Specifically, a reference fluctuation amplitude A0 and a corresponding reference gain coefficient K0 are set, and the real-time fluctuation amplitude is A. Then, the dynamic reference gain coefficient K = K0 × (A / A0). For example, if the reference fluctuation amplitude A0 = 0.2kV and the reference gain coefficient K0 = 100kVar / kV are set, when a real-time fluctuation amplitude A = 0.3kV is detected, the dynamic reference gain coefficient K = 100 × (0.3 / 0.2) = 150kVar / kV.

[0122] Multiplying the fuzzy membership value by the dynamic control gain yields the switching amount of the reactive power compensation device. In the example above, the switching amount of the reactive power compensation device = 0.595 × 150 = 89.25 kVar. Assuming the system is configured with multiple sets of reactive power compensation devices with capacities of 50 kVar, 30 kVar, 20 kVar, and 10 kVar respectively, based on the calculated switching amount of 89.25 kVar, the switching order is determined as follows: first the 50 kVar group, then the 30 kVar group, and finally the 10 kVar group, for a total of 90 kVar.

[0123] The high-frequency fluctuation threshold is set to 2% of the system's rated voltage, which is 0.2kV for a 10kV system. When the voltage fluctuation amplitude of the high-frequency component is detected to be greater than 0.2kV, the switching operation of the reactive power compensation device is executed. According to the determined switching sequence, the 50kVar group of compensation devices is switched on first. After switching on, wait 100ms and check the voltage fluctuation again. If the voltage fluctuation amplitude decreases from 0.3kV to 0.22kV, but is still higher than the threshold of 0.2kV, then switch on the 30kVar group of compensation devices. Wait another 100ms. If the voltage fluctuation amplitude decreases to 0.16kV, which is lower than the threshold of 0.2kV, the switching operation is stopped.

[0124] In actual operation, the system continuously monitors the grid voltage. When the high-frequency disturbance disappears or the system voltage returns to normal, the reactive power compensation devices are gradually disconnected based on the voltage recovery status. The disconnection sequence is the reverse of the activation sequence: smaller capacity groups are disconnected first, followed by larger capacity groups. After each disconnection, the voltage is checked to ensure it remains within the allowable range.

[0125] The aforementioned adaptive fuzzy control strategy enables rapid calculation of the required reactive power compensation for high-frequency disturbance components and precise switching control of the reactive power compensation device. This effectively suppresses voltage fluctuations in the power grid, improving power stability and power quality. Practical applications show that this method improves response speed by 40% and compensation accuracy by 25% compared to traditional control strategies, meeting the requirements for rapid response to high-frequency voltage fluctuations.

[0126] In one optional implementation, for low-frequency fluctuation components, putting standby reactive power compensation devices into operation according to predicted capacity demand includes:

[0127] The reactive capacity increment sequence is obtained by calculating the reactive capacity difference between adjacent sampling times within a preset time window, and the voltage deviation increment sequence is obtained by calculating the voltage deviation difference between adjacent sampling times within the preset time window. The first prediction weight is adaptively adjusted according to the changing trend of the reactive capacity increment sequence, and the second prediction weight is adaptively adjusted according to the changing trend of the voltage deviation increment sequence.

[0128] The reactive capacity prediction component is obtained by multiplying the reactive capacity increment sequence with the first prediction weight, and the voltage deviation prediction component is obtained by multiplying the voltage deviation increment sequence with the second prediction weight. The reactive capacity prediction component and the voltage deviation prediction component are added together to obtain the reactive capacity prediction value. Based on the reactive capacity prediction value, the activation capacity and activation order of the standby reactive power compensation device are determined in ascending order of capacity, and the hierarchical activation instruction of the reactive power compensation device is generated.

[0129] When the duration of voltage fluctuation of the low-frequency component exceeds the low-frequency fluctuation time limit, the standby reactive power compensation device is activated step by step according to the tiered activation command. After each activation, the voltage recovery status is detected until the voltage fluctuation amplitude is reduced to the steady-state range.

[0130] In one embodiment, during the operation of the power system, voltage and reactive power data are collected through distributed measurement devices. The collection frequency can be set to once every 100ms to ensure accurate understanding of the system's operating status. The collected data undergoes preliminary processing, including data normalization and outlier filtering, to ensure the accuracy of subsequent analysis.

[0131] The processed data is stored in the system's sliding time window, with a window length that can be set to 10 seconds, containing 100 sampling points. Based on the data within this window, the reactive capacity increment sequence and voltage deviation increment sequence are calculated. Specifically, for sampling times t and t-1, the reactive capacity difference Q(t) - Q(t-1) between the two times is calculated as the reactive capacity increment ΔQ(t) at time t; similarly, the voltage deviation difference V(t) - V(t-1) between time t and t-1 is calculated as the voltage deviation increment ΔV(t) at time t. The continuous ΔQ(t) and ΔV(t) are then used to construct the reactive capacity increment sequence and voltage deviation increment sequence, respectively.

[0132] Based on the changing trends of these two sequences, the prediction weights are adaptively adjusted. Specifically, the average value Avg of the reactive power capacity increment sequence over the most recent 30 sampling points is calculated. ΔQ and standard deviation Std ΔQ When Avg ΔQ When Avg is positive and shows an increasing trend, it indicates that the reactive power demand of the system is increasing, and the first prediction weight w1 will increase accordingly; when Avg ΔQ When the value is negative and shows a downward trend, it indicates that the reactive power demand of the system is decreasing, and w1 will decrease accordingly. The adjustment range of the weight w1 is from 0.5 to 0.9, and the specific value is determined based on the correlation between the current reactive power capacity increment and historical data. For example, when it is found that the average reactive power capacity increment of the most recent 30 sampling points is 5 Mvar / s and shows a stable growth trend, w1 will be set to 0.8.

[0133] Similarly, the average value Avg of the voltage deviation increment sequence over the most recent 30 sampling points is calculated. ΔV and standard deviation Std ΔV When Avg ΔV When the absolute value of is large and continues to increase, it indicates that the voltage deviation is widening, and the second prediction weight w2 will increase accordingly; when Avg ΔVWhen the voltage is close to zero and fluctuates little, it indicates that the voltage state is relatively stable, and w2 will decrease accordingly. The adjustment range of the weight w2 is from 0.3 to 0.7. For example, when the average increment of the voltage deviation over the most recent 30 sampling points is detected to be -0.02 pu / s and continues to decrease, w2 will be set to 0.6.

[0134] After determining the weights, the reactive power capacity prediction value is calculated. The latest value ΔQ(t) of the reactive power capacity increment sequence is multiplied by the first prediction weight w1 to obtain the reactive power capacity prediction component w1×ΔQ(t); the latest value ΔV(t) of the voltage deviation increment sequence is multiplied by the second prediction weight w2 and converted to the corresponding reactive power capacity unit to obtain the voltage deviation prediction component w2×ΔV(t)×k, where k is a conversion coefficient, determined based on characteristics, and can be taken as 100 Mvar / pu. These two prediction components are added together to obtain the reactive power capacity prediction value Q. pred = w1×ΔQ(t) + w2×ΔV(t)×k.

[0135] For example, if the measured reactive power capacity increment is 3 Mvar and the voltage deviation increment is -0.015 pu at a certain moment, the first prediction weight is 0.8, the second prediction weight is 0.6, and the conversion coefficient k is 100 Mvar / pu, then the calculated reactive power capacity prediction value is: 0.8×3 + 0.6×(-0.015)×100 = 2.4 - 0.9 = 1.5 Mvar.

[0136] Based on the calculated reactive power capacity forecast, the activation capacity and activation sequence of standby reactive power compensation devices are determined. Assume there are multiple standby reactive power compensation devices with capacities of 1 Mvar, 2 Mvar, 5 Mvar, and 10 Mvar, respectively. The activation scheme is determined according to the predicted reactive power demand, in ascending order of capacity. For the predicted demand of 1.5 Mvar, a 1 Mvar compensation device is selected as the first level, and a 2 Mvar compensation device as the second level, generating corresponding tiered activation instructions.

[0137] The duration of low-frequency voltage fluctuations also needs to be determined. When the detected duration of low-frequency voltage fluctuations exceeds the preset low-frequency fluctuation time limit (e.g., 3 seconds), the standby reactive power compensation devices are activated step by step according to the tiered activation instructions. After each level of compensation device is activated, a 200ms wait is waited before the voltage recovery is checked. If the voltage fluctuation amplitude has decreased to within the steady-state range (e.g., ±5% of the nominal voltage), activation is stopped; otherwise, the next level of compensation device is activated until the voltage returns to normal.

[0138] In a practical application, a substation experienced voltage fluctuations, with a low-frequency component fluctuation detected lasting for 4 seconds, predicting a reactive power compensation requirement of 2.5 Mvar. Initially, a 1 Mvar compensation device was activated. After waiting 200ms, it was found that the voltage fluctuation still exceeded the steady-state range, so a 2 Mvar compensation device was activated. At this point, the total activated capacity was 3 Mvar, exceeding the predicted requirement. The voltage fluctuation amplitude quickly decreased to within the steady-state range, and further activation of the compensation device was stopped.

[0139] This prediction-based tiered input method effectively avoids the problems of overcompensation or undercompensation in traditional methods, improves voltage stability and power quality, and also extends the service life of reactive power compensation devices.

[0140] This invention provides an embodiment of intelligent voltage regulation and control for offshore wind power booster stations, comprising:

[0141] The first unit is used to decompose the collected offshore wind turbine output power into high-frequency disturbance components and low-frequency fluctuation components through wavelet transform, determine the fuzzy rule base of the adaptive fuzzy controller based on the high-frequency disturbance components, and calculate the required reserve capacity of the reactive power compensation device based on the changing trend of the low-frequency fluctuation components.

[0142] The second unit is used to divide the offshore wind power booster station into multiple voltage control sub-regions according to the electrical connection relationship. Based on the real-time collected voltage data and operation data, the voltage correlation between each sub-region is calculated to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions is greater than the voltage coupling degree threshold, the sub-region and its adjacent sub-regions are classified into a joint control group.

[0143] The third unit is used to design an adaptive coordinated control strategy for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with the minimization of voltage deviation and the minimization of the number of equipment switching. The search capability of the particle swarm is adjusted by adaptive inertia weight, and the weight coefficients of the optimization objectives are dynamically updated so that the control strategy is adaptively adjusted with the operating state.

[0144] The fourth unit is used to quickly generate switching commands for reactive power compensation devices when voltage fluctuations are detected, for high-frequency disturbance components, by an adaptive fuzzy controller; for low-frequency fluctuation components, it puts standby reactive power compensation devices into operation according to the predicted capacity demand; and at the same time executes the optimal control scheme obtained by the adaptive coordinated control strategy.

[0145] A third aspect of the present invention provides an electronic device, comprising:

[0146] processor;

[0147] Memory used to store processor-executable instructions;

[0148] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0149] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0150] This invention can be a method, apparatus, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0151] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent voltage regulation and control of offshore wind power booster stations, characterized in that, include: The collected offshore wind turbine output power is decomposed into high-frequency disturbance components and low-frequency fluctuation components through wavelet transform. The fuzzy rule base of the adaptive fuzzy controller is determined based on the high-frequency disturbance components, and the required reserve capacity of the reactive power compensation device is calculated based on the changing trend of the low-frequency fluctuation components. Offshore wind power booster stations are divided into multiple voltage control sub-regions according to their electrical connections. The voltage correlation between each sub-region is calculated based on real-time collected voltage and operational data to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions is greater than the voltage coupling degree threshold, the sub-region and its adjacent sub-regions are classified into a joint control group. Based on the particle swarm optimization algorithm, an adaptive coordinated control strategy is designed for the joint control group. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with the minimization of voltage deviation and the minimization of the number of equipment switching. The search capability of the particle swarm is adjusted by adaptive inertial weighting, and the weight coefficients of the target are dynamically updated and optimized, so that the control strategy is adaptively adjusted according to the system operating state. This includes: calculating the change in voltage deviation between the current time and the previous control time of each node in the joint control group to obtain the voltage fluctuation trend, and counting the cumulative number of actions of transformer taps and reactive power compensation devices within the preset time window to obtain the equipment adjustment frequency. The voltage control priority coefficient is dynamically updated based on the voltage fluctuation trend, and the switching constraint coefficient is dynamically updated based on the equipment adjustment frequency. The product of the mean of the sum of squares of voltage deviations of each node in the joint control group and the voltage control priority coefficient is used as the voltage control sub-objective. The result of adding the change in transformer tap position and the change in switching status of reactive power compensation device at adjacent control times and multiplying it with the switching constraint coefficient is used as the switching count sub-objective. The voltage control sub-objective and the switching count sub-objective are added together to obtain the adaptive optimization objective. The particle swarm search iteration count is initialized to zero. A maximum iteration count threshold is set, and the ratio of the iteration count to the maximum iteration count threshold is calculated. The ratio is then substituted into the exponential decay function to obtain the adaptive search weight. Multiple sets of particle position vectors composed of transformer tap positions and reactive power compensation device switching states are generated. The adaptive search weight is combined with the historical optimal solution and the global optimal solution of the particle position vector to update the particle position vector. The adaptive optimization target value corresponding to the updated particle position vector is calculated. The number of iterations is increased to obtain the optimal transformer tap position and reactive power compensation device switching state. When voltage fluctuations are detected, the adaptive fuzzy controller quickly generates switching commands for reactive power compensation devices for high-frequency disturbance components and puts standby reactive power compensation devices into operation according to the predicted capacity demand for low-frequency fluctuation components; at the same time, the optimal control scheme obtained by the adaptive coordinated control strategy is executed.

2. The method according to claim 1, characterized in that, The collected offshore wind turbine output power is decomposed into high-frequency disturbance components and low-frequency fluctuation components using wavelet transform. The fuzzy rule base for the adaptive fuzzy controller is determined based on the high-frequency disturbance components. Based on the changing trend of the low-frequency fluctuation components, the required reserve capacity for the reactive power compensation device is calculated, including: Wavelet coefficients are obtained by high-pass filtering and low-pass filtering of the power output signal using wavelet basis functions, and then high-frequency disturbance components and low-frequency fluctuation components are obtained by orthogonal transformation and reconstruction of the wavelet coefficients. The disturbance amplitude, disturbance frequency, and disturbance duration are calculated based on the high-frequency disturbance components. The disturbance amplitude, disturbance frequency, and disturbance duration are then classified into levels. A trapezoidal membership function is used to map the disturbance amplitude, disturbance frequency, and disturbance duration to membership values ​​between zero and one. An IF-THEN form fuzzy rule table is generated based on the levels and the membership values ​​to obtain the fuzzy rule base of the adaptive fuzzy controller. Extract the data sequence of the low-frequency fluctuation component within a preset time window, and calculate the fluctuation mean, fluctuation standard deviation, and trend slope; use the fluctuation mean, fluctuation standard deviation, and trend slope to perform an autoregressive moving average prediction to obtain the predicted capacity, and take the maximum value of the product of the predicted capacity and the fluctuation standard deviation, and the product of the rated capacity and the preset ratio as the reserve capacity.

3. The method according to claim 1, characterized in that, Offshore wind power booster stations are divided into multiple voltage control sub-regions based on their electrical connections. Voltage correlations between sub-regions are calculated based on real-time collected voltage and operational data, generating a voltage coupling matrix. When the voltage coupling between any sub-region and its adjacent sub-regions exceeds a voltage coupling threshold, that sub-region and its adjacent sub-regions are grouped into a joint control group, including: A topological correlation matrix is ​​established based on the bus connection relationship of the offshore wind power booster station. The direct connection relationship between corresponding sub-regions is mapped to the element values ​​of the topological correlation matrix. Based on the topological correlation matrix, the offshore wind power booster station is divided into multiple voltage control sub-regions. Voltage and operating data of each voltage control sub-region are collected in real time. The voltage correlation coefficient is obtained by weighting the ratio of voltage fluctuation amplitude and phase difference between adjacent voltage control sub-regions within a preset time window. The power transmission coefficient is obtained by calculating the ratio of the maximum active power transmitted between adjacent voltage control sub-regions to the rated capacity of the sub-region. The impedance coupling coefficient is obtained by multiplying the reciprocal of the line impedance between adjacent voltage control sub-regions by the line voltage level. The voltage correlation coefficient, the power transmission coefficient, and the impedance coupling coefficient are weighted according to preset weights to generate a voltage coupling degree matrix. The voltage coupling degree threshold is calculated based on the non-zero elements in the voltage coupling degree matrix. It is then determined whether the voltage coupling degree between any voltage control sub-region and its adjacent voltage control sub-region is greater than the voltage coupling degree threshold. If so, the voltage control sub-region and its adjacent voltage control sub-region are merged into a joint control group.

4. The method according to claim 1, characterized in that, When voltage fluctuations are detected, the adaptive fuzzy controller quickly generates switching commands for the reactive power compensation device, targeting high-frequency disturbance components, including: The voltage deviation fuzzy sub-interval is divided based on the voltage fluctuation amplitude of the high-frequency component, and the voltage deviation change rate fuzzy sub-interval is divided based on the fluctuation frequency of the high-frequency component. The voltage deviation and voltage deviation rate of change are input into the Gaussian membership function, respectively. The membership degree of each sub-interval is calculated based on the fuzzy sub-interval of the voltage deviation and the fuzzy sub-interval of the voltage deviation rate of change. The sub-interval corresponding to the maximum membership degree is selected as the membership degree value of the voltage deviation and the membership degree value of the voltage deviation rate of change. The membership degree value of the voltage deviation and the membership degree value of the voltage deviation rate of change are multiplied to obtain the fuzzy membership degree composite value. The reference gain coefficient is adaptively adjusted according to the real-time fluctuation amplitude of the high-frequency component. The product of the voltage fluctuation amplitude and the reference gain coefficient is used as the dynamic control gain. The fuzzy membership comprehensive value is multiplied by the dynamic control gain to obtain the switching amount of the reactive power compensation device. When the voltage fluctuation amplitude of the high-frequency component is detected to be greater than the high-frequency fluctuation threshold, the switching order and switching time of each group of reactive power compensation devices are calculated based on the switching amount of the reactive power compensation device. The switching operation of the reactive power compensation devices is performed one group at a time according to the switching order. After each group is switched, the voltage recovery is detected. When the voltage fluctuation amplitude is reduced to below the high-frequency fluctuation threshold, the switching is stopped.

5. The method according to claim 4, characterized in that, To address low-frequency fluctuations, the standby reactive power compensation devices will be put into operation according to the predicted capacity demand, including: The reactive capacity increment sequence is obtained by calculating the reactive capacity difference between adjacent sampling times within a preset time window, and the voltage deviation increment sequence is obtained by calculating the voltage deviation difference between adjacent sampling times within the preset time window. The first prediction weight is adaptively adjusted according to the changing trend of the reactive capacity increment sequence, and the second prediction weight is adaptively adjusted according to the changing trend of the voltage deviation increment sequence. The reactive capacity prediction component is obtained by multiplying the reactive capacity increment sequence with the first prediction weight, and the voltage deviation prediction component is obtained by multiplying the voltage deviation increment sequence with the second prediction weight. The reactive capacity prediction component and the voltage deviation prediction component are added together to obtain the reactive capacity prediction value. Based on the reactive capacity prediction value, the activation capacity and activation order of the standby reactive power compensation device are determined in ascending order of capacity, and the hierarchical activation instruction of the reactive power compensation device is generated. When the duration of voltage fluctuation of the low-frequency fluctuation component is greater than the low-frequency fluctuation time limit, the standby reactive power compensation device is activated step by step according to the tiered activation instruction. After each activation, the voltage recovery status is detected until the voltage fluctuation amplitude is reduced to the steady-state range.

6. An intelligent voltage regulation and control system for offshore wind power booster stations, used to implement the method as described in any one of claims 1-5, characterized in that, include: The first unit is used to decompose the collected offshore wind turbine output power into high-frequency disturbance components and low-frequency fluctuation components through wavelet transform, determine the fuzzy rule base of the adaptive fuzzy controller based on the high-frequency disturbance components, and calculate the required reserve capacity of the reactive power compensation device based on the changing trend of the low-frequency fluctuation components. The second unit is used to divide the offshore wind power booster station into multiple voltage control sub-regions according to the electrical connection relationship. Based on the real-time collected voltage data and operation data, the voltage correlation between each sub-region is calculated to generate a voltage coupling degree matrix. When the voltage coupling degree between any sub-region and its adjacent sub-regions is greater than the voltage coupling degree threshold, the sub-region and its adjacent sub-regions are classified into a joint control group. The third unit is used to design an adaptive coordinated control strategy for the joint control group based on the particle swarm optimization algorithm. The transformer tap position and the switching state of the reactive power compensation device are used as optimization variables. A multi-objective optimization function is constructed with the minimization of voltage deviation and the minimization of the number of equipment switching. The search capability of the particle swarm is adjusted by adaptive inertia weight, and the weight coefficients of the optimization objectives are dynamically updated so that the control strategy is adaptively adjusted with the system operating state. The fourth unit is used to quickly generate switching commands for reactive power compensation devices when voltage fluctuations are detected, for high-frequency disturbance components, by an adaptive fuzzy controller; for low-frequency fluctuation components, it puts standby reactive power compensation devices into operation according to the predicted capacity demand; and at the same time executes the optimal control scheme obtained by the adaptive coordinated control strategy.

7. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 5.

Images