Offshore substation lightning protection intelligent control method and system

Abstract

The application provides a lightning protection intelligent control method and system for a marine booster station, relates to the technical field of control, and comprises the following steps: collecting various data of a lightning arrester through a sensor array, performing Fourier transform on voltage and current data to calculate power loss distribution, constructing an electric field interference degree matrix, and generating an optimal spacing position vector of the lightning arrester; extracting the fluctuation characteristics of atmospheric electric field intensity, calculating the polarization direction and intensity of the electric field, and dynamically adjusting the angle of the lightning arrester and the opening and closing timing; and evaluating the working state of the lightning arrester and correcting the control instruction. The application realizes intelligent and accurate control of lightning protection for the marine booster station, and effectively improves the lightning protection efficiency and safety.

Description

Technical Field

[0001] This invention relates to the field of control technology, and in particular to an intelligent control method and system for lightning protection of offshore substations. Background Technology

[0002] Offshore substations are a key component of the power transmission system for offshore wind farms. They primarily collect electricity generated by wind turbines and boost the voltage for long-distance transmission to the onshore power grid. Due to their location in the offshore environment, they face unique meteorological conditions, particularly frequent lightning strikes, which pose a serious threat to their safe and stable operation. Lightning activity at sea is characterized by high intensity and frequency; inadequate lightning protection measures can lead to equipment damage, system failure, and even fires, resulting in significant economic losses.

[0003] Traditional lightning protection for offshore substations mainly relies on fixed surge arrester installations and passive protection measures, such as lightning rods, lightning protection strips, and grounding grids. These methods are largely based on experience-based design and lack precise consideration of the unique characteristics of the offshore environment. As offshore wind farms expand into deeper waters and the capacity of individual turbines continues to increase, traditional lightning protection technologies face numerous challenges. Summary of the Invention

[0004] This invention provides an intelligent control method and system for lightning protection of offshore substations, which can solve the problems in the prior art.

[0005] A first aspect of the present invention provides an intelligent control method for lightning protection of offshore substations, comprising:

[0006] Data on surge arrester voltage, surge arrester current, atmospheric electric field strength, surge arrester temperature, and surge arrester grounding resistance of the offshore substation are collected using a sensor array.

[0007] Fourier transform is performed on the surge arrester voltage data and surge arrester current data to calculate the power loss distribution. An electric field interference matrix of the surge arrester is constructed based on the electromagnetic field finite element model. The power loss distribution and electric field interference matrix are input into an adaptive neural network to generate the optimal spacing position vector of the surge arrester. The position of the surge arrester is adjusted according to the optimal spacing position vector.

[0008] The atmospheric electric field intensity data is filtered to extract the electric field intensity fluctuation characteristics, and the electric field polarization direction and electric field polarization intensity are calculated; a surge arrester response characteristic model is established based on the electric field polarization direction and electric field polarization intensity, and the optimal conduction path and optimal conduction time are calculated; the polarization angle of the surge arrester is dynamically adjusted based on the optimal conduction path, and the opening and closing of the surge arrester is controlled based on the optimal conduction time;

[0009] The surge arrester temperature data and surge arrester grounding resistance data are input into the state evaluation model to calculate the surge arrester operating state parameters. Based on the surge arrester operating state parameters, the position adjustment command, the angle adjustment command, and the opening and closing control command are corrected.

[0010] Real-time monitoring of surge arrester operation data; when lightning overvoltage occurs, recording of lightning protection response time and protection effectiveness data.

[0011] Fourier transforms are performed on the surge arrester voltage and current data to calculate the power loss distribution. Based on the electromagnetic field finite element model, the surge arrester electric field interference matrix is ​​constructed, including:

[0012] Perform Fourier transform on the surge arrester voltage data and surge arrester current data to obtain the phase spectrum and amplitude spectrum;

[0013] Based on the phase spectrum and amplitude spectrum, the active power and reactive power components of each frequency band of the surge arrester are calculated, and a power loss distribution vector is established.

[0014] A three-dimensional electromagnetic field finite element model of the surge arrester is established, mesh elements are divided, and an electromagnetic field characteristic parameter matrix is ​​constructed. The electric field intensity distribution is calculated based on the electromagnetic field characteristic parameter matrix to obtain electric field interference data at different spatial locations. The electric field interference data is then organized into an electric field interference matrix.

[0015] The power loss distribution vector is mapped to the electric field interference matrix to establish a correlation model between power loss and electric field interference; based on the correlation model, the mutual influence coefficient between surge arresters is calculated, and surge arrester layout optimization suggestions are generated.

[0016] The power loss distribution and electric field interference matrix are input into an adaptive neural network to generate an optimal spacing vector for the surge arrester; adjusting the surge arrester position based on the optimal spacing vector includes:

[0017] An adaptive neural network is constructed, with the power loss distribution as the first input layer data and the electric field interference matrix as the second input layer data; the adaptive neural network is trained based on the backpropagation algorithm, with the optimal spacing data in the historical operation data of the surge arrester as training samples; when the training error of the adaptive neural network is less than a preset error threshold, the model training is completed.

[0018] The real-time power loss distribution and electric field interference matrix are input into the trained adaptive neural network model to calculate the optimal spacing position vector of the surge arrester group; based on the optimal spacing position vector, a spatial arrangement model of the surge arresters is constructed to generate the target coordinates of each surge arrester.

[0019] Calculate the deviation between the current position of each surge arrester and the target coordinates. When the position deviation exceeds the preset deviation threshold, trigger the position adjustment command.

[0020] Based on the position adjustment command, the surge arrester position adjustment device is controlled to move each surge arrester to the target coordinate position.

[0021] Based on the electric field polarization direction and the electric field polarization intensity, a surge arrester response characteristic model is established, and the optimal conduction path and optimal conduction time are calculated, including:

[0022] The electric field polarization direction is mapped to an azimuth vector, and the electric field polarization intensity is converted into an electric field gradient matrix; three-dimensional spatial electric field polarization distribution data are constructed based on the azimuth vector and the electric field gradient matrix.

[0023] The charge density distribution and potential gradient in the three-dimensional spatial electric field polarization distribution data are calculated to obtain the lightning channel initiation probability distribution; multiple conductive paths are simulated using the Monte Carlo method based on the lightning channel initiation probability distribution; the resistance loss and electromagnetic induction characteristics of each conductive path are calculated to determine the optimal conductive path with the minimum energy loss.

[0024] The change law of electric field polarization intensity with time is calculated based on the optimal conduction path to obtain the optimal response timing of the surge arrester; the optimal conduction time is determined according to the optimal response timing.

[0025] Inputting the surge arrester temperature data and surge arrester grounding resistance data into the state evaluation model, calculating the surge arrester operating state parameters, and correcting the position adjustment command, the angle adjustment command, and the opening / closing control command based on the surge arrester operating state parameters includes:

[0026] Time-frequency analysis is performed on the surge arrester temperature data to extract temperature change trend characteristics and temperature fluctuation period; statistical analysis is performed on the surge arrester grounding resistance data to calculate the grounding resistance change rate and stability coefficient; and the surge arrester operating status parameters are calculated based on the temperature change trend characteristics, temperature fluctuation period, grounding resistance change rate, and stability coefficient.

[0027] The temperature correction coefficient and grounding correction coefficient are calculated based on the operating state parameters of the surge arrester; when the temperature correction coefficient exceeds the preset temperature threshold, the execution speed of the position adjustment command is reduced according to the preset ratio; when the grounding correction coefficient exceeds the preset grounding threshold, the execution angle of the angle adjustment command is adjusted; the execution sequence of the opening and closing control command is determined according to the combined state of the temperature correction coefficient and the grounding correction coefficient.

[0028] The revised position adjustment command, angle adjustment command, and opening / closing control command are sent to the surge arrester execution system.

[0029] A second aspect of the present invention provides an intelligent control system for lightning protection of offshore substations, comprising:

[0030] The first unit is used to collect surge arrester voltage data, surge arrester current data, atmospheric electric field strength data, surge arrester temperature data, and surge arrester grounding resistance data of the offshore substation through a sensor array.

[0031] The second unit is used to perform Fourier transform on the surge arrester voltage data and surge arrester current data, calculate the power loss distribution, construct the surge arrester electric field interference matrix based on the electromagnetic field finite element model; input the power loss distribution and electric field interference matrix into an adaptive neural network to generate the optimal spacing position vector of the surge arrester; and adjust the position of the surge arrester according to the optimal spacing position vector.

[0032] The third unit is used to filter the atmospheric electric field strength data to extract the electric field strength fluctuation characteristics, and calculate the electric field polarization direction and electric field polarization intensity; establish a surge arrester response characteristic model based on the electric field polarization direction and electric field polarization intensity, calculate the optimal conduction path and optimal conduction time; dynamically adjust the polarization angle of the surge arrester based on the optimal conduction path, and control the opening and closing of the surge arrester based on the optimal conduction time;

[0033] The fourth unit is used to input the surge arrester temperature data and surge arrester grounding resistance data into the state evaluation model, calculate the surge arrester operating state parameters, and correct the position adjustment command, the angle adjustment command, and the opening and closing control command based on the surge arrester operating state parameters.

[0034] The fifth unit is used to monitor the operation data of the surge arrester in real time. When lightning overvoltage occurs, it records the lightning protection response time and protection effect data.

[0035] A third aspect of the embodiments of the present invention,

[0036] An electronic device is provided, comprising:

[0037] processor;

[0038] Memory used to store processor-executable instructions;

[0039] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0040] Fourth aspect of the present invention,

[0041] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0042] The beneficial effects of this application are as follows:

[0043] By collecting key data from all directions using a sensor array and processing and analyzing it using advanced algorithms such as Fourier transform, electromagnetic field finite element model and adaptive neural network, the intelligent optimization and adjustment of the lightning arrester position is realized, which significantly improves the lightning protection efficiency and reliability of offshore substations.

[0044] By accurately processing atmospheric electric field intensity data, extracting electric field fluctuation characteristics, calculating electric field polarization direction and intensity, and establishing a response characteristic model, dynamic adjustment of surge arrester angle and precise control of opening and closing time are realized. This enables the lightning protection system to make optimal responses to different lightning conditions, greatly improving the adaptability and protection effect of the lightning protection system.

[0045] By real-time monitoring and evaluation of surge arrester temperature and grounding resistance data, and by correcting adjustment commands based on operating status parameters, a closed-loop control mechanism is formed to ensure stable operation of the system in various complex marine environments. At the same time, it records lightning events and protection effects in real time, providing data support for continuous system optimization and significantly extending the service life of offshore booster station equipment. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating the intelligent control method for lightning protection of offshore substations according to an embodiment of the present invention. Detailed Implementation

[0047] 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.

[0048] 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.

[0049] Figure 1 This is a flowchart illustrating the intelligent control method for lightning protection of offshore substations according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0050] Data on surge arrester voltage, surge arrester current, atmospheric electric field strength, surge arrester temperature, and surge arrester grounding resistance of the offshore substation are collected using a sensor array.

[0051] Fourier transform is performed on the surge arrester voltage data and surge arrester current data to calculate the power loss distribution. An electric field interference matrix of the surge arrester is constructed based on the electromagnetic field finite element model. The power loss distribution and electric field interference matrix are input into an adaptive neural network to generate the optimal spacing position vector of the surge arrester. The position of the surge arrester is adjusted according to the optimal spacing position vector.

[0052] The atmospheric electric field intensity data is filtered to extract the electric field intensity fluctuation characteristics, and the electric field polarization direction and electric field polarization intensity are calculated; a surge arrester response characteristic model is established based on the electric field polarization direction and electric field polarization intensity, and the optimal conduction path and optimal conduction time are calculated; the polarization angle of the surge arrester is dynamically adjusted based on the optimal conduction path, and the opening and closing of the surge arrester is controlled based on the optimal conduction time;

[0053] The surge arrester temperature data and surge arrester grounding resistance data are input into the state evaluation model to calculate the surge arrester operating state parameters. Based on the surge arrester operating state parameters, the position adjustment command, the angle adjustment command, and the opening and closing control command are corrected.

[0054] Real-time monitoring of surge arrester operation data; when lightning overvoltage occurs, recording of lightning protection response time and protection effectiveness data.

[0055] In one optional implementation, Fourier transform is performed on the surge arrester voltage data and surge arrester current data to calculate the power loss distribution. The surge arrester electric field interference matrix is ​​then constructed based on the electromagnetic field finite element model, including:

[0056] Perform Fourier transform on the surge arrester voltage data and surge arrester current data to obtain the phase spectrum and amplitude spectrum;

[0057] Based on the phase spectrum and amplitude spectrum, the active power and reactive power components of each frequency band of the surge arrester are calculated, and a power loss distribution vector is established.

[0058] A three-dimensional electromagnetic field finite element model of the surge arrester is established, mesh elements are divided, and an electromagnetic field characteristic parameter matrix is ​​constructed. The electric field intensity distribution is calculated based on the electromagnetic field characteristic parameter matrix to obtain electric field interference data at different spatial locations. The electric field interference data is then organized into an electric field interference matrix.

[0059] The power loss distribution vector is mapped to the electric field interference matrix to establish a correlation model between power loss and electric field interference; based on the correlation model, the mutual influence coefficient between surge arresters is calculated, and surge arrester layout optimization suggestions are generated.

[0060] In this embodiment, to optimize surge arrester layout and reduce electromagnetic interference, a technical solution based on voltage and current data analysis and electromagnetic field simulation is proposed. This solution first performs Fourier transform on the surge arrester's voltage and current data, calculates the power loss distribution through frequency domain analysis, then establishes a finite element model of the electromagnetic field to evaluate the electric field interference level, and finally generates optimized surge arrester layout suggestions.

[0061] The specific implementation process of performing Fourier transform on surge arrester voltage and current data includes: acquiring time-domain data of voltage and current from surge arresters in a 110kV substation, with a sampling frequency of 20kHz and a sampling duration of 100ms. The acquired data is stored as digital signals via a data acquisition card. A Fast Fourier Transform (FFT) algorithm is used to transform these time-domain data, converting the time-domain signals into a frequency-domain representation. For the acquired voltage data v(t) and current data i(t), the voltage amplitude spectrum |V(f)| and phase spectrum φv(f), and the current amplitude spectrum |I(f)| and phase spectrum φi(f) are calculated, respectively. For example, at the fundamental frequency of 50Hz, the measured voltage amplitude is 8.5kV with a phase of 0 degrees; the current amplitude is 5.3mA with a phase of -27 degrees. At 150Hz, the voltage amplitude is 0.42kV with a phase of 30 degrees; the current amplitude is 0.85mA with a phase of -15 degrees.

[0062] Based on the obtained phase and amplitude spectrum data, the active and reactive power components of the surge arrester in each frequency band are calculated. For each frequency point f, the active power is calculated as the product of the voltage amplitude and the current amplitude multiplied by the cosine of the phase difference; the reactive power is calculated as the product of the voltage amplitude and the current amplitude multiplied by the sine of the phase difference. Taking the 50Hz fundamental frequency as an example, the calculated active power is 39.85W and the reactive power is 12.25Var; at a frequency of 150Hz, the active power is 0.32W and the reactive power is 0.08Var. The power data of each frequency point are organized into a power loss distribution vector P=[P1, P2, ...,Pn], where n represents the number of frequency points analyzed. In practical applications, n=40 is taken to cover the frequency range from 50Hz to 2000Hz.

[0063] When establishing a three-dimensional electromagnetic field finite element model of a surge arrester, a geometric model is first constructed based on the arrester's physical structure and material properties. A typical surge arrester is 1.5 meters high and 0.3 meters in diameter, consisting of a metal oxide varistor, an insulating bushing, and metal accessories. The conductivity of the metal oxide varistor is set to vary nonlinearly according to the electric field strength, the relative permittivity of the insulating bushing material is set to 5.6, and the loss tangent is 0.003. Tetrahedral elements are used to mesh the model, resulting in approximately 85,000 elements. The minimum mesh size is 2 mm to ensure sufficient mesh density in critical areas such as the varistor edges. The constructed electromagnetic field characteristic parameter matrix includes information such as the material properties, boundary conditions, and initial field distribution of each mesh node.

[0064] Based on the aforementioned electromagnetic field characteristic parameter matrix, the electric field distribution around the surge arrester was solved using the finite element method. During the calculation, a measured voltage was applied to the high-voltage end of the surge arrester, while the low-voltage end was grounded. The boundary value problem was solved to obtain the electric field intensity distribution E(x,y,z). The calculated electric field intensity results show that the maximum electric field intensity is 3.2 kV / m at a distance of 10 cm from the surge arrester surface, and decreases to 0.5 kV / m at a distance of 1 m. The calculated electric field intensity data at different spatial locations were organized into an electric field interference matrix D, with dimensions of 100×100×50, representing the electric field distribution in the three-dimensional space of x,y,z.

[0065] When mapping the power loss distribution vector to the electric field interference matrix, a spatial mapping algorithm is used to establish a correlation model between power loss and electric field interference. Through analysis of power loss and corresponding electric field distribution at 40 different frequency points, a linear regression model is established to describe the relationship between power loss and electric field strength. The calculation results show that at the 50Hz fundamental frequency, for every 1W increase in power loss, the electric field strength at a distance of 1m from the surge arrester increases by an average of approximately 0.012kV / m. Based on this correlation model, the mutual influence coefficient matrix C between different surge arresters is calculated, where the element Cij represents the electric field interference coefficient of the i-th surge arrester on the j-th surge arrester.

[0066] Finally, based on the above analysis results, optimization recommendations for surge arrester arrangement were generated. For three-phase surge arresters, it is recommended that the phase-to-phase distance be no less than 2.5m to ensure that the mutual interference coefficient is less than 0.15; for surge arrester groups installed in parallel, it is recommended that the minimum distance between adjacent surge arresters be 1.8m. For the arrangement of multiple surge arresters in a substation, it is recommended that the minimum distance between different groups of surge arresters be 4m to reduce the impact of electromagnetic interference. In a practical application case, after adjusting the surge arrester arrangement according to the optimization recommendations in a 110kV substation, the electric field interference between surge arresters was reduced by 28%, and the system operation stability was significantly improved.

[0067] This technical solution, through a combination of Fourier analysis and electromagnetic field simulation, provides a theoretical basis and practical guidance for the scientific arrangement of surge arresters in substations, effectively improving the reliability and stability of power systems.

[0068] In one optional implementation, the power loss distribution and electric field interference matrix are input into an adaptive neural network to generate an optimal spacing position vector for the surge arrester; adjusting the position of the surge arrester according to the optimal spacing position vector includes:

[0069] An adaptive neural network is constructed, with the power loss distribution as the first input layer data and the electric field interference matrix as the second input layer data; the adaptive neural network is trained based on the backpropagation algorithm, with the optimal spacing data in the historical operation data of the surge arrester as training samples; when the training error of the adaptive neural network is less than a preset error threshold, the model training is completed.

[0070] The real-time power loss distribution and electric field interference matrix are input into the trained adaptive neural network model to calculate the optimal spacing position vector of the surge arrester group; based on the optimal spacing position vector, a spatial arrangement model of the surge arresters is constructed to generate the target coordinates of each surge arrester.

[0071] Calculate the deviation between the current position of each surge arrester and the target coordinates. When the position deviation exceeds the preset deviation threshold, trigger the position adjustment command.

[0072] Based on the position adjustment command, the surge arrester position adjustment device is controlled to move each surge arrester to the target coordinate position.

[0073] To implement this method, an adaptive neural network model must first be constructed. This network employs a dual-input layer structure. The first input layer receives power loss distribution data, and the second input layer receives electric field interference matrix data. Power loss distribution data is typically represented as a three-dimensional matrix. For example, in a 100m × 100m × 50m space, the space can be divided into 1m³ cubic units, and the power loss value of each unit can be recorded. In practical applications, the power loss distribution of a transmission line area might be discrete data points such as [0.15, 0.23, 0.35, 0.42, 0.28, 0.19] kW / m³. The electric field interference matrix reflects the degree of electric field interference experienced at the arrester's installation location. It is usually represented as a two-dimensional or three-dimensional matrix. For example, in the same area, the value at a certain location in the electric field interference matrix might be [0.65, 0.72, 0.58, 0.43, 0.39, 0.67], representing the interference intensity coefficient at different locations.

[0074] The neural network's hidden layers are designed as three layers, containing 64, 128, and 64 neurons respectively. The output layer is designed as a vector representing the spacing of the surge arrester group, with the vector dimension related to the number of surge arresters. For example, for a group containing 8 surge arresters, the output vector is 8×3 dimensional, representing the three-dimensional spatial coordinates of each surge arrester. The ReLU activation function is used to enhance the network's ability to handle nonlinear relationships.

[0075] The backpropagation algorithm was used to optimize the network weights during training. Optimal spacing configurations recorded in historical operational data were used as training samples; for example, the optimal spacing data for surge arresters recorded in a substation was [(15.2m, 25.3m, 0m), (30.5m, 25.3m, 0m), (45.8m, 25.3m, 0m)]. The initial learning rate was set to 0.001, and a learning rate decay strategy was adopted, multiplying the learning rate by 0.95 every 100 training epochs. The batch size was set to 64, and the number of training epochs was 1000. During training, the mean squared error between the predicted spacing and the actual optimal spacing was calculated. When the error was less than a preset threshold of 0.05, the model training was considered complete. In a practical application, after 734 training epochs, the training error reached 0.048, which was lower than the preset threshold, at which point the model parameters were saved.

[0076] After the model training is completed, the real-time monitored power loss distribution and electric field interference matrix are input into the network. Assuming that the monitored power loss distribution at a certain moment is [0.18, 0.27, 0.39, 0.45, 0.31, 0.22] kW / m³, and the electric field interference matrix is ​​[0.62, 0.70, 0.55, 0.40, 0.36, 0.64], the optimal spacing position vector of the surge arrester group is obtained after inputting it into the network as [(14.8m, 24.9m, 0m), (30.1m, 25.0m, 0m), (45.5m, 25.2m, 0m)].

[0077] Based on the obtained optimal spacing location vector, a spatial arrangement model of surge arresters is constructed. This model considers the direction of the transmission line, terrain features, and the location constraints of other fixed facilities. The target coordinates of each surge arrester are generated through a three-dimensional spatial interpolation algorithm. For example, the target coordinates of a group of surge arresters near a certain transmission line are finally calculated as [(102.5m, 85.7m, 12.3m), (117.3m, 85.9m, 12.3m), (132.8m, 86.1m, 12.4m)].

[0078] The system calculates the deviation between the current position of each surge arrester and the target coordinates in real time. When the position deviation of a surge arrester exceeds a preset deviation threshold, a position adjustment command is triggered. The deviation calculation uses Euclidean distance, and the preset deviation threshold is set to 0.5m. For example, if the current position of a surge arrester is detected as (103.2m, 85.9m, 12.3m), while the target coordinates are (102.5m, 85.7m, 12.3m), the calculated deviation is 0.82m, which exceeds the preset threshold of 0.5m, and the system will trigger a position adjustment command.

[0079] Position adjustment commands are transmitted to the surge arrester position adjustment device via a control interface. The adjustment device includes a precision electric mechanism and a position feedback system, enabling precise movement of the surge arrester in three-dimensional space. The adjustment process employs a step-by-step movement strategy, with each movement not exceeding 0.1m, followed by a 0.5-second pause for position confirmation. The position adjustment of a single surge arrester is completed when the deviation between the surge arrester's position and the target coordinates is less than 0.05m. After all surge arresters have been adjusted, the system records the new position configuration data and adds it to the historical operation database for future model training optimization.

[0080] Through actual operation testing, after applying this method to a 330kV substation, the lightning protection efficiency of the surge arrester system increased by 17.8%, the system power loss decreased by 12.5%, and the electric field interference decreased by 21.3%. Especially during the thunderstorm season, this method can adjust the position of the surge arrester in real time according to weather changes, effectively improving the safe operation level of the power system.

[0081] This method not only improves the protection efficiency of surge arrester systems but also reduces maintenance costs and failure rates. The application of adaptive neural networks enables the system to learn and optimize, continuously refining surge arrester layout strategies as operational data accumulates, thus providing strong support for the safe and stable operation of power systems.

[0082] In one optional implementation, a surge arrester response characteristic model is established based on the electric field polarization direction and the electric field polarization intensity, and the optimal conduction path and optimal conduction time are calculated, including:

[0083] The electric field polarization direction is mapped to an azimuth vector, and the electric field polarization intensity is converted into an electric field gradient matrix; three-dimensional spatial electric field polarization distribution data are constructed based on the azimuth vector and the electric field gradient matrix.

[0084] The charge density distribution and potential gradient in the three-dimensional spatial electric field polarization distribution data are calculated to obtain the lightning channel initiation probability distribution; multiple conductive paths are simulated using the Monte Carlo method based on the lightning channel initiation probability distribution; the resistance loss and electromagnetic induction characteristics of each conductive path are calculated to determine the optimal conductive path with the minimum energy loss.

[0085] The change law of electric field polarization intensity with time is calculated based on the optimal conduction path to obtain the optimal response timing of the surge arrester; the optimal conduction time is determined according to the optimal response timing.

[0086] In one embodiment, optimizing the surge arrester's response first requires establishing an accurate electric field polarization model. The system acquires real-time electric field monitoring data, including the electric field polarization direction and intensity. The electric field polarization direction is typically represented by angles, such as azimuth α and elevation β, while the electric field polarization intensity is measured in kilovolts per meter (kV / m). To transform this data into a computable model, the system maps the electric field polarization direction to an azimuth vector V(θ,φ), where θ represents the horizontal azimuth and φ represents the vertical elevation. For example, when the detected electric field polarization direction is 85 degrees horizontally and 40 degrees vertically, the corresponding azimuth vector V is represented as (85, 40). Simultaneously, the system converts the electric field polarization intensity into an electric field gradient matrix G, which describes the rate of change of the electric field intensity in three-dimensional space. When the detected electric field polarization intensity is 25 kV / m, the system can construct a 9×9×9 electric field gradient matrix. The value at the center of the matrix is ​​25 kV / m, and the surrounding points are distributed according to the electric field attenuation law. For example, the gradient value at a distance of 1 meter from the center point may be 23.5 kV / m.

[0087] Based on the azimuth vector V and the electric field gradient matrix G, the system constructs three-dimensional spatial electric field polarization distribution data D. This data structure uses a three-dimensional grid with a resolution of 0.1 meters, covering a spatial range of 100 meters × 100 meters × 200 meters around the surge arrester. Each grid point contains attributes such as electric field strength, direction, and rate of change over time at that location. For example, at the grid point with coordinates (10.5, 20.3, 45.7), the electric field strength might be 18.7 kV / m, the direction vector might be (0.35, 0.62, 0.70), and the rate of change over time might be 3.2 kV / m·s. Through interpolation calculations, the system can obtain the electric field polarization characteristics at any point in continuous space, forming a complete three-dimensional electric field polarization distribution model.

[0088] After modeling the electric field polarization distribution, the system calculates the charge density distribution and potential gradient in the three-dimensional spatial electric field polarization distribution data. The charge density ρ is obtained by calculating the divergence of the electric field gradient, while the potential gradient is obtained by calculating the potential difference between adjacent points in three-dimensional space. In practical applications, the system divides a 100-meter-high area into 1000 calculation units, and the charge density and potential gradient of each unit are calculated and stored in the database. For example, in a calculation unit at a height of 50 meters and a horizontal position (25, 30), the charge density might be 0.85 μC / m³, and the potential gradient might be 12.6 kV / m². Based on these data, the system derives the lightning channel initiation probability distribution P, which reflects the probability that each point in space will become the initiation point of a lightning strike. Regions with high electric field strength, high charge density, and regions with abrupt changes in potential gradient usually have a higher probability of lightning channel initiation.

[0089] Based on the probability distribution of lightning channel initiation, the system uses the Monte Carlo method to simulate multiple possible conductive paths. Specifically, the system randomly generates 10,000 lightning initiation points and calculates the discharge development direction and velocity at each point based on the electric field direction and intensity. Each simulated path contains a series of three-dimensional coordinate points, with adjacent points spaced 0.5 meters apart, extending from the cloud initiation point to the ground-based surge arrester. For example, a simulated path might contain the coordinate sequence: (0,5,100)→(0.4,4.8,95.5)→...→(15.2,10.3,0). The system performs electrical characteristic analysis on each simulated path, including calculating parameters such as path resistance, capacitance, and inductance.

[0090] To determine the optimal conduction path, the system calculates the energy loss characteristics of each path. Path resistance loss is obtained by integrating the product of the resistance of each conductor segment and the square of the current. For a lightning channel with a diameter of approximately 1 cm, its resistivity is approximately 0.05 ohms / meter. Electromagnetic induction characteristics consider the influence of the magnetic field generated around the lightning channel on the conductivity efficiency, including both self-inductance and mutual inductance. The system calculates the total energy loss value E for each simulated path, and the path with the minimum loss is determined as the optimal conduction path. In one example, among the 10,000 paths simulated by the system, path number 3752 has the minimum total energy loss of 125 kilojoules, and this path is determined as the optimal conduction path.

[0091] After obtaining the optimal conduction path, the system calculates the change in electric field polarization intensity over time based on this path. The system divides the lightning development process into 500 time steps, each 10 microseconds, and calculates the change in electric field intensity at each moment. For example, at t=100 microseconds, the average electric field intensity on the optimal conduction path may be 32.7 kV / m, while at t=150 microseconds it decreases to 28.3 kV / m. By analyzing the time series of electric field polarization intensity, the system identifies the inflection point before the electric field intensity reaches its peak as the optimal triggering time for conduction. In the example above, the system detects that the electric field intensity begins to rise sharply at t=220 microseconds. Activating the surge arrester at this time can guide the lightning to discharge along the optimal path to the maximum extent, avoiding damage to the protected target.

[0092] The system determines the optimal conduction time T of the surge arrester based on the aforementioned optimal response timing. In practical applications, the system advances this time by 5-10 microseconds as the surge arrester activation time to ensure that the ionization channel is established before the lightning strikes. After receiving the time command, the surge arrester response module precisely controls the high-voltage discharge circuit to trigger at the specified time, achieving active guidance and protection against lightning.

[0093] In one optional implementation, the surge arrester temperature data and surge arrester grounding resistance data are input into a state evaluation model to calculate surge arrester operating state parameters. The correction of the position adjustment command, the angle adjustment command, and the opening / closing control command based on the surge arrester operating state parameters includes:

[0094] Time-frequency analysis is performed on the surge arrester temperature data to extract temperature change trend characteristics and temperature fluctuation period; statistical analysis is performed on the surge arrester grounding resistance data to calculate the grounding resistance change rate and stability coefficient; and the surge arrester operating status parameters are calculated based on the temperature change trend characteristics, temperature fluctuation period, grounding resistance change rate, and stability coefficient.

[0095] The temperature correction coefficient and grounding correction coefficient are calculated based on the operating state parameters of the surge arrester; when the temperature correction coefficient exceeds the preset temperature threshold, the execution speed of the position adjustment command is reduced according to the preset ratio; when the grounding correction coefficient exceeds the preset grounding threshold, the execution angle of the angle adjustment command is adjusted; the execution sequence of the opening and closing control command is determined according to the combined state of the temperature correction coefficient and the grounding correction coefficient.

[0096] The revised position adjustment command, angle adjustment command, and opening / closing control command are sent to the surge arrester execution system.

[0097] Regarding the specific implementation of the surge arrester status assessment and control adjustment method, this invention discloses in detail the technical solution of inputting surge arrester temperature data and surge arrester grounding resistance data into the status assessment model, calculating the surge arrester operating status parameters, and correcting the position adjustment command, angle adjustment command, and opening and closing control command accordingly.

[0098] The system collects temperature data recorded every 5 minutes from the surge arrester's temperature sensor, forming a time series. Wavelet transform is applied to this time series to decompose the temperature data into different frequency components. Linear regression analysis is performed on the decomposed low-frequency components to extract the slope of the temperature change trend as a characteristic feature; for example, the temperature change trend of a surge arrester over 24 hours is an increase of 0.5℃ / hour. Simultaneously, fast Fourier transform is performed on the mid-to-high-frequency components to identify the main frequency components and determine the temperature fluctuation period; for example, the main period of the surge arrester's temperature fluctuation is 120 minutes.

[0099] The system performs statistical analysis on the grounding resistance data of surge arresters. Grounding resistance data is collected hourly, with a continuous 72-hour data sample. The percentage change between any two adjacent measurements is calculated, and the average value is taken as the grounding resistance change rate. For example, the average grounding resistance change rate of a surge arrester is 0.8% / hour. The coefficient of variation is obtained by calculating the ratio of the standard deviation to the average value of the grounding resistance data. Its reciprocal is used as the stability coefficient. For example, a stability coefficient of 5.2 indicates that the grounding resistance is relatively stable.

[0100] Based on the extracted temperature change trend characteristics, temperature fluctuation period, grounding resistance change rate, and stability coefficient, the system calculates the surge arrester's operating parameters using a weighted summation method. Specifically: the temperature change trend characteristics are weighted at 0.4, the reciprocal of the temperature fluctuation period at 0.2, the grounding resistance change rate at 0.25, and the stability coefficient at 0.15, and then summed after normalization. In a practical case, if the normalized value of the temperature change trend characteristics is 0.7, the normalized value of the reciprocal of the temperature fluctuation period is 0.5, the normalized value of the grounding resistance change rate is 0.6, and the normalized value of the stability coefficient is 0.8, then the surge arrester's operating parameters are calculated as: 0.4×0.7+0.2×0.5+0.25×0.6+0.15×0.8=0.65.

[0101] Based on the surge arrester's operating parameters, the system calculates the temperature correction factor and the grounding correction factor. The temperature correction factor is obtained by dividing the weighted average of the temperature change trend characteristics and the reciprocal of the temperature fluctuation period by a preset reference value of 0.5. In the above case, the temperature correction factor is (0.7×0.7+0.3×0.5) / 0.5=1.12. The grounding correction factor is calculated by dividing the weighted average of the grounding resistance change rate and the reciprocal of the stability coefficient by a preset reference value of 0.3. In this case, the grounding correction factor is (0.6×0.6+0.4×(1 / 0.8)) / 0.3=1.53.

[0102] When the temperature correction factor exceeds the preset temperature threshold of 1.1, the system reduces the execution speed of the position adjustment command by a preset ratio. Specifically, the reduction ratio is equal to (temperature correction factor - 1) × 50%. In the above example, the temperature correction factor is 1.12, exceeding the preset threshold of 1.1, therefore the execution speed needs to be reduced by (1.12 - 1) × 50% = 6%. For example, if the original position adjustment command was to move 10 millimeters per second, the correction will reduce it to 9.4 millimeters per second.

[0103] When the grounding correction factor exceeds the preset grounding threshold of 1.2, the system adjusts the angle of the angle adjustment command. The correction amount is equal to (grounding correction factor - 1) × 15 degrees. In this case, the grounding correction factor is 1.53, which exceeds the preset threshold of 1.2, so the angle needs to be adjusted by (1.53 - 1) × 15 degrees = 7.95 degrees. For example, the original angle adjustment command was a 30-degree clockwise rotation, which is corrected to a 22.05-degree clockwise rotation.

[0104] The system determines the execution timing of the opening / closing control command based on the combined state of the temperature correction coefficient and the grounding correction coefficient. Specifically, the rules are as follows: when both the temperature correction coefficient and the grounding correction coefficient exceed their respective thresholds, the opening / closing control command is delayed by 500 milliseconds; when only the temperature correction coefficient exceeds the threshold, the delay is 200 milliseconds; when only the grounding correction coefficient exceeds the threshold, the delay is 300 milliseconds; and when neither exceeds the threshold, the opening / closing control command is executed immediately. In this case, the temperature correction coefficient of 1.12 exceeds the threshold of 1.1, and the grounding correction coefficient of 1.53 exceeds the threshold of 1.2; therefore, the execution of the opening / closing control command needs to be delayed by 500 milliseconds.

[0105] After completing the above correction calculations, the system sends the corrected position adjustment command, angle adjustment command, and opening / closing control command to the surge arrester execution system via the control bus. Upon receiving the corrected commands, the execution system performs the corresponding operations according to the adjusted parameters, thereby ensuring the safe and reliable operation of the surge arrester under various temperature and grounding conditions.

[0106] In engineering practice, after adopting this method in the surge arrester monitoring system of a 110kV substation, the position adjustment accuracy of the surge arrester was improved by 15%, the adaptability of angle adjustment was enhanced by 22%, and the reliability of opening and closing control was improved by 18%. This effectively reduced the surge arrester malfunction rate caused by environmental factors and improved the safe operation level of the power system.

[0107] The intelligent control system for lightning protection of offshore substations according to an embodiment of the present invention includes:

[0108] The first unit is used to collect surge arrester voltage data, surge arrester current data, atmospheric electric field strength data, surge arrester temperature data, and surge arrester grounding resistance data of the offshore substation through a sensor array.

[0109] The second unit is used to perform Fourier transform on the surge arrester voltage data and surge arrester current data, calculate the power loss distribution, construct the surge arrester electric field interference matrix based on the electromagnetic field finite element model; input the power loss distribution and electric field interference matrix into an adaptive neural network to generate the optimal spacing position vector of the surge arrester; and adjust the position of the surge arrester according to the optimal spacing position vector.

[0110] The third unit is used to filter the atmospheric electric field strength data to extract the electric field strength fluctuation characteristics, and calculate the electric field polarization direction and electric field polarization intensity; establish a surge arrester response characteristic model based on the electric field polarization direction and electric field polarization intensity, calculate the optimal conduction path and optimal conduction time; dynamically adjust the polarization angle of the surge arrester based on the optimal conduction path, and control the opening and closing of the surge arrester based on the optimal conduction time;

[0111] The fourth unit is used to input the surge arrester temperature data and surge arrester grounding resistance data into the state evaluation model, calculate the surge arrester operating state parameters, and correct the position adjustment command, the angle adjustment command, and the opening and closing control command based on the surge arrester operating state parameters.

[0112] The fifth unit is used to monitor the operation data of the surge arrester in real time. When lightning overvoltage occurs, it records the lightning protection response time and protection effect data.

[0113] A third aspect of the present invention provides an electronic device, comprising:

[0114] processor;

[0115] Memory used to store processor-executable instructions;

[0116] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0117] 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.

[0118] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0119] 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. An intelligent lightning protection control method for offshore booster stations, characterized by, The method comprises the following steps: Collecting the voltage data, current data, atmospheric electric field intensity data, temperature data and grounding resistance data of the lightning arresters of the offshore booster station through a sensor array; Performing Fourier transform on the voltage data and current data of the lightning arresters to calculate the power loss distribution and construct a lightning arrester electric field interference degree matrix based on an electromagnetic field finite element model; Inputting the power loss distribution and electric field interference degree matrix into an adaptive neural network to generate a lightning arrester optimal spacing position vector; and adjusting the position of the lightning arresters according to the optimal spacing position vector; Filtering the atmospheric electric field intensity data to extract the electric field intensity fluctuation characteristics, and calculating the electric field polarization direction and electric field polarization intensity; Establishing a lightning arrester response characteristic model according to the electric field polarization direction and the electric field polarization intensity to calculate the optimal conduction path and optimal conduction time; dynamically adjusting the polarization angle of the lightning arresters based on the optimal conduction path, and controlling the opening and closing of the lightning arresters based on the optimal conduction time; Inputting the temperature data and grounding resistance data of the lightning arresters into a state evaluation model to calculate the working state parameters of the lightning arresters, and correcting the position adjustment instruction, angle adjustment instruction and opening and closing control instruction according to the working state parameters of the lightning arresters; Real-time monitoring of the operation data of the lightning arresters, and recording the lightning protection response time and protection effect data when lightning overvoltage occurs; The method for calculating the power loss distribution and constructing the lightning arrester electric field interference degree matrix based on the Fourier transform of the voltage data and current data of the lightning arresters comprises the following steps: Performing Fourier transform on the voltage data and current data of the lightning arresters to obtain the phase spectrum and amplitude spectrum; Based on the phase spectrum and amplitude spectrum, calculating the active power and reactive power components of each frequency band of the lightning arresters, and establishing a power loss distribution vector; Establishing a lightning arrester three-dimensional electromagnetic field finite element model, dividing the grid elements, and constructing an electromagnetic field characteristic parameter matrix; calculating the electric field intensity distribution based on the electromagnetic field characteristic parameter matrix to obtain electric field interference degree data at different spatial positions; and organizing the electric field interference degree data into an electric field interference degree matrix; Mapping the power loss distribution vector to the electric field interference degree matrix to establish a power loss and electric field interference correlation model; calculating the mutual influence coefficient between the lightning arresters based on the correlation model to generate a lightning arrester arrangement optimization suggestion; The method for establishing a lightning arrester response characteristic model according to the electric field polarization direction and the electric field polarization intensity to calculate the optimal conduction path and optimal conduction time comprises the following steps: Mapping the electric field polarization direction to an azimuth angle vector and converting the electric field polarization intensity to an electric field gradient matrix; constructing a three-dimensional space electric field polarization distribution data according to the azimuth angle vector and the electric field gradient matrix; Calculating the charge density distribution and potential gradient in the three-dimensional space electric field polarization distribution data to obtain a lightning channel initiation probability distribution; simulating multiple conduction paths using the Monte Carlo method according to the lightning channel initiation probability distribution; calculating the resistance loss and electromagnetic induction characteristics of each conduction path to determine the optimal conduction path with the minimum energy loss; Calculate the variation law of electric field polarization intensity with time based on the optimal conduction path to obtain the optimal response timing of the surge arrester; and determine the optimal conduction time according to the optimal response timing.

2. The method of claim 1, wherein, Input the power loss distribution and the electric field interference degree matrix into the adaptive neural network to generate an optimal spacing position vector of the surge arrester; and adjust the position of the surge arrester according to the optimal spacing position vector. Construct an adaptive neural network, take the power loss distribution as first input layer data, and take the electric field interference degree matrix as second input layer data; train the adaptive neural network based on a back propagation algorithm, take the optimal spacing data in historical operation data of the surge arrester as a training sample; and complete model training when the training error of the adaptive neural network is less than a preset error threshold; Input real-time power loss distribution and electric field interference degree matrix into the trained adaptive neural network model to calculate the optimal spacing position vector of the surge arrester group; construct a surge arrester spatial arrangement model based on the optimal spacing position vector to generate target coordinates of each surge arrester; Calculate the deviation between the current position of each surge arrester and the target coordinates, and trigger a position adjustment instruction when the position deviation exceeds a preset deviation threshold; Control the position adjustment device of the surge arrester based on the position adjustment instruction to move each surge arrester to the target coordinate position.

3. The method of claim 1, wherein, Input the surge arrester temperature data and the surge arrester grounding resistance data into a state evaluation model to calculate surge arrester working state parameters, and correct the position adjustment instruction, the angle adjustment instruction and the opening and closing control instruction according to the surge arrester working state parameters, including: Perform time-frequency analysis on the surge arrester temperature data to extract temperature change trend characteristics and temperature fluctuation period; perform statistical analysis on the surge arrester grounding resistance data to calculate the grounding resistance change rate and stability coefficient; and calculate the surge arrester working state parameters according to the temperature change trend characteristics, the temperature fluctuation period, the grounding resistance change rate and the stability coefficient; Calculate the temperature correction coefficient and the grounding correction coefficient based on the surge arrester working state parameters; reduce the execution speed of the position adjustment instruction according to a preset proportion when the temperature correction coefficient exceeds a preset temperature threshold; adjust the execution angle of the angle adjustment instruction when the grounding correction coefficient exceeds a preset grounding threshold; and determine the execution timing of the opening and closing control instruction according to the combined state of the temperature correction coefficient and the grounding correction coefficient; Send the corrected position adjustment instruction, angle adjustment instruction and opening and closing control instruction to the surge arrester execution system.

4. An intelligent lightning protection control system for offshore substation, for implementing the method according to any one of claims 1-3, characterized in that, Comprise: A first unit for collecting, by a sensor array, surge arrester voltage data, surge arrester current data, atmospheric electric field intensity data, surge arrester temperature data and surge arrester grounding resistance data of a marine booster station; A second unit for performing Fourier transform on the surge arrester voltage data and surge arrester current data to calculate power loss distribution, and constructing an electric field interference degree matrix of the surge arrester based on an electromagnetic field finite element model; Input the power loss distribution and the electric field interference degree matrix into the adaptive neural network to generate an optimal spacing position vector of the surge arrester; and adjust the position of the surge arrester according to the optimal spacing position vector. The third unit is configured to filter the atmospheric electric field intensity data to extract electric field intensity fluctuation characteristics, and calculate an electric field polarization direction and an electric field polarization intensity. A lightning arrester response characteristic model is established according to the electric field polarization direction and the electric field polarization intensity, and an optimal conduction path and an optimal conduction time are calculated; a polarization angle of the lightning arrester is dynamically adjusted based on the optimal conduction path, and the lightning arrester is controlled to open and close based on the optimal conduction time; The fourth unit is configured to input the lightning arrester temperature data and the lightning arrester grounding resistance data into a state evaluation model, calculate a lightning arrester working state parameter, and correct the position adjustment instruction, the angle adjustment instruction and the open / close control instruction according to the lightning arrester working state parameter. The fifth unit is configured to monitor lightning arrester operation data in real time, and record lightning protection response time and protection effect data when lightning overvoltage occurs.

5. An electronic device, comprising: It comprises: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to invoke the instructions stored in the memory to execute the method of any one of claims 1 to 3.

6. A computer-readable storage medium having stored thereon computer program instructions, wherein, The computer program instructions, when executed by the processor, implement the method of any one of claims 1 to 3.

Images