Error online identification method and system based on multi-cvt vector collaborative matching

Abstract

This invention provides an online error identification method and system based on multi-CVT vector collaborative matching. The method includes: synchronously acquiring secondary-side voltage vector data of at least three CVTs of the same phase on the same bus at the same time section, and preprocessing the data to obtain target voltage vector data; determining the estimated value of the primary-side voltage after compensation for each CVT based on the target voltage vector data, rated transformation ratio, and error compensation vector to be solved; determining the average voltage vector based on the estimated value of the primary-side voltage after compensation, and constructing an optimization model with the objective of minimizing its deviation from the average voltage vector; solving the optimization model under preset constraints to obtain the optimal error compensation vector; and then determining the ratio difference estimate and angle difference estimate of each CVT accordingly, and identifying the operating status in combination with preset error limits. This invention can reduce the cost of CVT error monitoring and improve the ability to identify small errors.

Description

Technical Field

[0001] This invention relates to the technical field of online evaluation of the metering performance of high-voltage metering equipment, and in particular to an online error identification method and system based on multi-CVT vector collaborative matching. Background Technology

[0002] Voltage transformers, especially capacitive voltage transformers (CVTs), are key equipment in the metering and protection of power systems. Their metering accuracy directly affects the reliability of electricity metering results and related economic settlements. For CVTs installed on the busbar or outgoing line side of substations, any deviation in ratio or angle can lead to long-term accumulation of metering errors. Therefore, it is necessary to continuously evaluate the metering performance of CVTs.

[0003] Existing CVT metering performance monitoring methods mainly include two categories: periodic power outage testing and online monitoring device testing.

[0004] Periodic power outage testing typically involves conducting power outage tests on CVTs within a predetermined cycle, based on relevant verification procedures. While this method yields relatively accurate test results, it has at least the following drawbacks: First, the testing cycle is usually long, and even if the equipment exceeds tolerances between adjacent testing cycles, it is difficult to detect in a timely manner, resulting in a significant lag in condition perception. Second, each test requires a power outage and on-site implementation, which is costly and can disrupt power supply operations. Third, it is difficult to implement at remote sites or renewable energy plants.

[0005] Online monitoring devices typically employ dedicated online monitoring terminals on the secondary side of the CVT to continuously collect and analyze information such as voltage amplitude and phase. While enabling continuous monitoring, significant limitations remain: firstly, widespread deployment across a large power grid incurs high costs for hardware procurement, installation, and subsequent maintenance; secondly, they are highly dependent on the measurement accuracy and long-term stability of the online monitoring terminal itself, as terminal drift or calibration deviations can affect the final evaluation results. Furthermore, uploading large amounts of real-time data can introduce additional data transmission and security management challenges.

[0006] In addition to the methods mentioned above, existing research also includes schemes that perform independent analysis based on the secondary voltage data of a single CVT, such as judging the equipment status by monitoring harmonics, amplitude fluctuations, or phase change trends. However, these schemes usually only analyze a single device and are easily affected by common disturbances such as system-side voltage fluctuations and load changes. They are difficult to effectively separate the CVT's own small errors, especially when identifying slight out-of-tolerance or early offset states, where sensitivity and reliability remain limited.

[0007] Therefore, there is an urgent need for a method and system that does not require additional dedicated hardware, can utilize the group measurement information of multiple CVTs on the same bus, and can achieve online error identification under uninterrupted power supply conditions, so as to improve the ability to identify small CVT errors and the timeliness of operational status assessment. Summary of the Invention

[0008] To address the aforementioned shortcomings of the existing technology, the purpose of this invention is to provide an online error identification method and system based on multi-CVT vector collaborative matching, aiming to solve the technical problems of high cost, slow perception, and difficulty in accurately identifying minute errors in existing CVT error monitoring.

[0009] To achieve the above objectives, in a first aspect, the present invention provides an online error identification method based on multi-CVT vector collaborative matching, the steps of which include:

[0010] S1. Synchronously acquire the secondary side voltage vector data of at least three capacitive voltage transformers (CVTs) of the same phase on the same bus at the same time section, wherein the secondary side voltage vector data includes at least voltage amplitude and phase information.

[0011] S2. Preprocess the secondary voltage vector data to obtain target voltage vector data. Preferably, the preprocessing includes at least one of data validity verification, outlier removal, noise suppression, and time alignment.

[0012] S3. Based on the target voltage vector data, rated ratio and error compensation vector to be solved for each CVT, determine the estimated value of the primary side voltage after compensation for each CVT.

[0013] S4. Determine the average voltage vector based on the estimated value of the primary side voltage after compensation for each CVT, and construct an optimization model with the objective of minimizing the deviation of the estimated value of the primary side voltage after compensation for each CVT from the average voltage vector.

[0014] S5. Solve the optimization model under preset constraints to obtain the optimal error compensation vector for each CVT.

[0015] S6. Based on the optimal error compensation vector corresponding to each CVT, determine the ratio difference estimate and angle difference estimate of each CVT, and compare the ratio difference estimate and angle difference estimate with the preset error limit to identify the operating status of each CVT.

[0016] As a further improvement to the above technical solution, the secondary voltage vector data is derived from one or more of the following: an electricity information acquisition system, a synchronous phasor measurement unit (PMU), a merging unit, or a metering acquisition terminal.

[0017] As a further improvement to the above technical solution, in step S2, for secondary voltage vector data with time-scale deviations, time alignment is performed using interpolation or phase compensation based on system frequency; in the... The original secondary voltage vector of the CVT is Time scale deviation is The system frequency is In the case of the following formula, the first... Phase compensation is performed on the original secondary voltage vector of the CVT to obtain the time-aligned target voltage vector. :

[0018] ;

[0019] in, It is the imaginary unit.

[0020] As a further improvement to the above technical solution, before step S3, an error model is established between the CVT measurement value and the actual voltage on the system side. The system-side true voltage vector of the CVT is Rated ratio is The secondary side measured voltage vector is The difference is Angular difference is Random measurement noise is In the case of this, the error model satisfies:

[0021] ;

[0022] Furthermore, the first The measurement error of a CVT is represented as a complex error vector. ,satisfy:

[0023] ;

[0024] Wherein, the complex error vector The magnitude of the difference in magnitude corresponds to the magnitude of the complex error vector. The argument corresponds to the direction and magnitude of the angular difference.

[0025] As a further improvement to the above technical solution, steps S3 and S4 specifically include: in the first step... The target voltage vector after time alignment of the CVT is Rated ratio is The error compensation vector to be solved is In the case of the first Estimated primary voltage after compensation for each CVT satisfy:

[0026] ;

[0027] In theory, for the same phase, all ( It should approximate the same true value. Therefore, the degree of dispersion among them should be minimized. Let the total number of CVTs participating in collaborative matching be... Then the average voltage vector satisfy:

[0028] ;

[0029] in, .

[0030] As a further improvement to the above technical solution, the objective function of the optimization model in step S4... Represented as:

[0031] ;

[0032] in, Let be the objective function. For the first The weighting coefficients corresponding to the CVTs. The magnitude of a complex vector; the first Weighting coefficients corresponding to CVT It is determined based on at least one of the following: the historical reliability of the corresponding CVT, its years of service, and manufacturer information.

[0033] As a further improvement to the above technical solution, the preset constraints in step S5 include at least the following: Error compensation vector corresponding to CVT The modulus constraint and phase angle constraint satisfy:

[0034] ;

[0035] ;

[0036] in, For the first The rated voltage of the secondary side of the CVT. This is a preset percentage coefficient. This indicates taking the phase angle. and These are the preset lower limit of the phase angle and the preset upper limit of the phase angle, respectively.

[0037] As a further improvement to the above technical solution, the optimization model further includes a sparsity constraint term to ensure that the error compensation vector obtained by solving is sparsely distributed across multiple CVTs. The objective function after introducing the sparsity constraint... Represented as:

[0038] ;

[0039] in, To introduce the sparsity constraint into the objective function, This is the sparsity adjustment coefficient.

[0040] As a further improvement to the above technical solution, in step S5, the optimization model is solved using a particle swarm optimization algorithm, a genetic algorithm, or a gradient descent algorithm, and at least one of the following methods is used to improve the stability of the solution:

[0041] The optimization process is run independently multiple times, and the solution with the minimum objective function value is selected as the final solution.

[0042] The solution result from the previous time step is used as the initialization center for the current time step optimization solution.

[0043] As a further improvement to the above technical solution, in step S6, in the first... The optimal error compensation vector corresponding to the CVT is In the case of the following formula, the first one is determined. Estimated ratio of CVT difference Sum of angle difference estimates :

[0044] ;

[0045] ;

[0046] in, Indicates taking the real part, This indicates taking the imaginary part; and the estimated value of the ratio difference Sum of angle difference estimates Compare with the error limit for the corresponding accuracy class to determine the accuracy level. The operating status of the CVT is divided into normal, warning, or abnormal.

[0047] Secondly, the present invention also provides an online error identification system based on multi-CVT vector collaborative matching, for implementing the online error identification method described in the first aspect, the system comprising:

[0048] The data acquisition module is used to synchronously acquire the secondary side voltage vector data of at least three CVTs of the same phase on the same bus at the same time section.

[0049] The preprocessing module is used to perform data validity verification, outlier removal, noise suppression, and time alignment on the secondary side voltage vector data.

[0050] The collaborative matching analysis module is used to establish the CVT error model, determine the estimated value of the primary side voltage and the average voltage vector after compensation for each CVT, construct and solve the optimization model to obtain the optimal error compensation vector for each CVT.

[0051] The error calculation and status assessment module is used to determine the ratio difference estimate and angle difference estimate of each CVT based on the optimal error compensation vector corresponding to each CVT, and to identify the operating status of each CVT according to the preset error limit.

[0052] The early warning output module is used to output the identification results and early warning information.

[0053] Thirdly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the online error identification method based on multi-CVT vector collaborative matching as described in the first aspect.

[0054] Because the present invention adopts the above technical solutions, the beneficial effects of the present invention are as follows:

[0055] This invention provides an online error identification method based on multi-CVT vector collaborative matching. This invention directly utilizes the CVT voltage vector data in the existing acquisition system for analysis, without the need to add a dedicated online monitoring terminal for each CVT, thus reducing hardware investment, installation and implementation costs, and subsequent operation and maintenance costs.

[0056] This invention constructs a group collaborative matching model by using the synchronous measurement results of the same system voltage by multiple CVTs on the same bus and in the same phase. It can utilize the group redundancy information to reduce the impact of common disturbances such as system-side voltage fluctuations and load changes on the identification results, thereby improving the ability to identify small ratio differences and angle differences of CVTs.

[0057] This invention can complete online identification and status assessment of CVT errors without power interruption, which can shorten the anomaly detection cycle and reduce the measurement risks caused by long-term out-of-tolerance operation of CVT.

[0058] This invention constructs a complete error inversion link through complex error vector, compensated primary voltage estimate, average voltage vector, and weighted deviation minimization model, which can directly convert the final result into ratio difference estimate and angle difference estimate, and has good engineering interpretability and application feasibility.

[0059] This invention helps to make the solution results more consistent with the actual operating characteristics of CVT by setting the magnitude constraint and phase angle constraint of the error compensation vector, and introducing sparsity constraint terms when necessary, and improves the positioning capability and solution stability of abnormal equipment.

[0060] This invention is based on the fundamental idea of ​​performing group consistency analysis on the same physical quantity using similar measuring devices. It has good versatility and, after adaptive adjustments, can be extended to group status assessment scenarios for other voltage transformers or similar metering devices. Attached Figure Description

[0061] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0062] Figure 1 This is a flowchart illustrating the online error identification method based on multi-CVT vector collaborative matching disclosed in this invention 1;

[0063] Figure 2 This is a schematic diagram of the CVT error model and the principle of primary-side voltage consistency recovery after compensation disclosed in this invention.

[0064] Figure 3 This is a schematic diagram of the solution process of the optimization model disclosed in this invention.

[0065] The realization of the objective, functional characteristics and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0066] 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 a part of the embodiments of the present invention, and not all of them. 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.

[0067] It should be noted that the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0068] Example 1

[0069] See Figures 1-3 This invention provides an online error identification method based on multi-CVT vector cooperative matching, applicable to online condition assessment scenarios of multiple capacitive voltage transformers (CVTs) on the same bus and in the same phase within a substation. Specifically, under the same bus and phase conditions, the objects measured by multiple CVTs are essentially the same system-side voltage. Ideally, the measured results after conversion by each CVT should be consistent; when a CVT exhibits a ratio or phase shift, this consistency will be disrupted. Therefore, this invention does not directly analyze common system-side fluctuations as abnormal features, but rather uses the consistency deviation between the measurement results of multiple CVTs as the identification entry point, highlighting individual differences through group comparison. Compared to the traditional independent analysis method for a single CVT, the concept of this invention can more effectively weaken the impact of common disturbances such as system-side voltage fluctuations and load changes on the identification results, thereby improving the ability to identify small errors, especially slight out-of-tolerance conditions.

[0070] See Figure 1 The specific implementation steps are as follows:

[0071] Step S1: Synchronously acquire the secondary voltage vector data of at least three capacitive voltage transformers (CVTs) of the same phase on the same bus at the same time section.

[0072] Specifically, the secondary voltage vector data includes at least voltage magnitude and phase information.

[0073] By limiting the calculation to "the same busbar, the same phase, and the same time section," it is ensured that all CVTs involved in the calculation measure the same primary-side voltage at the physical level. This provides an objective physical benchmark for subsequent vector collaborative matching. Furthermore, requiring at least three CVTs to participate is to meet the redundancy requirements of subsequent optimization solutions. Utilizing cross-verification of group information effectively improves the sensitivity to identifying minute errors in individual CVTs.

[0074] Step S2: Preprocess the secondary voltage vector data to obtain the target voltage vector data.

[0075] Specifically, preprocessing operations include one or more combinations of data validity verification, outlier removal, noise suppression, and time alignment.

[0076] Field measurement data often contains transient interference, communication delays, or gross sampling errors. Time alignment processing eliminates the non-true phase shift caused by asynchronous sampling; combined with outlier removal and noise suppression, it effectively reduces the masking effect of electromagnetic interference and white noise on the extraction of minute errors. This step ensures the purity of the input data, which is a prerequisite for achieving accurate perception of minute errors.

[0077] Step S3: Based on the target voltage vector data, rated transformation ratio and error compensation vector to be solved for each CVT, determine the estimated value of the primary side voltage after compensation for each CVT.

[0078] Specifically, the target voltage vector data of the secondary side is multiplied by the rated transformation ratio, and the error compensation vector to be solved is superimposed to obtain the mathematical expression of the primary side voltage.

[0079] To achieve error identification when the actual voltage on the system side is unknown, this invention does not directly calculate the actual voltage. Instead, it introduces an error compensation quantity to be solved for each CVT, and determines the estimated value of the primary side voltage for each CVT after compensation based on this. By compensating the original measurement vector and converting it to the primary side, the measurement results of each CVT are transformed into comparable estimates under the same physical quantity. Since these estimated values ​​of the primary side voltage after compensation should theoretically be as consistent as possible, the degree of difference between multiple CVTs can be used to infer the error of each CVT. By setting this "estimated value of the primary side voltage after compensation" step, this invention transforms the problem of "online error identification" into the problem of "consistency recovery of multiple CVTs," thereby avoiding the limitation of having to introduce an external standard source or dedicated calibration device, and enhancing the non-invasiveness and engineering applicability of the solution.

[0080] Step S4: Determine the average voltage vector based on the estimated value of the primary side voltage after compensation for each CVT, and construct an optimization model with the objective of minimizing the deviation of the estimated value of the primary side voltage after compensation for each CVT from the average voltage vector.

[0081] Specifically, since the true primary voltage of the same phase on the same bus is unique, the estimated primary voltage of multiple CVTs after compensation should theoretically be highly consistent, i.e., equal to the true voltage. Therefore, the average of the estimated primary voltage of each CVT is used as a virtual benchmark, and an objective function is constructed to minimize the sum of squares or the sum of absolute values ​​of the deviations between the estimated values ​​of each CVT and this virtual benchmark.

[0082] This step cleverly utilizes the relative consistency between multiple CVTs, allowing the construction of closed-loop equations without relying on an absolutely accurate external physical reference source. This fundamentally overcomes the dependence of traditional methods on high-precision hardware and significantly reduces error monitoring costs.

[0083] Step S5: Solve the optimization model under preset constraints to obtain the optimal error compensation vector for each CVT.

[0084] Specifically, the preset constraints can be set as the physical reasonable boundaries of the transformer error, such as the limit range of ratio difference and angle difference. Preferably, a nonlinear optimization algorithm or least squares method is used to perform global optimization on the above model.

[0085] Reasonable physical constraints can effectively prevent optimization algorithms from diverging or getting trapped in unrealistic local optima. By extracting error parameters through iterative solutions of a purely mathematical model, the limitations of power outage periodic inspections are completely eliminated, enabling online error identification under uninterrupted power conditions.

[0086] Step S6: Based on the optimal error compensation vector corresponding to each CVT, determine the ratio difference estimate and angle difference estimate of each CVT, and compare the ratio difference estimate and angle difference estimate with the preset error limit to identify the operating status of each CVT.

[0087] Specifically, the abstract error compensation vector is converted into standard ratio (relative amplitude error) and angle (phase error) indicators in the power system, and compared with the error limits in national standards or enterprise regulations to determine whether the equipment is in a normal, out-of-tolerance alarm, or fault state. This visualizes complex vector parameters into intuitive ratio and angle indicators, facilitating direct access by the operation and maintenance system and understanding by field personnel. By comparing limits in real time, the traditional "lagging maintenance" relying on periodic power outages for testing is transformed into data-driven "state awareness," greatly improving the timeliness of operational status assessment and avoiding the risk of protection malfunctions or metering inaccuracies caused by CVT operation with defects.

[0088] This invention forms a complete technical chain around the core concept of "recovering consistency constraints and identifying errors based on the group measurement information of multiple CVTs on the same bus." Synchronous data acquisition and preprocessing ensure the comparability of input data; error model establishment provides an analytical basis for compensation and inversion; the setting of the primary voltage estimate and average voltage vector after compensation enables a quantifiable group consistency reference among multiple CVTs; the combination of optimization and constraint conditions ensures that the error identification process balances physical rationality and algorithm stability; finally, through the output results of ratio difference, angle difference, and state identification, a closed-loop processing from raw measurement data to operational status assessment is achieved. It is precisely because of the clear synergy among the above technical features that this invention can effectively balance cost control, online identification, and the ability to resolve minute errors without requiring additional dedicated hardware.

[0089] In a preferred embodiment, the secondary voltage vector data may be derived from one or more of the following: an electricity consumption information acquisition system, a synchronous phasor measurement unit (PMU), a merging unit, or a metering acquisition terminal.

[0090] In one embodiment, existing data acquisition links in the substation are preferentially utilized to obtain secondary side voltage vector data of at least three CVTs on the same bus and in the same phase at the same time section. Specifically, the voltage amplitude and phase data corresponding to each CVT can be directly read from the electricity consumption information acquisition system. This approach eliminates the need for additional dedicated online monitoring hardware, allowing for direct reuse of existing metering and acquisition infrastructure, thereby reducing system deployment costs and on-site modification workload. It is well-suited for implementation in sites with well-established metering automation infrastructure.

[0091] In another embodiment, the secondary-side voltage vector data can also be derived from a synchronous phasor measurement unit (PMU). Since the PMU has good time synchronization capabilities and can output voltage phasor data with a unified time reference, using the PMU as a data source is beneficial for improving the time consistency between vector data from multiple CVTs, reducing additional errors caused by sampling timescale deviations, and thus providing a more stable data foundation for subsequent consistency comparisons and identification of minor errors among multiple CVTs.

[0092] In another embodiment, the secondary voltage vector data may also originate from a merging unit. The merging unit is typically connected to the substation's digital measurement link and can output synchronized sampled values ​​or corresponding electrical quantity information. Obtaining CVT secondary voltage vector data through a merging unit facilitates the use of existing information exchange channels between primary equipment and secondary systems within the substation, making it easier to implement the invention in a digital substation scenario and improving the convenience of data access.

[0093] In another embodiment, the secondary-side voltage vector data can also originate from a metering acquisition terminal. For sites already configured with metering acquisition terminals, the voltage data of each CVT can be collected through these terminals, and the corresponding amplitude and phase information can be further extracted. This approach facilitates integration with existing metering operation and maintenance platforms or condition monitoring platforms, and allows for direct embedding of the invention into existing business processes.

[0094] Furthermore, in practical applications, the secondary-side voltage vector data is not limited to originating from a single device or system, but can also originate from two or more of the aforementioned data sources. For example, high-precision synchronous phasor data can be provided by a PMU, while historical operating data and supplementary data can be provided by an electricity consumption information acquisition system or metering acquisition terminal. By jointly accessing data from multiple sources, the completeness of data coverage can be improved, and it is also beneficial to flexibly select appropriate data access methods according to the data conditions of different sites, thereby enhancing the engineering applicability of the present invention.

[0095] Therefore, by limiting the source of secondary voltage vector data to one or more of the following: electricity consumption information acquisition system, synchronous phasor measurement unit (PMU), merging unit, or metering acquisition terminal, this invention can fully utilize existing data resources in the substation to complete online CVT error identification without relying on additional dedicated monitoring devices. This not only helps reduce implementation costs but also provides an available and reusable data foundation for subsequent consensus analysis of multiple CVTs, thereby improving the applicability of this invention in actual power grid scenarios.

[0096] In a preferred embodiment, in step S2, for secondary voltage vector data with time-scale deviations, time alignment can be performed using interpolation or phase compensation based on system frequency. This processing step is included because the present invention subsequently requires collaborative comparison of voltage vectors from multiple CVTs on the same bus and in the same phase. If the data from different CVTs are not at the same reference time, even if each CVT itself does not have obvious metering anomalies, additional phase shifts may be introduced due to sampling time differences, thus affecting the consistency judgment of the primary voltage estimate after compensation. By completing time alignment in the preprocessing stage, the additional errors caused by asynchronous sampling can be largely extracted from the original data, thereby providing a more reliable data foundation for subsequent error compensation vector solving and small error identification. This has a positive effect on improving the timeliness and stability of operational status assessment.

[0097] Specifically, let the first The original secondary voltage vector of the CVT is The timescale deviation is The system frequency is Then the following formula can be used to calculate the first... Phase compensation is performed on the original secondary voltage vector of the CVT to obtain the time-aligned target voltage vector. :

[0098] ;

[0099] in, Indicates the first The secondary voltage vector obtained by the CVT at the original sampling time; This represents the time-aligned target voltage vector; Indicates the first The timescale deviation of the CVT data relative to the unified reference time; Indicates the system frequency; Represents the imaginary unit; Indicates exponentiation; This represents the CVT number. According to the above formula, the corresponding phase rotation is calculated based on the system frequency and time scale deviation, and the original secondary voltage vector is converted to a unified reference time point. After this processing, the voltage vectors of different CVTs can be compared under the same time reference, which helps reduce the interference of time scale deviation on group consistency analysis.

[0100] In practical applications, when the time scale deviation is small and the sampling points are relatively continuous, interpolation can also be used to complete time alignment; however, when the system frequency is known and the phasor representation is relatively stable, the phase compensation method based on the system frequency described above is preferred. Compared to directly using unaligned data for analysis, this processing can more accurately reflect the true phase relationship between each CVT, avoiding misjudging time asynchrony as equipment error, thereby helping to improve the accuracy of this invention in identifying slight ratio and angle shifts.

[0101] In a preferred embodiment, before step S3, an error model is first established between the CVT measurement value and the actual voltage on the system side. The purpose of setting the error model is to uniformly represent the deviation relationship between the CVT measurement result and the actual voltage on the system side, thereby providing an analytical basis for subsequent error compensation vector solution and primary side voltage consistency recovery after compensation. Since the technical problem to be solved by this invention is not only to identify obvious out-of-tolerance errors, but also to improve the ability to identify small ratio and phase deviations under uninterrupted power conditions, it is necessary to incorporate the amplitude and phase errors of each CVT into the same mathematical framework before entering the cooperative matching solution. This approach avoids the one-sidedness caused by empirical judgment based solely on a single amplitude fluctuation or a single phase change, and is conducive to improving the interpretability and consistency of subsequent identification results.

[0102] Specifically, let the first The system-side true voltage vector of the CVT is Rated ratio is The secondary side measured voltage vector is The difference is Angular difference is Randomly measured noise is Then the error model satisfies:

[0103] ;

[0104] in, Indicates the first The actual system-side voltage vector corresponding to the CVT;

[0105] Indicates the first The rated gear ratio of the CVT;

[0106] Indicates the first The secondary side voltage vector of the CVT is measured.

[0107] Indicates the first The difference between the CVT and the platform;

[0108] Indicates the first The angle difference of the CVT;

[0109] Indicates the first Random measurement noise vector of a CVT;

[0110] Indicates the difference of angles The resulting phase rotation factor;

[0111] Represents the imaginary unit; Indicates the CVT number.

[0112] As can be seen from the error model, the first... The actual measurement results of a CVT can be considered as the result of the system-side true voltage after being converted to the rated transformation ratio, and then superimposed with the effects of ratio difference, phase difference, and random measurement noise. This model allows the amplitude and phase offsets in CVT measurements to be unified into a single vector expression, which is beneficial for subsequent analysis of differences between multiple CVTs within the same coordinate frame. Compared to treating ratio difference and phase difference separately, this approach is more suitable for the scenario described in this invention, which involves online error identification based on the consistency of multiple CVT groups.

[0113] Furthermore, to facilitate subsequent collaborative matching solutions, the first... The measurement error of a CVT is represented as a complex error vector. ,satisfy:

[0114] ;

[0115] in, Indicates the first The complex error vector of the CVT;

[0116] Indicates the first The voltage vector that a CVT should correspond to under ideal, error-free conditions;

[0117] This represents the equivalent ideal measurement result after subtracting the complex error vector.

[0118] In this representation, the complex error vector The magnitude of the difference in magnitude corresponds to the magnitude of the complex error vector. The argument corresponds to the direction and magnitude of the angular difference. That is to say, It is not simply a scalar deviation, but a comprehensive vector quantity that simultaneously contains information on amplitude error and phase error. By introducing a complex error vector, on the one hand, the measurement error of the CVT can be explicitly separated from the original measurement vector, and on the other hand, a direct interface is provided for transforming the error identification problem into an error compensation vector solution problem.

[0119] In practical applications, when the ratio and angle differences of each CVT are small, the aforementioned error model and complex error vector representation can effectively reflect the deviation between the CVT measurement results and the ideal values, providing a foundation for constructing the primary voltage estimate after subsequent compensation. Especially in scenarios where this invention requires the identification of minor errors, standardizing the measured values ​​using this error model helps distinguish between equipment-specific errors and common system disturbances, thereby improving the targeting of subsequent collaborative matching optimization. Furthermore, since the error model is directly based on the relationship between CVT measurements, rated transformer ratio, actual system voltage, and ratio and angle differences, its physical meaning is clear, facilitating integration with subsequent primary voltage estimates, average voltage vectors, and state identification steps. By establishing this error model first and then conducting subsequent group collaborative matching analysis, the overall technical route of this invention can form a closed-loop processing procedure of "original measurement - error modeling - error compensation - consistency restoration - state identification," thereby improving the logical integrity and engineering applicability of the solution.

[0120] In a preferred embodiment, steps S3 and S4 specifically include: in the first... The target voltage vector after time alignment of the CVT is Rated ratio is The error compensation vector to be solved is In the case of the first Estimated primary voltage after compensation for each CVT satisfy:

[0121] ;

[0122] in, Indicates the first The target voltage vector of the CVT after time alignment; Indicates the first The rated gear ratio of the CVT; Indicates the first The error compensation vector to be solved for the CVT; Indicates the first The estimated primary voltage after compensation for each CVT; Indicates the CVT number.

[0123] This approach eliminates the need for an external standard source, allowing the direct utilization of the relative consistency among multiple CVTs to reflect equipment error status. This reduces implementation costs and improves the feasibility of online identification. Theoretically, for the same phase, all ( It should approximate the same true value. In other words, under ideal conditions of zero error or sufficient error compensation, the dispersion between the compensated primary voltage estimates for each CVT should be as small as possible. Based on this, this invention does not directly judge a single CVT individually, but first uses multiple CVTs to form a group central quantity as a reference for consistency analysis. Let the total number of CVTs participating in the collaborative matching be... Then the average voltage vector satisfy:

[0124] ;

[0125] in, This represents the average voltage vector of the primary side voltage estimates after compensation from multiple CVTs participating in coordinated matching; This represents the total number of CVTs participating in the collaborative matching process, and ; Indicates the first Estimated primary voltage after compensation of the CVT.

[0126] Using average voltage vector Using a single CVT as a reference for group consistency, rather than selecting a specific CVT as a fixed benchmark, helps avoid bias in the overall judgment caused by deviations in a single device. At the same time, it can make fuller use of the redundant information formed by multiple CVTs simultaneously measuring the same system voltage on the same bus and in the same phase, thereby enhancing the ability to suppress common disturbances on the system side.

[0127] Furthermore, after obtaining the compensated primary voltage estimates for each CVT... and average voltage vector Then, we can focus on "each Compared to The subsequent optimization solution is based on the objective of "minimizing the dispersion." In other words, this invention transforms the online CVT error identification problem into a problem of restoring the consistency of the primary side voltage after compensation: when a CVT has a ratio or angle deviation, its corresponding estimated value of the primary side voltage after compensation will deviate from the group center; by solving for a suitable error compensation vector, this deviation can be reduced, and the error state of each CVT can be deduced from this. This technical approach can more effectively separate the equipment's own error from the background of common system fluctuations, and is therefore more suitable for the online identification of small errors and slight out-of-tolerance states.

[0128] As a preferred embodiment, after obtaining the compensated primary voltage estimate for each CVT... and average voltage vector Subsequently, an optimization model is constructed with the objective of minimizing the deviation of the compensated primary voltage estimate of each CVT from the average voltage vector. The basic consideration for this approach is that, for multiple CVTs on the same bus and in the same phase, when there is no significant measurement deviation among the CVTs, or when their measurement deviations have been reasonably compensated, the compensated primary voltage estimate of each CVT should be as close as possible to the true voltage of the same system side. Therefore, the smaller the dispersion of each compensated primary voltage estimate relative to the central quantity of the group, the more reasonable the current error compensation result. Based on this idea, this invention does not directly use the fluctuation of the original measured value as the judgment criterion, but rather uses the degree of consistency restoration among the compensated results of multiple CVTs as the optimization objective, thereby more effectively suppressing the impact of common disturbances on the system side on the identification results.

[0129] In a preferred embodiment, the objective function of the optimization model Represented as:

[0130] ;

[0131] in, Represent the objective function; Indicates the first The weighting coefficients corresponding to each CVT; Indicates the first Estimated primary voltage after compensation for the CVT; This represents the average voltage vector of the primary side voltage estimates after compensation from multiple CVTs participating in coordinated matching; Represents the magnitude of a complex vector; This indicates the total number of CVTs participating in the collaborative matching process; Indicates the CVT number.

[0132] After constructing the above objective function, the optimization solution process can directly revolve around "restoring the group consistency of multiple CVTs for the same system voltage", rather than relying on a single CVT as a fixed reference. Therefore, it is beneficial to reduce the impact of the inaccuracy of a single device reference on the overall identification result.

[0133] Furthermore, the weighting coefficients The weighting can be determined based on at least one of the following: the historical reliability, service life, and manufacturer information of the corresponding CVT. For example, for CVTs with a longer service life, higher historical deviation risk, or lower reliability, a relatively high or targeted weight can be set to introduce necessary engineering priors during the optimization process. By setting weighting coefficients, the degree of influence of different CVTs in the recovery of group consistency can be adaptively adjusted, thereby making the optimization results more consistent with the actual operating conditions of the field equipment and helping to improve the targeting of error identification.

[0134] To ensure the solution results conform to engineering practice, the objective function is adjusted. During optimization, preset constraints are applied to the error compensation vectors corresponding to each CVT. Preferably, these preset constraints include at least magnitude constraints and phase angle constraints on the error compensation vectors. That is, the magnitude and direction of the error compensation are limited to a reasonable range to avoid the optimization algorithm obtaining compensation results that significantly deviate from actual operating conditions. By introducing such constraints, on the one hand, the error compensation results can be physically interpretable; on the other hand, it helps to reduce the impact of noise, local anomaly sampling, or search path fluctuations on the final results, thereby improving the stability of the optimization solution.

[0135] Specifically, the first Error compensation vector corresponding to CVT The following modulus and phase angle constraints must be satisfied:

[0136] ;

[0137] ;

[0138] in, Indicates the first Error compensation vector of the CVT; Indicates the first The magnitude of the CVT error compensation vector; Indicates the first The rated voltage of the secondary side of the CVT; This represents a preset percentage coefficient used to limit the upper limit of the error compensation vector magnitude; This indicates taking the phase angle; Indicates the lower limit of the preset phase angle; Indicates the preset upper limit of the phase angle; Indicates the CVT number.

[0139] In one embodiment, considering that the probability of multiple CVTs on the same busbar simultaneously experiencing significant anomalies is generally low, a sparsity constraint term can be further introduced into the above objective function to enhance the ability to locate a few abnormal devices. Objective function with added sparsity constraint Represented as:

[0140] ;

[0141] in, This represents the objective function after introducing sparsity constraints; Indicates the sparsity adjustment coefficient; Indicates the first Error compensation vector of the CVT; This represents the sparsity regularization term used to sum the magnitudes of all CVT error compensation vectors involved in the collaborative matching; the meanings of the other symbols are the same as described above.

[0142] Objective function after introducing sparsity constraints The technical significance lies in minimizing the deviation between the estimated primary voltage and the average voltage vector after compensation for each CVT, while also suppressing unnecessary large compensation amounts for a large number of CVTs in the group. This approach results in optimization that concentrates larger compensation amounts on a few CVTs with actual deviations, while keeping the majority of normal CVTs with smaller compensation amounts, thus better reflecting the common operating condition of "a few devices malfunctioning, most devices functioning normally" in real-world operation. For the technical problem of "difficulty in accurately identifying minute errors" that this invention aims to address, this sparsity constraint helps improve the clarity of locating abnormal devices and reduce misjudgments within the group.

[0143] In the specific solution process, particle swarm optimization, genetic algorithm, or gradient descent algorithm are used to solve the optimization model to obtain the optimal error compensation result for each CVT. Since the aforementioned objective function and constraints together constitute a nonlinear constrained optimization problem, it is often difficult to simultaneously balance solution efficiency and engineering applicability if solved directly using analytical methods. However, by using particle swarm optimization, genetic algorithm, or gradient descent algorithm, an error compensation solution that minimizes the objective function within a given constraint range can be searched, thereby making the estimated primary voltage values ​​of multiple CVTs as consistent as possible. With this processing, the present invention can complete the error compensation calculation using existing acquired data without adding dedicated hardware, providing a foundation for subsequent ratio difference, angle difference estimation, and state identification.

[0144] In one embodiment, a particle swarm optimization algorithm is preferably used for solving the problem. Specifically, all error compensation vectors to be solved are used as optimization variables, and the aforementioned objective function value is used as the fitness evaluation criterion. Under the condition of satisfying preset constraints, the candidate compensation result at the current moment is obtained through iterative search of the solution space by the particle swarm. The advantage of using the particle swarm optimization algorithm is that the implementation method is relatively direct, which is convenient for handling multi-variable collaborative optimization problems and is suitable for the joint solution of error compensation of multiple CVTs in this invention. For the genetic algorithm, a better solution can be gradually approximated through population selection, crossover and mutation operations; for the gradient descent algorithm, the compensation result that satisfies the constraints can be obtained by iteratively updating the variables to be solved along the descent direction of the objective function. Different algorithms can be used to implement the collaborative matching solution of this invention, and their common goal is to restore the group consistency of multiple CVTs for the same system voltage.

[0145] As a preferred embodiment and a further improvement to the above technical solution, after obtaining the optimal error compensation vector for each CVT through optimization, the ratio difference estimate and angle difference estimate for each CVT are further calculated based on the optimal error compensation vector. The purpose of this step is to further convert the compensation results obtained from the aforementioned collaborative matching solution into error indicators with greater engineering significance in CVT metrological performance evaluation, enabling the identification results to directly correspond to the ratio difference and angle difference criteria in existing metrological verification and operation and maintenance. After this processing, the output results of the present invention no longer remain at the level of abstract optimization variables, but form metrological error quantities that can be directly used for condition assessment, thereby improving the interpretability and practical application value of the identification results.

[0146] In a preferred embodiment, let the first... The optimal error compensation vector corresponding to the CVT is , No. The rated ratio of the CVT is The average voltage vector of the primary side voltage estimates after compensation from multiple CVTs participating in the coordinated matching is: Then the first Estimated ratio of CVT difference Sum of angle difference estimates They can be represented as:

[0147] ;

[0148] ;

[0149] in, Indicates the first The optimal error compensation vector obtained by optimizing the CVT; Indicates the first The rated gear ratio of the CVT; This represents the average voltage vector of the primary side voltage estimates after compensation from multiple CVTs participating in coordinated matching; Indicates the first The estimated ratio difference of the CVTs; Indicates the first The estimated angle difference of the CVT; This indicates the operation of taking the real part; This indicates the operation of taking the imaginary part; Indicates the CVT number.

[0150] Using the above formula, the first... After normalizing the optimal error compensation vector of a CVT relative to the primary side group reference, its real part is used to characterize the error component in the amplitude direction, i.e., the ratio difference estimate; its imaginary part is used to characterize the error component in the phase direction, i.e., the angle difference estimate. Through this processing, the compensation result obtained by the aforementioned cooperative matching optimization model can be further mapped into an error index with clear physical meaning, thereby realizing the transformation from "group consistency restoration" to "individual equipment error quantification".

[0151] Furthermore, due to These values ​​are derived from the group center values ​​of the primary side voltage estimates of multiple CVTs under the same bus and phase. Therefore, the above ratio difference estimates and angle difference estimates are not based on isolated measurements of a single device, but on the collaborative analysis of multiple CVTs as a group. This approach can more effectively reduce the impact of system-side common fluctuations, load changes, and common-mode disturbances, thus making it easier to identify slight ratio difference offsets and small angle difference changes.

[0152] In practical applications, after calculating the first... Estimated ratio of CVT difference Sum of angle difference estimates Then, it can be further compared with a preset error limit to determine the operating status of the CVT. The preset error limit can be determined according to the verification procedure for the corresponding accuracy level. Preferably, the operating status can be divided into three levels: normal, caution, and abnormal. and When all indicators are within the allowable range, it is considered normal; when it has not exceeded the error limit but is close to the preset threshold, it is considered a warning; when any indicator exceeds the error limit, it is considered abnormal. This process not only identifies whether there are error anomalies but also distinguishes the degree of error deviation, thus providing maintenance personnel with more targeted early warning information and a basis for handling.

[0153] To further illustrate the inventive concept of this invention, a section I of a 110kV busbar in a 220kV substation is used as an implementation scenario. This section of the busbar connects four outgoing lines L1, L2, L3, and L4, each equipped with a capacitive voltage transformer (CVT), denoted as CVT1, CVT2, CVT3, and CVT4. All four CVTs are of the same model and have the same rated transformation ratio. The substation's energy acquisition system synchronously acquires the fundamental vector data of the A-phase voltage on the secondary side of the four CVTs at a rate of one frame per second through a merging unit or independent acquisition terminal. This data includes amplitude and phase information, with phase measurement using a unified synchronous clock as a reference. This embodiment demonstrates that the present invention can directly utilize data from existing acquisition links within the substation for analysis, eliminating the need for additional dedicated online monitoring hardware, thereby reducing implementation costs and improving engineering applicability.

[0154] First, the acquired synchronous data frames undergo validity verification, outlier screening, and noise suppression. For historical data with timescale discrepancies, further time alignment is performed. Let the... The original secondary voltage vector of the CVT is The timescale deviation is The system frequency is Then the time-aligned target voltage vector It can be represented as:

[0155] ;

[0156] This process can unify data from different sampling times to the same reference time point, reducing the impact of asynchronous sampling on subsequent phase comparison and error identification.

[0157] After preprocessing, an error model is established between the actual voltage on the system side and the CVT measurement value. Let the actual voltage vector of phase A on the system side be... , No. The rated ratio of the CVT is The secondary side measured voltage vector is The difference is Angular difference is Randomly measured noise is Then we have:

[0158] ;

[0159] Furthermore, when the ratio difference and angle difference are small, the first... Complex error vector of CVT for:

[0160] ;

[0161] And there are:

[0162] ;

[0163] in, Indicates the first The complex error vector of a CVT represents the absolute vector deviation of its measured value from the ideal value. This error model allows for the integration of ratio error and angle error into a single vector framework, providing a foundation for subsequent compensation solutions.

[0164] Subsequently, the first Error compensation vector of CVT Construct the estimated value of the primary side voltage after compensation. :

[0165] ;

[0166] Let the total number of CVTs participating in the collaborative matching be... Then the average voltage vector for:

[0167] ;

[0168] Furthermore, constructing a system based on various Compared to The objective function is to minimize the deviation. :

[0169] ;

[0170] In this embodiment, the weighting coefficients are set according to the years of operation of the equipment: CVT1 and CVT2 have been in operation for 12 years, and their weights are set to 1.2; CVT3 and CVT4 have been in operation for 8 years, and their weights are set to 1.0. By setting this, the engineering prior of the equipment's years of operation can be introduced into the optimization process, making the solution results more consistent with the actual situation on site.

[0171] During the solution phase, reasonable constraints are imposed on the error compensation vector, and a particle swarm optimization algorithm is used for solving. To improve the stability of the solution, this embodiment employs five independent runs, selecting the result with the minimum objective function value as the final solution; simultaneously, the optimization result from the previous time step is used as the population initialization center for the current time step. After approximately 150 iterations, the algorithm converges and obtains the optimal error compensation vector for each CVT. This process helps reduce the impact of random initialization and differences in local search paths on the results, thereby improving the stability of the results in continuous online recognition scenarios.

[0172] Obtain the optimal error compensation vector for each CVT. Then, the estimated ratio difference and angle difference of each CVT were further calculated:

[0173] ;

[0174] ;

[0175] in, Indicates the first The estimated ratio difference of the CVTs; Indicates the first The estimated angle difference of the CVT; Indicates taking the real part; This indicates the imaginary part is taken. Through this step, the compensation results obtained from collaborative matching can be directly converted into the ratio difference and angle difference indices commonly used in engineering, which facilitates integration with existing verification procedures and operation and maintenance criteria.

[0176] In this embodiment, it is assumed that the accuracy class of all four CVTs is 0.2, and their status is assessed based on the error limits for the corresponding accuracy class. The results show that CVT1 is classified as "normal"; CVT2 and CVT4 are generally in a "normal" state, but CVT4 exhibits a certain negative deviation and requires continuous monitoring; although CVT3 is not exceeding the limit, its deviation has reached 41.5% of the limit, therefore it can be classified as "attention," and a suggestion is made to strengthen monitoring or arrange a pre-test. This result demonstrates that the present invention can not only identify equipment with obvious deviations but also identify its offset trend before the equipment enters a serious abnormality, thereby improving the timeliness of identifying minor errors and early anomalies.

[0177] As can be seen from the above embodiments, this invention is based on the objective fact that multiple CVTs simultaneously measure the same system voltage under the same bus and phase. It sequentially processes data preprocessing, error model establishment, construction of primary voltage estimates after compensation, group consistency optimization, and conversion of ratio and phase differences, ultimately outputting the status identification results of each CVT. This technical approach does not require additional dedicated online monitoring hardware, and can effectively balance implementation cost, online identification capability, and the ability to resolve minor errors, making it suitable for online evaluation scenarios of CVT operating status in substations.

[0178] Example 2

[0179] This invention also provides an online error identification system based on multi-CVT vector collaborative matching, used to implement the online error identification method described in Embodiment 1. The system can be deployed in a metering and maintenance platform, a condition monitoring platform, or a data analysis platform connected to an in-station acquisition link, to achieve online error identification and operational status assessment of multiple CVTs on the same bus and under the same phase without power outages.

[0180] Specifically, the system includes a data acquisition module, a preprocessing module, a collaborative matching analysis module, an error calculation and status assessment module, and an early warning output module. These modules can be implemented using software, hardware, or a combination of both. The modules are connected via data interfaces to form a continuous processing flow from raw voltage vector data acquisition to final early warning output.

[0181] The data acquisition module is used to synchronously acquire secondary-side voltage vector data of at least three CVTs of the same phase on the same bus at the same time section. Preferably, the data acquisition module is connected to one or more of the following: an electricity consumption information acquisition system, a synchronous phasor measurement unit (PMU), a merging unit, or a metering acquisition terminal, to acquire secondary-side voltage vector data containing voltage amplitude and phase information. This module allows for direct reuse of existing acquisition links within the station, eliminating the need for additional dedicated online monitoring terminals, thereby reducing system implementation costs and providing a synchronous data foundation for subsequent group consistency analysis of multiple CVTs.

[0182] The preprocessing module is used to perform data validity verification, outlier removal, noise suppression, and time alignment on the secondary voltage vector data. Specifically, the preprocessing module can check the validity of the voltage amplitude and phase ranges, identify and remove abnormal sampling points, suppress noise in data with significant random fluctuations, and align data with time scale deviations. This module improves the consistency and comparability of the input data, reduces the impact of abnormal sampling, random noise, and time asynchrony on subsequent error identification results, and thus provides a more reliable data foundation for online identification of minor errors.

[0183] The collaborative matching analysis module is used to establish a CVT error model, determine the estimated primary-side voltage and average voltage vector of each CVT after compensation, construct and solve an optimization model to obtain the optimal error compensation vector for each CVT. Specifically, the collaborative matching analysis module first establishes an error model based on the secondary-side voltage vector, rated turns ratio, and error relationship of each CVT. Then, it introduces the error compensation amount to be solved, constructs the estimated primary-side voltage of each CVT after compensation, and forms an average voltage vector. Further, it constructs an optimization model with the objective of minimizing the deviation of the estimated primary-side voltage of each CVT after compensation from the average voltage vector. Then, under preset constraints, it uses a particle swarm optimization algorithm, a genetic algorithm, or a gradient descent algorithm to solve the optimization model to obtain the optimal error compensation vector for each CVT. Through this module, the isolated analysis problem of a single CVT can be transformed into a group consistency recovery problem of multiple CVTs on the same bus, thus improving the ability to identify minor errors.

[0184] The error calculation and status assessment module is used to determine the ratio difference estimate and angle difference estimate of each CVT based on the optimal error compensation vector corresponding to each CVT, and to identify the operating status of each CVT according to preset error limits. Specifically, the error calculation and status assessment module converts the optimal error compensation result into the ratio difference estimate and angle difference estimate of each CVT, and compares them with the preset error limits of the corresponding accuracy level to identify whether each CVT is in a normal, alert, or abnormal state. Through this module, the intermediate results obtained from collaborative matching analysis can be further converted into error indicators and status conclusions that are easy for operation and maintenance personnel to understand and use, thereby improving the engineering practicality of the identification results.

[0185] The early warning output module is used to output identification results and early warning information. Preferably, the early warning output module can output the ratio difference estimate, angle difference estimate, status level, and corresponding early warning prompts of each CVT in the form of page display, report, alarm list, SMS, message push, or work order. Through this module, the analysis results of the present invention can be directly converted into actionable operation and maintenance prompts, thereby supporting on-site status monitoring, early warning linkage, and subsequent maintenance arrangements.

[0186] Furthermore, during system operation, the data acquisition module, preprocessing module, collaborative matching analysis module, error calculation and status assessment module, and early warning output module work together according to a predetermined process: the data acquisition module is responsible for acquiring the raw voltage vector data; the preprocessing module is responsible for outputting the aligned and cleaned target data; the collaborative matching analysis module is responsible for solving the optimal error compensation vector; the error calculation and status assessment module is responsible for completing error quantification and status identification; and the early warning output module is responsible for displaying the results and outputting early warning information.

[0187] The system revolves around the core concept of "restoring consistency and identifying errors by utilizing group measurement information from multiple CVTs on the same busbar." It modularly implements data acquisition, data preprocessing, collaborative matching analysis, error quantification, and early warning output. Without requiring additional dedicated online monitoring hardware, it effectively balances implementation cost, online identification capability, and timeliness of operational status assessment, making it suitable for online monitoring and early warning scenarios of CVT metering status in substations.

[0188] Example 3

[0189] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein, when the program is executed, it controls the device on which the storage medium is located to perform some or all of the steps in Embodiment 1.

[0190] The computer-readable storage medium may include high-speed RAM memory, and may also include nonvolatile memory (NVM), such as at least one disk storage device. It is understood that the storage medium can be any machine-readable medium capable of storing program code, such as random access memory (RAM), magnetic disk, hard disk, solid state disk (SSD), or nonvolatile memory.

[0191] Those skilled in the art will understand that embodiments of the present invention can be provided as methods or storage media. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0192] The above are merely preferred embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct or indirect applications in other related technical fields, are within the patent protection scope of the present invention.

Claims

1. An online error identification method based on multi-CVT vector collaborative matching, characterized in that, The steps include: S1. Synchronously acquire the secondary side voltage vector data of at least three capacitive voltage transformers (CVTs) of the same phase on the same bus at the same time section, wherein the secondary side voltage vector data includes at least voltage amplitude and phase information. S2. Preprocess the secondary voltage vector data to obtain the target voltage vector data; S3. Based on the target voltage vector data, rated ratio and error compensation vector to be solved for each CVT, determine the estimated value of the primary side voltage after compensation for each CVT. S4. Determine the average voltage vector based on the estimated value of the primary side voltage after compensation for each CVT, and construct an optimization model with the objective of minimizing the deviation of the estimated value of the primary side voltage after compensation for each CVT from the average voltage vector. S5. Solve the optimization model under preset constraints to obtain the optimal error compensation vector for each CVT. S6. Based on the optimal error compensation vector corresponding to each CVT, determine the ratio difference estimate and angle difference estimate of each CVT, and compare the ratio difference estimate and angle difference estimate with the preset error limit to identify the operating status of each CVT.

2. The online error identification method based on multi-CVT vector collaborative matching according to claim 1, characterized in that, In step S2, for secondary voltage vector data with time-scale deviations, time alignment is performed using interpolation or phase compensation based on system frequency; in the... The original secondary voltage vector of the CVT is Time scale deviation is The system frequency is In the case of the following formula, the first... Phase compensation is performed on the original secondary voltage vector of the CVT to obtain the time-aligned target voltage vector. : ； in, It is the imaginary unit.

3. The online error identification method based on multi-CVT vector collaborative matching according to claim 1 or 2, characterized in that, Before step S3, an error model is established between the CVT measurement value and the actual voltage on the system side. The system-side true voltage vector of the CVT is Rated ratio is The secondary side measured voltage vector is The difference is Angular difference is Random measurement noise is In the case of this, the error model satisfies: ； Furthermore, the first The measurement error of a CVT is represented as a complex error vector. ,satisfy: ； Wherein, the complex error vector The magnitude of the difference in magnitude corresponds to the magnitude of the complex error vector. The argument corresponds to the direction and magnitude of the angular difference.

4. The online error identification method based on multi-CVT vector collaborative matching according to claim 1 or 2, characterized in that, Steps S3 and S4 specifically include: in the first step The target voltage vector after time alignment of the CVT is Rated ratio is The error compensation vector to be solved is In the case of the first Estimated primary voltage after compensation for each CVT satisfy: ； Let the total number of CVTs participating in the collaborative matching be... Then the average voltage vector satisfy: ； in, .

5. The online error identification method based on multi-CVT vector collaborative matching according to claim 4, characterized in that, The objective function of the optimization model in step S4 Represented as: ； in, Let be the objective function. For the first The weighting coefficients corresponding to the CVTs. The magnitude of a complex vector; the first Weighting coefficients corresponding to CVT It is determined based on at least one of the following: the historical reliability of the corresponding CVT, its years of service, and manufacturer information.

6. The online error identification method based on multi-CVT vector collaborative matching according to claim 1 or 2, characterized in that, The preset constraints in step S5 include at least the following: Error compensation vector corresponding to CVT The modulus constraint and phase angle constraint satisfy: ； ； in, For the first The rated voltage of the secondary side of the CVT. This is a preset percentage coefficient. This indicates taking the phase angle. and These are the preset lower limit of the phase angle and the preset upper limit of the phase angle, respectively.

7. The online error identification method based on multi-CVT vector collaborative matching according to claim 6, characterized in that, The optimization model further includes a sparsity constraint term to ensure that the solved error compensation vector is sparsely distributed across multiple CVTs. The objective function after introducing the sparsity constraint... Represented as: ； in, To introduce the sparsity constraint into the objective function, This is the sparsity adjustment coefficient.

8. The online error identification method based on multi-CVT vector collaborative matching according to claim 1 or 2, characterized in that, In step S5, the optimization model is solved using a particle swarm optimization algorithm, a genetic algorithm, or a gradient descent algorithm, and at least one of the following methods is used to improve the stability of the solution: The optimization process is run independently multiple times, and the solution with the minimum objective function value is selected as the final solution. The solution result from the previous time step is used as the initialization center for the current time step optimization solution.

9. The online error identification method based on multi-CVT vector collaborative matching according to claim 1 or 2, characterized in that, In step S6, at the first The optimal error compensation vector corresponding to the CVT is In the case of the following formula, the first one is determined. Estimated ratio of CVT difference Sum of angle difference estimates : ； ； in, Indicates taking the real part, This indicates taking the imaginary part; and the estimated value of the ratio difference Sum of angle difference estimates Compare with the error limit for the corresponding accuracy class to determine the accuracy level. The operating status of the CVT is divided into normal, warning, or abnormal.

10. An online error identification system based on multi-CVT vector collaborative matching, used to implement the online error identification method according to any one of claims 1-9, characterized in that, The system includes: The data acquisition module is used to synchronously acquire the secondary side voltage vector data of at least three CVTs of the same phase on the same bus at the same time section. The preprocessing module is used to perform data validity verification, outlier removal, noise suppression, and time alignment on the secondary side voltage vector data. The collaborative matching analysis module is used to establish the CVT error model, determine the estimated value of the primary side voltage and the average voltage vector after compensation for each CVT, construct and solve the optimization model to obtain the optimal error compensation vector for each CVT. The error calculation and status assessment module is used to determine the ratio difference estimate and angle difference estimate of each CVT based on the optimal error compensation vector corresponding to each CVT, and to identify the operating status of each CVT according to the preset error limit. The early warning output module is used to output the identification results and early warning information.

Images