A parameter self-correction-based hybrid signal injection type capacitance current compensation method, system, device and medium

Abstract

The application discloses a kind of hybrid signal injection type capacitance current compensation method, system, equipment and medium based on parameter self-correction, is specifically related to electrical automation and digital technology field.The method comprises: by main station to distribution network inject wideband signal, response signal and zero sequence signal are collected by column device;Based on U1, I1 data, using improved recursive least square method on-line real-time identification line equivalent capacitance and resistance to ground, realize parameter self-correction;According to zero sequence voltage amplitude automatic switching signal injection method and displacement voltage method calculate capacitance current, and utilize Kalman filtering to carry out data fusion calibration;Finally based on identification parameter, through parameter-compensation amount mapping model dynamic determination arc suppression coil adjustment step size, carry out fast closed loop compensation.The application realizes on-line real-time identification and self-correction of distribution network capacitance parameter to ground, reaches fast, accurate compensation capacitance current under all-weather operation condition, ensures that system residual flow meets requirements.

Description

Technical Field

[0001] This invention relates to the fields of electrical automation and digital technology, and more specifically to a hybrid signal injection capacitor current compensation method, system, device, and medium based on parameter self-calibration. Background Technology

[0002] In the operation of power distribution networks, single-phase grounding faults are the most frequent type of fault. The capacitive current generated at the fault grounding point is the core cause of arc grounding, overvoltage, line insulation breakdown, and even the expansion of the fault range, seriously threatening the safe and stable operation of the distribution network. To suppress the capacitive current during single-phase grounding faults, arc suppression coils are commonly installed in distribution networks, and precise capacitive current compensation is implemented. The inductive current generated by the arc suppression coil cancels the capacitive capacitive current, reducing the residual current at the grounding point to a safe range and achieving arc self-extinguishing. Therefore, accurate and rapid measurement and compensation of the system's capacitive current to ground has become the core key to resonant grounding protection technology in distribution networks.

[0003] Currently, the mainstream capacitor current compensation methods in the power distribution network field are mainly divided into two categories: signal injection method and displacement voltage method. Both methods revolve around the detection and calculation of capacitor current. However, due to their different technical principles, they each have significant technical shortcomings in practical engineering applications, making it difficult to adapt to the complex and ever-changing all-weather operating conditions of the power distribution network. The specific details are as follows:

[0004] 1. Signal Injection Method: This method injects a current signal of a specific frequency into the system and calculates the capacitance to ground by measuring the response voltage. While this method offers high accuracy when the system is balanced, its measurement results are heavily dependent on a pre-set line parameter model. In actual operation, the line's capacitance to ground and insulation resistance can change due to environmental temperature, humidity, aging, and network structure variations, leading to a significant decrease in the accuracy of compensation based on fixed parameters. Long-term operating errors can exceed 10%.

[0005] 2. Displacement Voltage Method: This method utilizes the zero-sequence voltage, or displacement voltage, generated by the system's inherent asymmetry to calculate the capacitor current. While this method does not rely on precise line parameters, it is highly susceptible to interference from system operating conditions. Under normal disturbances such as load fluctuations, distributed power source switching, and intermittent asymmetry, fluctuations in the zero-sequence voltage can be misjudged, leading to malfunctions or slow responses in the compensation device. Tuning times often exceed 50ms, and the residual current is difficult to stably control below the safe limit.

[0006] Besides the inherent defects of the two mainstream methods mentioned above, existing improvements to capacitor current compensation have not fundamentally resolved the core technical contradictions. For example, the "Distributed Arc Suppression Coil Capacitor Current Measurement Method" disclosed in patent CN106877306A optimizes data acquisition sources through distributed measurement nodes, which improves the comprehensiveness of data to some extent. However, this solution still fails to solve the problem of time-varying drift of line parameters and still uses a single measurement and calculation method. It fails to effectively avoid the shortcomings of poor anti-interference and slow response, and cannot adapt to the complex operating conditions of the distribution network.

[0007] In summary, existing capacitive current compensation methods have significant shortcomings in terms of parameter adaptability, anti-interference capability, response speed, and long-term operational stability. There is an urgent need for a capacitive current compensation method that can automatically adapt to parameter changes, intelligently resist operational interference, and achieve rapid and accurate compensation to solve the key technical challenges in current distribution network grounding protection. Summary of the Invention

[0008] In view of the above problems, this invention is proposed to provide a hybrid signal injection capacitor current compensation method, system, device, and medium based on parameter self-calibration to overcome or at least partially solve the above problems. This method aims to achieve online real-time identification and self-calibration of the distribution network's ground capacitance parameters, and integrates the advantages of both active injection and passive monitoring measurement modes, ultimately achieving rapid and accurate capacitor current compensation under all-weather operating conditions, ensuring that the system residual current meets requirements.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, embodiments of the present invention provide a hybrid signal injection capacitor current compensation method based on parameter self-calibration, comprising the following steps: S1. Inject broadband signals into the distribution network through the main station, and collect and upload the response voltage U1, current I1, zero-sequence voltage U0 and zero-sequence current I0 by the pole-mounted device. S2. Based on the response voltage U1 and current I1 data, the recursive least squares method with forgetting factor is used to identify the equivalent capacitance C to ground and the equivalent resistance R to ground in real time, and the parameters are updated when the parameter drift exceeds the threshold. S3. Calculate the capacitor current by automatically switching between signal injection method and displacement voltage method based on the amplitude of the zero-sequence voltage U0, and fuse the two results through Kalman filtering to output the calibrated estimated value of capacitor current Îc. S4. Based on the current equivalent capacitance to ground C and the equivalent resistance to ground R, the arc suppression coil adjustment step size ΔL is determined through a mapping model. Combined with the estimated capacitor current Îc, an adjustment command is generated, and closed-loop control is performed until the system residual current Ires meets the target.

[0010] Furthermore, step S2 includes: S21: Determine the excitation-response model of the dynamic system, and set the initial parameters for the recursive least squares method, including the initial parameter estimates θ(0), the initial estimation error covariance matrix P(0), the forgetting factor λ, and the parameter drift threshold ΔC. th or ΔR th ; S22: For each sampling time k, construct a regression vector φ(k) based on the collected response voltage U1 and current I1 data; S23: Perform recursive calculations, including calculating the gain matrix K(k), updating the parameter estimates θ(k), and updating the estimation error covariance matrix P(k); S24: Monitor the drift of the parameter estimate. When the drift of the equivalent capacitance to ground C or the equivalent resistance to ground R exceeds the parameter drift threshold, update the parameter database.

[0011] Furthermore, in step S21, the excitation-response model of the dynamic system is expressed as:

[0012] Where y(k) is the I1 measurement at time k; φ(k) is the regression vector at time k, composed of U1 and its historical data. ; denoted as , where is the parameter to be identified; e(k) represents the measurement noise.

[0013] Furthermore, step S23 includes: Perform the following calculations in sequence: (1) Calculate the gain matrix:

[0014] Where K(k) is a 2×1 vector and P(k) is the error covariance matrix; (2) Update parameter estimates :

[0015] in, y(k) represents the optimal estimates of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R at time k; y(k) represents the measured value of I1 at time k; φ(k) represents the regression vector composed of U1 and its historical data. This represents the prediction error; These are the optimal estimates of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R at time k-1. (3) Update the estimated error covariance matrix:

[0016] Where P(k) is the estimation error covariance matrix at time k; P(k) reflects the uncertainty of parameter estimation; λ∈[0.95,0.99].

[0017] Further, step S24 includes: calculating the relative drift ΔR and ΔC of the equivalent capacitance C to ground and the equivalent resistance R to ground relative to the reference values ​​R0 and C0, respectively. The formula for calculating the relative drift is:

[0018]

[0019] in, This is an estimate of the equivalent capacitance to ground. This is an estimate of the equivalent resistance to ground; if ΔR is greater than or equal to the stated ΔR... th Or ΔC is greater than or equal to the stated ΔC th When this happens, it triggers an update of the parameters of the equivalent capacitance to ground C or the equivalent resistance to ground R.

[0020] Furthermore, in step S3, automatically switching between the signal injection method and the displacement voltage method to calculate the capacitor current based on the amplitude of the zero-sequence voltage U0 specifically includes: When the zero-sequence voltage U0 is less than or equal to the voltage threshold, the branch capacitor current Ic1 is calculated using the signal injection method; the voltage threshold is related to the state of the system. When the zero-sequence voltage U0 is greater than the voltage threshold, the displacement voltage method is used to calculate the total system capacitance current Ic_total.

[0021] Furthermore, in step S3, a Kalman filter is used to fuse the branch capacitor current Ic1 and the total system capacitor current Ic_total. The filtering formula is as follows:

[0022] Where z(k) is the branch capacitor current I c1 The measurement information is related to the total system capacitance current Ic_total. is the state prediction value at time k-1; K(k) is the Kalman gain.

[0023] Secondly, embodiments of the present invention provide a hybrid signal injection capacitor current compensation system based on parameter self-calibration, the system comprising: Hybrid signal injection and acquisition module: used to inject broadband signals into the distribution network through the master station, and to collect and upload response voltage U1, current I1, zero-sequence voltage U0 and zero-sequence current I0 by pole-mounted devices. Online line parameter identification and self-correction module: Based on the response voltage U1 and current I1 data, it uses a recursive least squares method with a forgetting factor to identify the equivalent capacitance C and equivalent resistance R of the line to ground in real time, and updates the parameters when the parameter drift exceeds the threshold. The multi-method fusion capacitor current calculation module is used to automatically switch between signal injection method and displacement voltage method to calculate capacitor current based on the amplitude of the zero-sequence voltage U0, and fuse the two results through Kalman filtering to output a calibrated capacitor current estimate Îc. Dynamic compensation adjustment and closed-loop control module: It is used to determine the arc suppression coil adjustment step size ΔL based on the current equivalent capacitance C to ground and the equivalent resistance R to ground through a mapping model, and generate adjustment commands by combining the estimated capacitor current Îc, and perform closed-loop control until the system residual current Ires meets the target.

[0024] Thirdly, embodiments of the present invention provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-described method for hybrid signal injection capacitor current compensation based on parameter self-calibration as described in the first aspect embodiment.

[0025] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the hybrid signal injection capacitor current compensation method based on parameter self-calibration described in the first aspect embodiment above.

[0026] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a hybrid signal injection capacitor current compensation method based on parameter self-calibration, which has the following beneficial effects: 1. High precision and strong adaptability By relying on an improved recursive least squares method, online real-time identification and dynamic updating of the equivalent capacitance and resistance of the line to ground are achieved. This fundamentally solves the parameter drift problem caused by changes in environmental temperature and humidity, line aging, and network structure adjustments in traditional compensation methods. It ensures that the capacitance current measurement model always maintains a high degree of matching with the actual physical system of the distribution network, completely eliminating the dependence on fixed parameter models. This significantly improves the accuracy of capacitance current measurement. Simultaneously, the algorithm can track changes in line parameters in real time and complete self-correction without manual intervention to adjust model parameters. It possesses strong adaptability to the complex and ever-changing operating conditions of the distribution network, ensuring stable measurement and compensation accuracy throughout the entire lifecycle.

[0027] 2. Good robustness and fast response This paper innovatively proposes an intelligent fusion strategy combining the signal injection method and the displacement voltage method. Using a zero-sequence voltage of 5V as the critical value, it accurately distinguishes between the steady-state and transient states of the system, achieving seamless automatic switching between the two measurement methods. This fully leverages the advantages of the signal injection method's accurate steady-state measurement and the displacement voltage method's resistance to parameter dependence, while avoiding the measurement drawbacks of a single method under different operating conditions. Simultaneously, a Kalman filter algorithm is introduced to perform data fusion calibration of the measurement results from both methods. By dynamically balancing the reliability of predicted and real-time measured values, it effectively filters out various noise interferences during system operation. This allows the algorithm to maintain high measurement accuracy even under complex disturbances such as load fluctuations, distributed power supply switching, and intermittent asymmetry, significantly improving its anti-interference capability and robustness. Furthermore, the algorithm directly outputs the optimal adjustment step size of the arc suppression coil through a parameter-compensation mapping model. Combined with fast closed-loop adjustment logic, the tuning response time is strictly controlled within 20ms, a significant improvement over traditional methods. This enables rapid and accurate compensation of capacitor current, effectively suppressing arc reignition at the grounding point and preventing fault expansion.

[0028] 3. Highly practical engineering applications This algorithm adopts a purely software-based deployment model, which can be directly integrated into the collaborative control framework of existing distribution network master stations and pole-mounted integrated arc suppression devices. It eliminates the need for large-scale modifications or replacements of primary equipment in the distribution network, significantly reducing hardware costs and construction difficulty. The algorithm has broad adaptability and can be seamlessly applied to neutral point grounding systems via arc suppression coils in distribution networks across all voltage levels from 10kV to 35kV. It is not limited by the distribution network structure or outgoing line scale, and can meet the capacitive current compensation needs of distribution networks of different regions and specifications. Furthermore, the algorithm's execution logic is highly compatible with the data acquisition, communication, and control processes of existing distribution network automation equipment, facilitating on-site debugging and subsequent maintenance. It is easy to scale up and apply in power systems, providing core algorithmic support for building a fast-response, accurate compensation distribution network grounding protection system, and possesses extremely high engineering application value and practical significance for promotion. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0030] Figure 1 This is an overall flowchart provided in an embodiment of the present invention.

[0031] Figure 2 This is a specific flowchart provided in the embodiments of the present invention.

[0032] Figure 3 This is a schematic diagram of the module structure provided in an embodiment of the present invention.

[0033] Figure 4 This is a structural diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0034] 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.

[0035] Example 1: like Figure 1 As shown in the figure, this invention discloses a hybrid signal injection capacitor current compensation method based on parameter self-calibration, including the following steps: S1. The control signal generator injects a broadband voltage signal into the distribution network system and receives the voltage U1 and current I1 in response to the injected signal, as well as the system zero-sequence voltage U0 and zero-sequence current I0, which are synchronously collected by the pole-mounted device. The specific uses and advantages of this step are as follows: it actively injects broadband signals into the distribution network system, synchronously collects response voltage U1, response current I1, zero-sequence voltage U0, and zero-sequence current I0, providing high-precision basic data for subsequent parameter identification and capacitor current calculation; the broadband encoded signal enhances anti-interference capability, synchronous collection ensures data consistency, amplitude is controlled within a safe range to avoid affecting the normal operation of the system, and encrypted uploading ensures communication security.

[0036] S2. Based on the received data sequence of response voltage U1 and current I1, online identification is performed using the recursive least squares method with forgetting factor, and the equivalent capacitance to ground C and equivalent resistance to ground R of each line are estimated in real time; when the estimated value of the equivalent capacitance to ground C and equivalent resistance to ground R changes more than the reference value by a preset threshold, the parameter database is updated. The specific uses and advantages of this step are as follows: Based on the response voltage U1 and current I1 data sequence, the recursive least squares method with forgetting factor is used to identify the equivalent capacitance C and equivalent resistance R of the line to ground in real time, and the model is automatically updated when the parameter drift exceeds the threshold; the forgetting factor enhances the tracking ability of time-varying parameters, and the real-time identification ensures that the model is always consistent with the physical system, fundamentally solving the parameter drift problem; the threshold triggering mechanism avoids invalid calculations, and provides an accurate line parameter basis for capacitor current calculation and compensation adjustment.

[0037] S3. Determine the system operating status based on the amplitude of the zero-sequence voltage U0: If the zero-sequence voltage U0 is less than or equal to the first threshold, the capacitor current is calculated using the signal injection method; if the zero-sequence voltage U0 is greater than the first threshold, the capacitor current is calculated using the displacement voltage method; and fuse the calculation results obtained by the two methods through a Kalman filter to output the calibrated capacitor current estimate Îc. The specific uses and advantages of this step are as follows: it automatically switches between signal injection method and displacement voltage method to calculate the capacitor current based on the amplitude of zero-sequence voltage U0, and then fuses the calculation results of the two methods through Kalman filter to output the calibrated estimated value of capacitor current Îc; the dual-mode intelligent switching takes into account both steady-state accuracy and transient response, overcomes the limitations of a single method, and Kalman filter fusion eliminates measurement noise, suppresses occasional errors, ensures measurement accuracy, and provides a reliable basis for accurate compensation.

[0038] S4. Based on the latest estimated equivalent capacitance to ground C and equivalent resistance to ground R, determine the inductance adjustment step size ΔL of the arc suppression coil through a preset parameter-compensation mapping model; based on the estimated capacitance current Îc and the adjustment step size ΔL, generate an arc suppression coil inductance adjustment command and send it to the corresponding pole-mounted device to adjust the arc suppression coil inductance in a closed-loop control manner until the system residual current Ires meets the preset target value.

[0039] The specific uses and advantages of this step are as follows: Based on the newly identified equivalent capacitance to ground C and equivalent resistance to ground R parameters, the adjustment step size ΔL of the arc suppression coil inductance is determined through a preset parameter-compensation mapping model. Combined with the estimated capacitance current Îc, an adjustment command is generated, and the arc suppression coil inductance is controlled in a closed loop until the system residual current Ires meets the preset target. The mapping model realizes the rapid mapping from parameters to compensation quantities, and the closed-loop control ensures that the residual current accurately meets the target and avoids under-compensation or over-compensation. The smooth adjustment of the step size reduces transient impacts, and the full automation improves the reliability and intelligence level of the distribution network grounding protection.

[0040] This invention has several outstanding advantages: Firstly, by actively injecting broadband signals into the distribution network from the master station, and simultaneously acquiring response voltage, current, and zero-sequence electrical quantities via pole-mounted devices, it achieves full-domain online and real-time sensing of distribution network parameters, enabling parameter acquisition without power outages and significantly improving the observability and operational flexibility of the distribution network. Secondly, it employs a recursive least squares method with a forgetting factor to identify the equivalent capacitance and resistance of the line to ground in real time, enabling rapid tracking of line parameter changes and maintaining stable output even under parameter drift, load fluctuations, and environmental interference. This method features high identification accuracy, fast dynamic response, and strong anti-interference capability. Thirdly, it automatically switches between the signal injection method and the displacement voltage method based on the zero-sequence voltage amplitude, and utilizes Kalman filtering to fuse the results of both algorithms, achieving complementary advantages under different operating conditions. This effectively suppresses measurement noise and system errors, significantly improving the accuracy and reliability of capacitor current estimation. Fourthly, based on the real-time identified parameters, it determines the arc suppression coil adjustment step size through a mapping model, forming a closed-loop control with the calibrated capacitor current. The adjustment process is smooth without overshoot and responds rapidly, quickly controlling the system residual current within the target range, thus enhancing the system's adaptive capability and fault suppression capability. The overall solution integrates active sensing, precise identification, intelligent fusion, and closed-loop regulation, and has advantages such as high degree of automation, stable and reliable operation, wide applicability, and strong practicality. It can effectively improve the safe and stable operation level of the distribution network resonant grounding system.

[0041] The following is a detailed description of the hybrid signal injection capacitor current compensation based on parameter self-calibration of the present invention: like Figure 2 As shown, a hybrid signal injection capacitor current compensation method based on parameter self-calibration is applied to a collaborative control system consisting of a distribution network master station and distributed pole-mounted integrated arc suppression devices. The master station is responsible for centralized calculation and decision-making, while the pole-mounted devices are responsible for local signal acquisition and execution. The method includes the following steps: S1. Mixed signal injection and acquisition: S11, The distribution network master station controls a broadband signal generator to inject an adjustable coded broadband voltage signal with a frequency ranging from 20Hz to 50Hz into the system bus. The amplitude Umax of this signal is limited to between 1% and 3% of the rated phase voltage of the system to ensure that it does not affect the normal operation of the power grid.

[0042] S12, the pole-mounted integrated arc suppression device has a built-in high-precision acquisition unit that simultaneously measures two types of signals: one is the specific frequency band response voltage U1 and response current I1 generated by the injected signal; the other is the inherent zero-sequence voltage U0 and zero-sequence current I0 of the power grid. The acquired data is transmitted to the main station in real time through a secure encrypted channel, i.e., using the preferred national cryptographic SM4 algorithm, with a sampling period of no more than 10ms.

[0043] This invention utilizes a distribution network master station to control a broadband signal generator, injecting a frequency-adjustable, amplitude-limited coded broadband voltage signal into the system bus. Simultaneously, it employs the high-precision acquisition unit of a pole-mounted integrated arc suppression device to synchronously measure the specific frequency band response voltage U1 and response current I1 generated by the injected signal, as well as the inherent zero-sequence voltage U0 and zero-sequence current I0 of the power grid. This achieves precise injection and efficient acquisition of the mixed signal, providing reliable data support for subsequent applications such as distribution network fault diagnosis and operational status analysis, while ensuring the safety and stability of power grid operation.

[0044] S2. Online identification and self-calibration of line parameters: S21. Define the stimulus-response model of the dynamic system: The U1 (excitation) and I1 (response) received by the master station satisfy a linear relationship, that is, the circuit model based on the line-to-ground equivalent resistance R and the line-to-ground equivalent capacitance C can be expressed as:

[0045] Where y(k) is the I1 measurement (response) at time k; φ(k) is the regression vector at time k (composed of current / historical data of U1). ); The parameters to be identified are the equivalent capacitance to ground C and the equivalent resistance to ground R; e(k) is the measurement noise.

[0046] Before the identification begins, i.e. at k=0, set the initial values: The initial parameter estimate is θ(0), which can be set as the line reference value or a small random number; The initial estimation error covariance matrix is Where α is a large positive number, such as 10 6 I is a 2×2 identity matrix, which guarantees the initial convergence rate; The forgetting factor is λ∈[0.95, 0.99], which can be selected according to the time-varying rate; if the parameter changes rapidly, a smaller value should be selected. The thresholds are ΔCth=5% and ΔRth=8%, which are the drift thresholds relative to the baseline value.

[0047] S22. Based on the constructed excitation-response model and the y(k) expression, a real-time data acquisition and refresh mechanism is established to continuously receive synchronously acquired data at time k uploaded by the column-mounted device, and to classify, extract, and update the response data and excitation data in real time.

[0048] The response data is y(k), which is the current measured value of current I1; The excitation data is the current / historical data of the response voltage U1, and a regression vector is constructed. .

[0049] This step establishes a real-time data acquisition and refresh mechanism based on the constructed excitation-response model and y(k) expression. It classifies, extracts, and updates the synchronous data at time k uploaded by the pole-mounted device, extracts the I1 measurement value as the response data y(k), and integrates the current and historical data of the response voltage U1 to construct a regression vector, providing real-time and matching input data for subsequent line parameter recursive identification.

[0050] S23, Recursive Calculation: Calculate according to the following formula, that is, update once for each new set of data received: (1) Calculate the gain matrix:

[0051] Where K(k) is a 2×1 vector used to adjust the "step size" of parameter estimation, and P(k) is the error covariance matrix; (2) Update parameter estimates :

[0052] in, y(k) represents the optimal estimates of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R at time k; y(k) represents the measured value of I1 at time k; φ(k) represents the regression vector composed of the response voltage U1 and its historical data. To correct the prediction error, the previous parameter estimate is corrected using K(k). These are the optimal estimates of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R at time k-1. (3) Update the estimated error covariance matrix:

[0053] Where P(k) is the estimation error covariance matrix at time k; P(k) reflects the uncertainty of parameter estimation, and the forgetting factor λ decays the weight of old data to enhance the time-varying tracking ability. λ∈[0.95, 0.99] is used to enhance the algorithm's ability to track time-varying parameters.

[0054] S24. Identification Result Monitoring and Parameter Update: Each time I get That is, after determining the equivalent capacitance C to ground and the equivalent resistance R to ground, calculate the relative drift with respect to the reference values ​​(R0, C0):

[0055]

[0056] If ΔR ≥ 8% or ΔC ≥ 5%, then "parameter update" is triggered, updating the parameters of the equivalent capacitance to ground C and equivalent resistance to ground R of the line model for subsequent analysis / control; if the threshold is not exceeded, then the identification of the next set of data continues.

[0057] This step uses the excitation-response data of the injected signal as a basis to construct a linear identification model of the equivalent capacitance C and equivalent resistance R of the line to ground. By configuring reasonable initial parameters and a forgetting factor, an improved recursive least squares method with a forgetting factor is used to achieve online real-time recursive identification of the equivalent resistance and capacitance of the line. At the same time, a parameter drift threshold is set to monitor the identification results in real time. When the parameter drift exceeds the threshold, the line model parameters are automatically updated. This fundamentally solves the problem of decreased compensation accuracy caused by the drift of line parameters due to environmental factors, aging, and other factors in traditional methods, and realizes the self-correction of line parameters. This provides a real-time and accurate parameter basis for subsequent accurate calculation and compensation of capacitance current.

[0058] S3. Calculation of capacitor current using a fusion of multiple methods: S31. The main station determines the system operating status: When the zero-sequence voltage U0 ≤ 5V, that is, when the system is operating normally or in a slightly unbalanced state, the branch capacitor current is calculated using the signal injection method. The formula is , where f is the frequency of the injected signal; S32. When the zero-sequence voltage U0 > 5V, i.e., when the system is disturbed, automatically switch to the displacement voltage method to calculate the total system capacitance current Ic_total. The calculation formula is as follows: , where Xc is the system equivalent capacitive reactance; S33. Kalman filtering is used to fuse the branch capacitor current Ic1 and the total system capacitor current Ic_total. The filtering formula is as follows:

[0059] This formula combines the measured information (z(k)) of the branch capacitor current Ic1 and the total system capacitor current Ic_total with the previous state information. By combining the two data sources and using the Kalman gain K(k) to weigh the reliability of the predicted values ​​and the reliability of the measured values, a fused result is obtained that utilizes information from multiple data sources while suppressing errors.

[0060] in, That is K(k) is the Kalman gain, ensuring that the measurement error is ≤2%.

[0061] The system intelligently determines the operating status of the distribution network system using a zero-sequence voltage of 5V as the threshold. When the zero-sequence voltage U0≤5V, the system is determined to be in normal operation or a slightly asymmetrical steady state. At this time, the signal injection method is used to accurately calculate the capacitive current of each branch. When the zero-sequence voltage U0>5V, the system is determined to be in a disturbance transient state due to factors such as load fluctuations and power supply switching. The system immediately switches to the displacement voltage method to calculate the total capacitive current of the entire system. To address the measurement characteristics of the two methods under different operating conditions, a Kalman filter algorithm is introduced to perform data fusion processing on the branch capacitance current obtained by the signal injection method and the total capacitance current obtained by the displacement voltage method. Relying on the state prediction and measurement update mechanism of the Kalman filter, multi-source measurement information and historical state information are organically combined. By dynamically balancing the reliability weights of the predicted value and the real-time measurement value through Kalman gain, noise interference in the measurement process is effectively filtered out. The measurement advantages of different methods are integrated, and finally, a calibrated and accurate capacitance current estimate is output. This not only avoids the disadvantages of the single signal injection method being susceptible to parameter drift and the displacement voltage method being susceptible to system disturbance interference, but also achieves high accuracy and high anti-interference of capacitance current calculation under steady-state and transient operating conditions, ensuring a reduction in overall measurement error and providing reliable data support for subsequent accurate compensation.

[0062] S4. Dynamic compensation adjustment: S41. Based on the line parameters identified in S2, construct a parameter-compensation mapping model. The model input is the equivalent capacitance C to ground and the equivalent resistance R to ground, and the output is the optimal adjustment step size ΔL of the arc suppression coil inductance value, with a value range of 0.05-0.1mH. S42. The main station calculates the target compensation inductance based on the fused capacitor current calibration value and sends an adjustment command to the column-mounted integrated arc suppression device to control the arc suppression coil inductance value to be adjusted from L0 to L0±ΔL. S43. Calculate the adjusted system residual current Ires in real time. When Ires meets the condition, stop the adjustment; if Ires does not meet the condition, repeat steps S401-S402 until the residual current requirement is met. The total response time of the entire tuning process is strictly controlled within 20ms.

[0063] Based on the real-time parameters of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R obtained through online identification in step S2, a parameter-compensation mapping model is established. Using the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R as inputs, the optimal adjustment step size ΔL of the arc suppression coil inductance is precisely output, with ΔL limited to a reasonable range of 0.05-0.1mH. The main station, combined with the estimated capacitance current after Kalman filtering calibration, calculates the target compensation inductance of the arc suppression coil and issues adjustment commands to the pole-mounted integrated arc suppression device, adjusting the inductance value from the initial value L0 by ΔL. After adjustment, the system residual current Ires is monitored and calculated in real time. The inductance parameters are iteratively adjusted repeatedly using closed-loop control until the residual current meets the preset requirements. The closed-loop response time of the entire tuning compensation is strictly controlled within 20ms, achieving dynamic, rapid, and accurate compensation of the capacitance current and ensuring the effective suppression of capacitance current during grounding faults in the distribution network.

[0064] The specific implementation of the present invention is illustrated in detail below through two examples: (I) Normal operation scenario of 10kV distribution network: In a 10kV distribution network system, the rated phase voltage is approximately 5.77kV. The main substation injects a 25Hz coded broadband signal into the busbar, with an amplitude set to 0.5kV, approximately 8.7% of the rated value. A pole-mounted device located on a certain outgoing line collects a response voltage U1 = 0.48kV and a response current I1 = 0.20A.

[0065] After data upload, the main station uses an improved RLS algorithm with λ=0.98 to identify the current equivalent capacitance to ground of the line, C=1.2μF, and the equivalent resistance to ground, R=5Ω. Compared with the baseline value from last week, ΔC=2% (<5%), and the parameters were not updated. At this time, the system zero-sequence voltage U0=3V, indicating a steady-state mode. Using the signal injection method, I_c1=2. π 25 1.2e-6 0.48 ≈ 0.0905A.

[0066] After Kalman filtering calibration, the estimated current of this branch capacitor was Îc = 0.089A. The mapping model, based on (C = 1.2μF, R = 5Ω), output an adjustment step size ΔL = 0.05mH. The master station command fine-tuned the inductance of the arc suppression coil from L0 = 100.00mH to 100.05mH. After adjustment, the measured residual current Ires = 0.8A, meeting the residual current requirement, and the entire process took 15ms.

[0067] (II) 35kV distribution network disturbance scenarios A disturbance occurred in a 35kV distribution network system due to the switching of a nearby line. A 40Hz signal with an amplitude of 2.5kV was injected into the main substation. The pole-mounted device measured U1=2.4kV and I1=0.30A.

[0068] The main station identifies C=0.8μF, R=8Ω, and parameter drift ΔR=6% (<8%). At this time, the system zero-sequence voltage rises to U0=6V, which is determined to be a disturbance mode, and automatically switches to the displacement voltage method. The system power frequency is 50Hz, and the equivalent capacitance to ground is C=0.8μF. Therefore, the system equivalent capacitive reactance is: Xc=1 / (2π 50 0.8e-6) ≈ 3979Ω. The total capacitive current Ic_total is calculated as 6 / 3979 ≈ 1.51mA. This calculation is for schematic purposes only; the actual calculation is a vector calculation. After Kalman filtering and fusing this value with injection method measurements from other lines, the total calibration capacitive current output is Îc = 1.45A. The mapping model outputs ΔL = 0.1mH based on the current parameters. The master station command adjusts the arc suppression coil inductance from 200.0mH to 199.9mH. After adjustment, the residual current Ires = 0.9A, meeting the standard, with a total response time of 18ms.

[0069] Two implementation examples were conducted for two typical operating conditions: steady-state operation of a 10kV distribution network and disturbance operation of a 35kV distribution network. These examples verified the practical application effect of a hybrid signal injection-based capacitor current compensation algorithm with parameter self-calibration under different voltage levels and operating states. The algorithm followed a process of signal injection acquisition, parameter identification and monitoring, calculation based on zero-sequence voltage switching, Kalman filtering fusion, adjustment of the step size by the mapping model output, and closed-loop adjustment of the arc suppression coil inductance. When parameter drift did not exceed the threshold, the reference parameters were used. Under steady-state conditions, the signal injection method was used; under disturbance conditions, the displacement voltage method was automatically switched to calculate the capacitor current. After filtering and calibration, the inductance was fine-tuned. Precise adjustment of the arc suppression coil was achieved in both scenarios. The residual current after adjustment met the preset requirements, and the overall tuning response times were 15ms and 18ms, respectively, both within the 20ms requirement. This fully verified the algorithm's technical advantages in 10kV-35kV distribution networks: strong adaptability, fast response speed, and high compensation accuracy. It also demonstrated its practical engineering value in achieving stable and accurate capacitor current compensation under both steady-state and disturbance conditions.

[0070] Example 2: like Figure 3 As shown, based on the same inventive concept, this embodiment of the invention also provides a hybrid signal injection capacitor current compensation system based on parameter self-calibration, the system comprising: Hybrid signal injection and acquisition module: used to inject broadband signals into the distribution network through the master station, and to collect and upload response voltage U1, current I1, zero-sequence voltage U0 and zero-sequence current I0 by pole-mounted devices. Online line parameter identification and self-correction module: Based on the response voltage U1 and current I1 data, it uses a recursive least squares method with a forgetting factor to identify the equivalent capacitance C and equivalent resistance R of the line to ground in real time, and updates the parameters when the parameter drift exceeds the threshold. The multi-method fusion capacitor current calculation module is used to automatically switch between signal injection method and displacement voltage method to calculate capacitor current based on the amplitude of the zero-sequence voltage U0, and fuse the two results through Kalman filtering to output a calibrated capacitor current estimate Îc. Dynamic compensation adjustment and closed-loop control module: It is used to determine the arc suppression coil adjustment step size ΔL based on the current equivalent capacitance C to ground and the equivalent resistance R to ground through a mapping model, and generate adjustment commands by combining the estimated capacitor current Îc, and perform closed-loop control until the system residual current Ires meets the target.

[0071] Since these systems and the principles of the problems they solve are similar to the aforementioned hybrid signal injection capacitor current compensation method based on parameter self-calibration, the implementation of this system can refer to the implementation of the aforementioned method, and the repetitions will not be repeated.

[0072] Example 3: Based on the same inventive concept, the present invention also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; When the processor executes a program stored in the memory, it is able to implement the hybrid signal injection capacitor current compensation method based on parameter self-calibration as described in any one of Embodiments 1.

[0073] like Figure 4 As shown, the electronic device may include: a processor 10, a communication interface 20, a memory 30, and a communication bus 40, wherein the processor 10, the communication interface 20, and the memory 30 communicate with each other via the communication bus 40. The processor 10 can call logic instructions in the memory 30 to execute a hybrid signal injection capacitor current compensation method based on parameter self-calibration, the method including: S1. Inject broadband signals into the distribution network through the main station, and collect and upload the response voltage U1, current I1, zero-sequence voltage U0 and zero-sequence current I0 by the pole-mounted device. S2. Based on the response voltage U1 and current I1 data, the recursive least squares method with forgetting factor is used to identify the equivalent capacitance C to ground and the equivalent resistance R to ground in real time, and the parameters are updated when the parameter drift exceeds the threshold. S3. Calculate the capacitor current by automatically switching between signal injection method and displacement voltage method based on the amplitude of the zero-sequence voltage U0, and fuse the two results through Kalman filtering to output the calibrated estimated value of capacitor current Îc. S4. Based on the current equivalent capacitance to ground C and the equivalent resistance to ground R, the arc suppression coil adjustment step size ΔL is determined through a mapping model. Combined with the estimated capacitor current Îc, an adjustment command is generated, and closed-loop control is performed until the system residual current Ires meets the target.

[0074] Example 4: This invention also provides a computer-readable storage medium containing a program for executing the parameter self-calibration-based hybrid signal injection capacitor current compensation method of Embodiment 1 described above. This program can be executed on a processor.

[0075] Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0076] The program stored on this medium is loaded into the processor's memory and executed to perform various functions. This storage medium, connected to hardware devices, enables the computer to perform the steps of Embodiment 1 described above.

[0077] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0078] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A hybrid signal injection capacitor current compensation method based on parameter self-calibration, characterized in that, Includes the following steps: S1. Inject broadband signals into the distribution network through the main station, and collect and upload the response voltage U1, current I1, zero-sequence voltage U0 and zero-sequence current I0 by the pole-mounted device. S2. Based on the response voltage U1 and current I1 data, the recursive least squares method with forgetting factor is used to identify the equivalent capacitance C to ground and the equivalent resistance R to ground in real time, and the parameters are updated when the parameter drift exceeds the threshold. S3. Calculate the capacitor current by automatically switching between signal injection method and displacement voltage method based on the amplitude of the zero-sequence voltage U0, and fuse the two results through Kalman filtering to output the calibrated estimated value of capacitor current Îc. S4. Based on the current equivalent capacitance to ground C and the equivalent resistance to ground R, the arc suppression coil adjustment step size ΔL is determined through a mapping model. Combined with the estimated capacitor current Îc, an adjustment command is generated, and closed-loop control is performed until the system residual current Ires meets the target.

2. The method as described in claim 1, characterized in that, Step S2 includes: S21: Determine the excitation-response model of the dynamic system, and set the initial parameters for the recursive least squares method, including the initial parameter estimates θ(0), the initial estimation error covariance matrix P(0), the forgetting factor λ, and the parameter drift threshold ΔC. th or ΔR th ; S22: For each sampling time k, construct a regression vector φ(k) based on the collected response voltage U1 and current I1 data; S23: Perform recursive calculations, including calculating the gain matrix K(k), updating the parameter estimates θ(k), and updating the estimation error covariance matrix P(k); S24: Monitor the drift of the parameter estimate. When the drift of the equivalent capacitance to ground C or the equivalent resistance to ground R exceeds the parameter drift threshold, update the parameter database.

3. The method as described in claim 2, characterized in that, In step S21, the excitation-response model of the dynamic system is expressed as: Where y(k) is the I1 measurement at time k; φ(k) is the regression vector at time k, composed of U1 and its historical data. ; denoted as , where is the parameter to be identified; e(k) represents the measurement noise.

4. The method as described in claim 2, characterized in that, Step S23 includes: Perform the following calculations in sequence: (1) Calculate the gain matrix: Where K(k) is a 2×1 vector and P(k) is the error covariance matrix; (2) Update parameter estimates : in, y(k) represents the optimal estimates of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R at time k; y(k) represents the measured value of I1 at time k; φ(k) represents the regression vector composed of U1 and its historical data. This represents the prediction error; These are the optimal estimates of the line-to-ground equivalent capacitance C and the line-to-ground equivalent resistance R at time k-1. (3) Update the estimated error covariance matrix: Where P(k) is the estimation error covariance matrix at time k; P(k) reflects the uncertainty of parameter estimation; λ∈[0.95,0.99].

5. The method as described in claim 2, characterized in that, Step S24 includes: calculating the relative drift ΔR and ΔC of the equivalent capacitance C to ground and the equivalent resistance R to ground relative to the reference values ​​R0 and C0, respectively. The formula for calculating the relative drift is: in, This is an estimate of the equivalent capacitance to ground. This is an estimate of the equivalent resistance to ground; if ΔR is greater than or equal to the stated ΔR... th Or ΔC is greater than or equal to the stated ΔC th When this happens, it triggers an update of the parameters of the equivalent capacitance to ground C or the equivalent resistance to ground R.

6. The method as described in claim 1, characterized in that, In step S3, the capacitor current is calculated by automatically switching between the signal injection method and the displacement voltage method based on the amplitude of the zero-sequence voltage U0. Specifically, this includes: When the zero-sequence voltage U0 is less than or equal to the voltage threshold, the branch capacitor current Ic1 is calculated using the signal injection method; the voltage threshold is related to the state of the system. When the zero-sequence voltage U0 is greater than the voltage threshold, the displacement voltage method is used to calculate the total system capacitance current Ic_total.

7. The method as described in claim 6, characterized in that, In step S3, a Kalman filter is used to fuse the branch capacitor current Ic1 and the total system capacitor current Ic_total. The filtering formula is as follows: Where z(k) is the branch capacitor current I c1 The measurement information is related to the total system capacitance current Ic_total. is the state prediction value at time k-1; K(k) is the Kalman gain.

8. A hybrid signal injection capacitor current compensation system based on parameter self-calibration, characterized in that, The system comprises: Hybrid signal injection and acquisition module: used to inject broadband signals into the distribution network through the master station, and to collect and upload response voltage U1, current I1, zero-sequence voltage U0 and zero-sequence current I0 by pole-mounted devices. Online line parameter identification and self-correction module: Based on the response voltage U1 and current I1 data, it uses a recursive least squares method with a forgetting factor to identify the equivalent capacitance C and equivalent resistance R of the line to ground in real time, and updates the parameters when the parameter drift exceeds the threshold. The multi-method fusion capacitor current calculation module is used to automatically switch between signal injection method and displacement voltage method to calculate capacitor current based on the amplitude of the zero-sequence voltage U0, and fuse the two results through Kalman filtering to output a calibrated capacitor current estimate Îc. Dynamic compensation adjustment and closed-loop control module: It is used to determine the arc suppression coil adjustment step size ΔL based on the current equivalent capacitance C to ground and the equivalent resistance R to ground through a mapping model, and generate adjustment commands by combining the estimated capacitor current Îc, and perform closed-loop control until the system residual current Ires meets the target.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the hybrid signal injection capacitor current compensation method based on parameter self-calibration as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the hybrid signal injection capacitor current compensation method based on parameter self-calibration as described in any one of claims 1 to 7.

Images