Early turn-to-turn short circuit monitoring method, device and equipment of transformer and storage medium

Abstract

The invention discloses an early fault monitoring method, device and equipment for a transformer and a storage medium, and the method comprises the steps: arranging magneto-optical measurement sensors at the two ends and the middle part of a winding in an iron core, obtaining the magnetic leakage induction intensity of the sensors and the current of a high-voltage winding, constructing a coefficient analysis model, and calculating the fault of the transformer; and identifying each coefficient of the current coefficient analysis model through a least square parameter algorithm to obtain an identification result, calculating an identification coefficient change rate based on the identification result and a leakage magnetic induction coefficient of the winding in a normal state, and when the transformer is not no-load and the iron core is not magnetized and saturated, if the identification coefficient change rate is greater than a change rate threshold value, determining that the transformer is in a non-load state. And determining that the transformer has turn-to-turn short circuit. It can be seen that the coefficient analytical model reflects the analytical calculation relation between the current and the magnetic flux leakage induction intensity, through the combination of the analytical model and the recursive least square algorithm, signal offset caused by load fluctuation and power grid disturbance is effectively filtered out, and the stability and accuracy of transformer early-stage turn-to-turn short circuit recognition are improved.

Description

Technical Field

[0001] The present application relates to the technical field of transformer monitoring, and more specifically, to a method, device, equipment, and storage medium for early turn-to-turn short circuit monitoring of a transformer. Background Art

[0002] Large power transformers, mostly single-phase in design, are critical equipment in power systems. Failures can easily cause widespread power outages. Most failures stem from deteriorating winding insulation. Minor interturn shorts can be subtle, with subtle changes in terminal current, making them difficult to identify in a timely manner and easily developing into serious incidents.

[0003] The current detection method is still mainly based on abnormal signal recognition, but due to the signal offset caused by transformer load fluctuations and grid disturbances, the accuracy and stability of early transformer turn-to-turn short circuit recognition are poor.

[0004] How to accurately monitor the early inter-turn short circuit of power transformers, timely identify the early faults of transformers, and ensure the safe operation of power systems are issues that need attention. Summary of the Invention

[0005] In view of the above problems, the present application provides a transformer early fault monitoring method, device, equipment and storage medium to accurately identify early turn-to-turn short circuits of transformers and ensure the safe operation of the power system.

[0006] In order to achieve the above objectives, the following specific plans are proposed:

[0007] A method for monitoring early-stage turn-to-turn short circuits in a transformer is provided, the method being applied to a monitoring module connected to a plurality of magneto-optical measurement sensors, wherein the magneto-optical measurement sensors are mounted at both ends of a winding inside a transformer core and at the axial center of the winding inside the core, the winding including a high-voltage winding.

[0008] The method includes:

[0009] respectively obtaining the leakage magnetic induction intensity of each magneto-optical measurement sensor and the first current flowing through the high-voltage winding;

[0010] constructing a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors;

[0011] Identifying the coefficients of the current magnetic field coefficient analytical model by a least squares parameter algorithm to obtain coefficient identification results;

[0012] Calculating an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state;

[0013] When the transformer is not no-loaded and the iron core is not magnetized and saturated, if the identification coefficient change rate is greater than a change rate threshold, it is determined that an inter-turn short circuit exists in the transformer.

[0014] Optionally, the identifying the coefficients of the current magnetic field coefficient analytical model by a least squares parameter algorithm to obtain coefficient identification results includes:

[0015] Acquire multiple groups of sampling data, and select one group of the sampling data as the target group of sampling data;

[0016] Initialize the forgetting factor, magnetic flux leakage prediction error, coefficient identification matrix and covariance matrix;

[0017] Using a recursive least squares algorithm to solve a model coefficient calculation formula, calculating a gain matrix and a magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, updating the magnetic flux leakage prediction error matrix, the coefficient identification matrix, and the covariance matrix to obtain a new magnetic flux leakage prediction error matrix, a new coefficient identification matrix, and a new covariance matrix;

[0018] Adaptively updating the forgetting factor through a forgetting factor adaptive update function to obtain a new forgetting factor;

[0019] Determining whether there is any sample data that has not been used for recursive calculation;

[0020] If so, one of the groups of sampled data that have not been used for recursive calculation is selected as a new target group of sampled data, and the process returns to the step of using the recursive least squares algorithm to solve the model coefficient calculation formula, calculate the gain matrix and the magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group of sampled data, and update the magnetic flux leakage prediction error matrix, the coefficient identification matrix, and the covariance matrix to obtain a new magnetic flux leakage prediction error matrix, a new coefficient identification matrix, and a new covariance matrix.

[0021] If not, the coefficient identification result of the current magnetic field coefficient analytical model is determined according to the coefficient identification matrix.

[0022] Optionally, the current magnetic field coefficient analytical model is:

[0023]

[0024] in, is a magnetic leakage matrix constructed based on the magnetic leakage induction intensity of each magneto-optical measurement sensor, is a current matrix constructed based on the first current, is the coefficient identification matrix.

[0025] Optionally, the recursive least squares algorithm is used to calculate the model coefficients:

[0026]

[0027] Among them, represents the target group sampling data, is the magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the magnetic leakage matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the current matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the previous target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the gain matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the covariance matrix under the previous target group sampling data, is the covariance matrix under the target group sampling data, is the identity matrix, is the forgetting factor under the target group sampling data, is the matrix transpose symbol;

[0028] The forgetting factor adaptive update function is:

[0029]

[0030] in, is the upper limit of the forgetting factor, is the lower limit of the forgetting factor, is the initial magnetic flux leakage prediction error, is the rounding function, is the magnetic flux leakage prediction error matrix The total error of the leakage magnetic output prediction results of each element.

[0031] Optionally, the method further includes:

[0032] When the transformer is unloaded or satisfies a first condition, the transformer is locked, wherein the first condition is:

[0033]

[0034] in, is the high voltage side current of the transformer, is the rated current of the transformer.

[0035] Optionally, the method further includes:

[0036] When the core is magnetized and saturated or the transformer meets a second condition, the transformer is locked, wherein the second condition is that the percentage of the second harmonic content of the differential current of the transformer is greater than a percentage threshold, and the differential current is the processed vector difference of the secondary currents of the current transformers on each side of the transformer.

[0037] Optionally, the change rate threshold is:

[0038]

[0039] in, is the change rate threshold, is the noise weighting coefficient, is the accuracy error of the current transformer sampling the winding, is the measurement accuracy error of the magneto-optical measurement sensor.

[0040] A transformer early turn short circuit monitoring device is applied to a monitoring module, wherein the monitoring module is connected to a plurality of magneto-optical measurement sensors, wherein each magneto-optical measurement sensor is respectively installed at both ends of a winding inside the iron core of the transformer and at the axial center of the winding inside the iron core, wherein the winding includes a high-voltage winding;

[0041] The device includes:

[0042] a measurement information acquisition unit, configured to respectively acquire the leakage magnetic induction intensity of each of the magneto-optical measurement sensors and the first current flowing through the high-voltage winding;

[0043] a current magnetic field coefficient analytical model construction unit, configured to construct a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors;

[0044] A system identification unit, configured to identify the coefficients of the current magnetic field coefficient analytical model by using a least squares parameter algorithm to obtain coefficient identification results;

[0045] an identification coefficient change rate calculation unit, configured to calculate an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state;

[0046] The turn-to-turn short circuit judgment unit is configured to determine that a turn-to-turn short circuit exists in the transformer if the identification coefficient change rate is greater than a change rate threshold when the transformer is not no-loaded and the iron core is not magnetized and saturated.

[0047] Optionally, the system identification unit includes:

[0048] a target group sampling data selection unit, configured to obtain multiple groups of sampling data and select one of the groups of sampling data as the target group sampling data;

[0049] Parameter initialization unit, used to initialize the forgetting factor, magnetic flux leakage prediction error, coefficient identification matrix and covariance matrix;

[0050] a matrix calculation and updating unit, configured to solve a model coefficient calculation formula using a recursive least squares algorithm, calculate a gain matrix and a magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, update the magnetic flux leakage prediction error matrix, the coefficient identification matrix, and the covariance matrix, and obtain a new magnetic flux leakage prediction error matrix, a new coefficient identification matrix, and a new covariance matrix;

[0051] a forgetting factor updating unit, configured to adaptively update the forgetting factor by using a forgetting factor adaptive updating function to obtain a new forgetting factor;

[0052] an adopted data existence judgment unit, configured to judge whether the sampled data has not been used for recursive calculation, and if so, to execute a new target adopted data determination unit; if not, to execute a coefficient identification result determination unit;

[0053] The new target adopted data determination unit is used to select one group of sample data from each group of sample data that has not been used for recursive calculation as a new target group of sample data, and return the selected group to execute the matrix calculation update unit;

[0054] The coefficient identification result determination unit is used to determine the coefficient identification result of the current magnetic field coefficient analytical model according to the coefficient identification matrix.

[0055] Optionally, the device further includes:

[0056] The first locking unit is configured to lock the transformer when the transformer is unloaded or a first condition is satisfied, wherein the first condition is:

[0057]

[0058] in, is the high voltage side current of the transformer, is the rated current of the transformer.

[0059] Optionally, the device further includes:

[0060] a second locking unit, configured to lock the transformer when the core is magnetized to saturation or the transformer satisfies a second condition, wherein the second condition is that a percentage of a second harmonic content of a differential current of the transformer is greater than a percentage threshold, and the differential current is a processed vector difference of secondary currents of current transformers on each side of the transformer.

[0061] A transformer early turn short circuit monitoring device includes a memory and a processor;

[0062] The memory is used to store programs;

[0063] The processor is used to execute the program to implement the various steps of the transformer early turn-to-turn short circuit monitoring method as described above.

[0064] A storage medium stores a computer program, which, when executed by a processor, implements the various steps of the transformer early turn-to-turn short circuit monitoring method as described above.

[0065] By means of the above technical solution, the present application sets magneto-optical measurement sensors at both ends of the winding and in the middle of the winding in the core, respectively obtains the leakage magnetic induction intensity of each sensor and the first current flowing through the high-voltage winding, constructs a current magnetic field coefficient analytical model, identifies the coefficients of the current magnetic field coefficient analytical model through the least squares parameter algorithm, obtains the coefficient identification result, calculates the identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding under normal conditions, and calculates the identification coefficient change rate when the transformer is not no-loaded and the core is not magnetized and saturated. If the identification coefficient change rate is greater than the change rate threshold, it is determined that the transformer has an inter-turn short circuit. It can be seen that the current magnetic field coefficient analytical model reflects the analytical calculation relationship between current and leakage magnetic induction intensity, and can perform real-time quantitative analysis of the local magnetic field disturbance caused by a slight inter-turn short circuit in the transformer winding. By combining the analytical model with the recursive least squares algorithm, the signal offset caused by load fluctuations and power grid disturbances can be effectively filtered out, thereby improving the stability and accuracy of early inter-turn short circuit identification of the transformer. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present application. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0067] Figure 1 A schematic diagram of the arrangement of the magneto-optical measurement sensor provided in an embodiment of the present application;

[0068] Figure 2A schematic diagram of a process for implementing early transformer fault monitoring provided in an embodiment of the present application;

[0069] Figure 3 A schematic diagram of a flow chart for identifying an analytical model of current magnetic field coefficients using a least squares parameter algorithm provided in an embodiment of the present application;

[0070] Figure 4 A schematic diagram of the structure of a device for realizing early fault monitoring of a transformer provided in an embodiment of the present application;

[0071] Figure 5 A schematic diagram of the structure of a device for implementing early fault monitoring of a transformer provided in an embodiment of the present application. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0073] The solution of the present application can be implemented based on a terminal with data processing capabilities, which can be a monitoring module. The monitoring module is connected to a plurality of magneto-optical measurement sensors to obtain leakage magnetic induction intensity data transmitted by the magneto-optical measurement sensors in real time.

[0074] The arrangement of the magneto-optical measurement sensors may be such that some are installed at both ends of the winding inside the iron core of the transformer, and some are installed at the axial center of the winding inside the iron core. Figure 1 As shown, magneto-optical measurement sensors x1 and x2 are installed at the ends of the winding inside the transformer core to effectively monitor inter-turn short-circuit faults in the upper and lower end regions. Magneto-optical measurement sensor y3 can be installed in the axial center of the winding inside the core to identify inter-turn short-circuit anomalies that may occur in the middle section of the winding. The winding can include high-voltage windings and low-voltage windings.

[0075] Each core window of the transformer can be set as follows Figure 1 The measurement points (x1, x2, y3) are shown. It can be understood that this measurement point layout takes into account both fault sensitivity and structural feasibility, providing support for the accurate positioning and effective monitoring of inter-turn short circuits at different locations.

[0076] Next, combine Figure 1 The transformer early turn short circuit monitoring method of the present application may include the following steps:

[0077] Step S110 , respectively obtaining the leakage magnetic induction intensity of each magneto-optical measurement sensor and the first current flowing through the high-voltage winding.

[0078] Specifically, the first current flowing through the high-voltage winding and the second current flowing through the low-voltage winding can be collected by current transformers, and the current transformers can be respectively arranged at the outlets of the respective windings.

[0079] Step S120: constructing a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each magneto-optical measurement sensor.

[0080] It is understandable that the leakage magnetic induction intensity sampling data and the current sampling data are used to form the leakage magnetic induction intensity matrix B and the current matrix I, and the relationship between the two is expressed in the standard form of the least squares algorithm as follows:

[0081]

[0082] in, 、 、 They represent the magnitude of the leakage magnetic induction intensity corresponding to the measuring points x1, x2 and y3 respectively. and Represent the first current and the second current respectively, the matrix is the coefficient matrix, 、 They represent the coefficient of the i-th measuring point of the high-voltage winding and the coefficient of the i-th measuring point of the low-voltage winding, i=1,2,3 respectively.

[0083] Further understanding shows that after a short circuit occurs between turns of the transformer, since the short-circuit turn current cannot actually be obtained, the final reaction is that the coefficient relationship between the current and the magnetic field changes:

[0084]

[0085] in, is the short-circuit turn current, is the coefficient of the magnetic field at each measuring point corresponding to the short-circuit turn current.

[0086] Further understanding shows that due to the complex collinearity between the high and low voltage winding currents, the identification results will be unstable. Ignoring the influence of the excitation current, the secondary current is converted to the primary side to reduce the number of parameters to be determined. Therefore, the analytical model of the current magnetic field coefficient that needs to be identified is:

[0087] in, is a magnetic leakage matrix constructed based on the magnetic leakage induction intensity of each magneto-optical measurement sensor, is a current matrix constructed based on the first current, is the coefficient identification matrix.

[0088] More specifically, the analytical model of the current magnetic field coefficient can be:

[0089]

[0090] It can be seen that the current magnetic field coefficient analytical model can include the leakage magnetic induction intensity matrix constructed by the leakage magnetic induction intensity of each magneto-optical measurement sensor. , the current matrix constructed by the first current , and the coefficient identification matrix , , n is the transformation ratio of the transformer.

[0091] Step S130 : Identify the coefficients of the current magnetic field coefficient analytical model using a least squares parameter algorithm to obtain coefficient identification results.

[0092] Understandably, traditional online monitoring methods often have a low detection rate when a transformer experiences a minor inter-turn short circuit due to the small change in the main circuit current. This method, however, relies on the analytically calculated change in the relationship between current and magnetic field during an inter-turn short circuit, and by identifying model coefficients, it can capture this subtle change. Furthermore, existing current and voltage monitoring methods are susceptible to interference from external load fluctuations in the distribution network, CT saturation, and excitation current, while vibration signal monitoring is susceptible to interference from environmental vibration and mechanical noise. Online identification of model parameters using a recursive least squares algorithm, based on an analytical model between current and leakage magnetic induction intensity, can effectively filter out signal offsets caused by these interfering factors, improve the stability of identifying true fault characteristics, and reduce misjudgments and missed reports.

[0093] Step S140: Calculate the identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state.

[0094] Among them, the identification coefficient change rate can represent the dynamic change degree of the coefficient in the analytical relationship model between the transformer winding current and the leakage magnetic field, reflecting the abnormal fluctuation of the winding structure integrity and the inter-turn insulation state.

[0095] Specifically, the identification coefficient change rate can be calculated using the following formula:

[0096]

[0097] in, is the rate of change of identification coefficient, is the coefficient identification result, is the leakage magnetic induction coefficient of the winding under normal conditions.

[0098] It is understandable that when the transformer is operating normally, there is a stable analytical relationship between the current in the transformer winding and the leakage magnetic field distribution, and the rate of change of the identification coefficient is in an extremely small range. When an early fault such as a slight inter-turn short circuit occurs in the winding, the short-circuit turn current will change the original electromagnetic coupling law, causing the analytical relationship between the current and the leakage magnetic field to shift, which manifests as a significant change in the model coefficient.

[0099] Step S150: When the transformer is not no-loaded and the iron core is not magnetized and saturated, if the identification coefficient change rate is greater than the change rate threshold, it is determined that there is an inter-turn short circuit in the transformer.

[0100] It is understandable that when the transformer is not no-loaded and the core is not magnetized and saturated, the winding fault is the only factor that affects the increase in the coefficient change rate r. At this time, the criterion for the existence of an inter-turn short circuit in the transformer can be:

[0101]

[0102] in, is the rate of change threshold, is the noise weighting coefficient, which can be 1.3. The accuracy error of the current transformer sampling the winding, its value depends on the sampling capability of the current transformer itself. is the measurement accuracy error of the magneto-optical measurement sensor, and its value depends on the measurement capability of the magneto-optical measurement sensor itself.

[0103] For example, taking the commonly used Class 1.0 measuring current transformer as an example, its accuracy error does not exceed 3% under various operating conditions such as different primary load currents and a slight increase in terminal current caused by a slight inter-turn short circuit.

[0104] The present embodiment provides a method for monitoring early-stage inter-turn short circuits in transformers. Magneto-optical measurement sensors are provided at both ends and in the middle of the winding within the core. The leakage magnetic induction intensity of each sensor and the first current flowing through the high-voltage winding are respectively obtained. A current magnetic field coefficient analytical model is constructed. The coefficients of the current magnetic field coefficient analytical model are identified using a least squares parameter algorithm to obtain coefficient identification results. Based on the coefficient identification results and the leakage magnetic induction coefficient of the winding under normal conditions, the identification coefficient change rate is calculated. When the transformer is not no-loaded and the core is not magnetized and saturated, if the identification coefficient change rate is greater than a change rate threshold, it is determined that the transformer has an inter-turn short circuit. Thus, the current magnetic field coefficient analytical model reflects the analytical calculation relationship between current and leakage magnetic induction intensity, and can perform real-time quantitative analysis of the local magnetic field disturbance caused by a slight inter-turn short circuit in the transformer winding. By combining the analytical model with a recursive least squares algorithm, signal offsets caused by load fluctuations and grid disturbances can be effectively filtered out, thereby improving the stability and accuracy of early-stage inter-turn short circuit identification in transformers.

[0105] In some embodiments of the present application, the process of obtaining the coefficient identification result by identifying the coefficients of the current magnetic field coefficient analytical model using the least squares parameter algorithm in step S130 is introduced. The process may include:

[0106] Step S210: Acquire multiple groups of sampling data, and select one group of sampling data as the target group of sampling data.

[0107] Specifically, each set of sampling data may include the leakage magnetic induction intensity measured by each magneto-optical measurement sensor and the sampling current flowing through the high-voltage winding.

[0108] Among them, the current of the sampling data can be expressed as , k represents the current k-th sampling data, and the leakage magnetic output of the sampling data can be expressed as , each element in the matrix represents the leakage magnetic induction intensity at the x1, x2, and y3 measuring points of the k-th sampling data.

[0109] Step S220 , initializing the forgetting factor, magnetic flux leakage prediction error, coefficient identification matrix and covariance matrix.

[0110] Among them, the initial value of the coefficient identification matrix can be , is a sufficiently small constant, specifically The initial value of the covariance matrix can be , is a sufficiently large constant, is the identity matrix.

[0111] Step S230: Use the recursive least squares algorithm to solve the model coefficient calculation formula, calculate the gain matrix and leakage magnetic field prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, update the leakage magnetic field prediction error matrix, coefficient identification matrix and covariance matrix, and obtain a new leakage magnetic field prediction error matrix, a new coefficient identification matrix and a new covariance matrix.

[0112] Specifically, the calculation formula for solving the model coefficients using the recursive least squares algorithm can be:

[0113]

[0114] Among them, here represents the target group sampling data, is the leakage magnetic field prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the magnetic flux leakage matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the current matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the previous target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the gain matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the covariance matrix under the previous target group sampling data, is the covariance matrix under the target group sampling data, is the identity matrix, is the forgetting factor under the target group sampling data, is the matrix transpose symbol.

[0115] Step S240: Adaptively update the forgetting factor using a forgetting factor adaptive update function to obtain a new forgetting factor.

[0116] Specifically, the forgetting factor adaptive update function can be:

[0117]

[0118] in, is the upper limit of the forgetting factor, which can be 0.999. is the lower limit of the forgetting factor, is the initial magnetic flux leakage prediction error, is the rounding function, is the magnetic flux leakage prediction error matrix The total error of the leakage magnetic output prediction results of each element.

[0119] Step S250 , determine whether there is any sample data that has not been used for recursive calculation. If so, execute step S260 ; if not, execute step S270 .

[0120] Step S260 : Select one of the groups of sample data that have not been used for recursive calculation as a new target group of sample data, and return to step S230 .

[0121] Step S270: Determine the coefficient identification result of the current magnetic field coefficient analytical model according to the coefficient identification matrix.

[0122] The transformer early fault monitoring method provided in this embodiment, through the combination of an analytical model and a recursive least squares algorithm, effectively filters out signal offsets caused by load fluctuations and grid disturbances, improving the robustness of identifying true fault characteristics. By continuously tracking and evaluating model coefficients, it avoids misjudgments and missed detections caused by single-value anomalies, significantly enhancing the system's robustness in practical applications.

[0123] Considering the no-load condition of the transformer, the identification result is only the coefficient corresponding to the high-voltage winding. Based on this, the early fault monitoring method of the transformer provided by this application also includes a solution for the response action to be performed when the transformer is no-loaded, specifically:

[0124] When the transformer is unloaded or the first condition is met, the transformer is locked.

[0125] Among them, the first condition is:

[0126]

[0127] in, is the high voltage side current of the transformer, is the rated current of the transformer.

[0128] Understandably, when the transformer is in the no-load state, the low-voltage winding current is extremely low. At this point, the coefficients in the analytical relationship model between current and leakage magnetic field primarily reflect the characteristics of the high-voltage winding and fail to fully capture the electromagnetic coupling patterns during normal operation. If the transformer is not locked out at this point, abnormal coefficient fluctuations (not caused by fault factors) may lead to a misinterpretation of a fault. By locking out the transformer, interference from no-load conditions on model coefficient identification can be eliminated, ensuring that monitoring is initiated only when the transformer is operating normally and under load, thereby improving the reliability of fault diagnosis.

[0129] Considering the situation of magnetic saturation of the transformer core, the change of the core magnetic permeability causes nonlinear changes in the magnetic field distribution. Based on this, the early fault monitoring method of the transformer provided by the present application also includes a solution for the response action to be performed when the core is magnetically saturated, specifically:

[0130] When the core is magnetized to saturation or the transformer meets the second condition, the transformer is locked.

[0131] Among them, the second condition is the percentage of the second harmonic content of the transformer differential current The differential current may represent the processed vector difference of the secondary currents of the current transformers on each side of the transformer.

[0132] It is understandable that when the transformer core is saturated, the core's magnetic permeability changes significantly, resulting in a nonlinear magnetic field distribution around the winding. This nonlinearity disrupts the analytical relationship model between the current and the leakage magnetic field during normal operation, causing the coefficients identified by the recursive least squares algorithm to fluctuate abnormally (not due to faults such as inter-turn short circuits). If the transformer is not locked out, these coefficient changes caused by non-fault factors may be mistaken for faults, triggering false alarms. By locking out the transformer, monitoring can be suspended under core saturation conditions, ensuring that monitoring is only activated under normal operating conditions when the magnetic field distribution conforms to the linear analytical model, further improving the reliability and accuracy of fault diagnosis.

[0133] The following describes an apparatus for implementing early turn-to-turn short-circuit monitoring of a transformer provided in an embodiment of the present application. The apparatus for implementing early turn-to-turn short-circuit monitoring of a transformer described below and the method for implementing early turn-to-turn short-circuit monitoring of a transformer described above can refer to each other.

[0134] See also Figure 4 , Figure 4 This is a schematic diagram of the structure of a device for realizing early turn-to-turn short-circuit monitoring of a transformer disclosed in an embodiment of the present application.

[0135] like Figure 4 As shown, the device may include:

[0136] A measurement information acquisition unit 11 is used to respectively acquire the leakage magnetic induction intensity of each of the magneto-optical measurement sensors and the first current flowing through the high-voltage winding;

[0137] a current magnetic field coefficient analytical model construction unit 12, configured to construct a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors;

[0138] The system identification unit 13 is used to identify the coefficients of the current magnetic field coefficient analytical model by a least squares parameter algorithm to obtain a coefficient identification result;

[0139] an identification coefficient change rate calculation unit 14, configured to calculate an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state;

[0140] The turn-to-turn short circuit judgment unit 15 is configured to determine that a turn-to-turn short circuit exists in the transformer if the identification coefficient change rate is greater than a change rate threshold when the transformer is not no-loaded and the iron core is not magnetized and saturated.

[0141] Optionally, the system identification unit includes:

[0142] a target group sampling data selection unit, configured to obtain multiple groups of sampling data and select one of the groups of sampling data as the target group sampling data;

[0143] Parameter initialization unit, used to initialize the forgetting factor, magnetic flux leakage prediction error, coefficient identification matrix and covariance matrix;

[0144] a matrix calculation and updating unit, configured to solve a model coefficient calculation formula using a recursive least squares algorithm, calculate a gain matrix and a magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, update the magnetic flux leakage prediction error matrix, the coefficient identification matrix, and the covariance matrix, and obtain a new magnetic flux leakage prediction error matrix, a new coefficient identification matrix, and a new covariance matrix;

[0145] a forgetting factor updating unit, configured to adaptively update the forgetting factor by using a forgetting factor adaptive updating function to obtain a new forgetting factor;

[0146] an adopted data existence judgment unit, configured to judge whether the sampled data has not been used for recursive calculation, and if so, to execute a new target adopted data determination unit; if not, to execute a coefficient identification result determination unit;

[0147] The new target adopted data determination unit is used to select one group of sample data from each group of sample data that has not been used for recursive calculation as a new target group of sample data, and return the selected group to execute the matrix calculation update unit;

[0148] The coefficient identification result determination unit is used to determine the coefficient identification result of the current magnetic field coefficient analytical model according to the coefficient identification matrix.

[0149] Optionally, the current magnetic field coefficient analytical model is:

[0150]

[0151] in, is a magnetic leakage matrix constructed based on the magnetic leakage induction intensity of each magneto-optical measurement sensor, is a current matrix constructed based on the first current, is the coefficient identification matrix.

[0152] Optionally, the recursive least squares algorithm is used to calculate the model coefficients:

[0153]

[0154] Among them, represents the target group sampling data, is the magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the magnetic leakage matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the current matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the previous target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the gain matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the covariance matrix under the previous target group sampling data, is the covariance matrix under the target group sampling data, is the identity matrix, is the forgetting factor under the target group sampling data, is the matrix transpose symbol;

[0155] The forgetting factor adaptive update function is:

[0156]

[0157] in, is the upper limit of the forgetting factor, is the lower limit of the forgetting factor, is the initial magnetic flux leakage prediction error, is the rounding function, is the magnetic flux leakage prediction error matrix The total error of the leakage magnetic output prediction results of each element.

[0158] Optionally, the device further includes:

[0159] The first locking unit is configured to lock the transformer when the transformer is unloaded or a first condition is satisfied, wherein the first condition is:

[0160]

[0161] in, is the high voltage side current of the transformer, is the rated current of the transformer.

[0162] Optionally, the device further includes:

[0163] a second locking unit, configured to lock the transformer when the core is magnetized to saturation or the transformer satisfies a second condition, wherein the second condition is that a percentage of a second harmonic content of a differential current of the transformer is greater than a percentage threshold, and the differential current is a processed vector difference of secondary currents of current transformers on each side of the transformer.

[0164] Optionally, the change rate threshold is:

[0165]

[0166] in, is the change rate threshold, is the noise weighting coefficient, is the accuracy error of the current transformer sampling the winding, is the measurement accuracy error of the magneto-optical measurement sensor.

[0167] The device for early turn-to-turn short circuit monitoring of a transformer provided in the embodiment of the present application can be applied to equipment for early turn-to-turn short circuit monitoring of a transformer, such as a monitoring module. Optionally, Figure 5 The hardware structure diagram of the equipment for early turn-to-turn short circuit monitoring of transformer is shown in FIG. Figure 5 The hardware structure of the device for early turn-to-turn short circuit monitoring of a transformer may include: at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4;

[0168] In the embodiment of the present application, the number of the processor 1, the communication interface 2, the memory 3, and the communication bus 4 is at least one, and the processor 1, the communication interface 2, and the memory 3 communicate with each other through the communication bus 4;

[0169] The processor 1 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention;

[0170] The memory 3 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory;

[0171] The memory stores a program, and the processor can call the program stored in the memory, wherein the program is used to:

[0172] respectively obtaining the leakage magnetic induction intensity of each magneto-optical measurement sensor and the first current flowing through the high-voltage winding;

[0173] constructing a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors;

[0174] Identifying the coefficients of the current magnetic field coefficient analytical model by a least squares parameter algorithm to obtain coefficient identification results;

[0175] Calculating an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state;

[0176] When the transformer is not no-loaded and the iron core is not magnetized and saturated, if the identification coefficient change rate is greater than a change rate threshold, it is determined that an inter-turn short circuit exists in the transformer.

[0177] Optionally, the detailed functions and extended functions of the program may refer to the above description.

[0178] An embodiment of the present application further provides a storage medium, which may store a program suitable for execution by a processor, wherein the program is used to:

[0179] respectively obtaining the leakage magnetic induction intensity of each magneto-optical measurement sensor and the first current flowing through the high-voltage winding;

[0180] constructing a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors;

[0181] Identifying the coefficients of the current magnetic field coefficient analytical model by a least squares parameter algorithm to obtain coefficient identification results;

[0182] Calculating an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state;

[0183] When the transformer is not no-loaded and the iron core is not magnetized and saturated, if the identification coefficient change rate is greater than a change rate threshold, it is determined that an inter-turn short circuit exists in the transformer.

[0184] Optionally, the detailed functions and extended functions of the program may refer to the above description.

[0185] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or device comprising the element.

[0186] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referenced to each other.

[0187] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one 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 present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for monitoring early turn-to-turn short circuit of a transformer, characterized in that: Applied to a monitoring module, the monitoring module is connected to a plurality of magneto-optical measurement sensors, wherein each magneto-optical measurement sensor is respectively installed at both ends of a winding inside the iron core of a transformer and at the axial center of the winding inside the iron core, wherein the winding includes a high-voltage winding; The method includes: respectively obtaining the leakage magnetic induction intensity of each magneto-optical measurement sensor and the first current flowing through the high-voltage winding; constructing a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors; Identifying the coefficients of the current magnetic field coefficient analytical model by a least squares parameter algorithm to obtain coefficient identification results; Calculating an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state; When the transformer is not no-loaded and the iron core is not magnetized and saturated, if the identification coefficient change rate is greater than a change rate threshold, it is determined that an inter-turn short circuit exists in the transformer.

2. The method according to claim 1, characterized in that The coefficients of the current magnetic field coefficient analytical model are identified by the least squares parameter algorithm to obtain coefficient identification results, including: Acquire multiple groups of sampling data, and select one group of the sampling data as the target group of sampling data; Initialize the forgetting factor, magnetic flux leakage prediction error, coefficient identification matrix and covariance matrix; Using a recursive least squares algorithm to solve a model coefficient calculation formula, calculating a gain matrix and a magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, updating the magnetic flux leakage prediction error matrix, the coefficient identification matrix, and the covariance matrix to obtain a new magnetic flux leakage prediction error matrix, a new coefficient identification matrix, and a new covariance matrix; Adaptively updating the forgetting factor through a forgetting factor adaptive update function to obtain a new forgetting factor; Determining whether there is any sample data that has not been used for recursive calculation; If so, one of the groups of sampled data that have not been used for recursive calculation is selected as a new target group of sampled data, and the process returns to the step of using the recursive least squares algorithm to solve the model coefficient calculation formula, calculate the gain matrix and the magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group of sampled data, and update the magnetic flux leakage prediction error matrix, the coefficient identification matrix, and the covariance matrix to obtain a new magnetic flux leakage prediction error matrix, a new coefficient identification matrix, and a new covariance matrix. If not, the coefficient identification result of the current magnetic field coefficient analytical model is determined according to the coefficient identification matrix.

3. The method according to claim 2, characterized in that The current magnetic field coefficient analytical model is: in, is a magnetic leakage matrix constructed based on the magnetic leakage induction intensity of each magneto-optical measurement sensor, is a current matrix constructed based on the first current, is the coefficient identification matrix.

4. The method according to claim 3, characterized in that The recursive least squares algorithm is used to calculate the model coefficients: Among them, represents the target group sampling data, is the magnetic flux leakage prediction error matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the magnetic leakage matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the current matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the previous target group sampling data, is the coefficient identification matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the gain matrix of the current magnetic field coefficient analytical model under the target group sampling data, is the covariance matrix under the previous target group sampling data, is the covariance matrix under the target group sampling data, is the identity matrix, is the forgetting factor under the target group sampling data, is the matrix transpose symbol; The forgetting factor adaptive update function is: in, is the upper limit of the forgetting factor, is the lower limit of the forgetting factor, is the initial magnetic flux leakage prediction error, is the rounding function, is the magnetic flux leakage prediction error matrix The total error of the leakage magnetic output prediction results of each element.

5. The method according to any one of claims 1 to 4, characterized in that Also includes: When the transformer is unloaded or satisfies a first condition, the transformer is locked, wherein the first condition is: in, is the high voltage side current of the transformer, is the rated current of the transformer.

6. The method according to any one of claims 1 to 4, characterized in that Also includes: When the core is magnetized and saturated or the transformer meets a second condition, the transformer is locked, wherein the second condition is that the percentage of the second harmonic content of the differential current of the transformer is greater than a percentage threshold, and the differential current is the processed vector difference of the secondary currents of the current transformers on each side of the transformer.

7. The method according to any one of claims 1 to 4, characterized in that The change rate threshold is: in, is the change rate threshold, is the noise weighting coefficient, is the accuracy error of the current transformer sampling the winding, is the measurement accuracy error of the magneto-optical measurement sensor.

8. A transformer early turn short circuit monitoring device, characterized in that: Applied to a monitoring module, the monitoring module is connected to a plurality of magneto-optical measurement sensors, wherein each magneto-optical measurement sensor is respectively installed at both ends of a winding inside the iron core of a transformer and at the axial center of the winding inside the iron core, wherein the winding includes a high-voltage winding; The device includes: a measurement information acquisition unit, configured to respectively acquire the leakage magnetic induction intensity of each of the magneto-optical measurement sensors and the first current flowing through the high-voltage winding; a current magnetic field coefficient analytical model construction unit, configured to construct a current magnetic field coefficient analytical model according to the first current and the leakage magnetic induction intensity of each of the magneto-optical measurement sensors; A system identification unit, configured to identify the coefficients of the current magnetic field coefficient analytical model by using a least squares parameter algorithm to obtain coefficient identification results; an identification coefficient change rate calculation unit, configured to calculate an identification coefficient change rate based on the coefficient identification result and the leakage magnetic induction coefficient of the winding in a normal state; The turn-to-turn short circuit judgment unit is configured to determine that a turn-to-turn short circuit exists in the transformer if the identification coefficient change rate is greater than a change rate threshold when the transformer is not no-loaded and the iron core is not magnetized and saturated.

9. A transformer early turn short circuit monitoring device, characterized in that: including memory and processor; The memory is used to store programs; The processor is used to execute the program to implement each step of the early turn-to-turn short circuit monitoring method for a transformer as described in any one of claims 1 to 7.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, each step of the method for early turn-to-turn short circuit monitoring of a transformer as claimed in any one of claims 1 to 7 is implemented.

Images