Transformer winding fault and phase identification method based on power loss change rate

Abstract

This invention belongs to the field of transformer fault identification technology, specifically relating to a method for identifying transformer winding faults and phases based on the power loss change rate. The steps include: collecting basic transformer information and obtaining nameplate parameters; establishing a theoretical power loss model during normal transformer operation; calculating the actual total power loss of the transformer windings and the actual power loss of each phase winding, and calculating the total power loss change rate and the power loss change rate of each phase respectively; determining the winding insulation fault state based on the maximum value of the total power loss change rate and preset loss and temperature thresholds; when a winding insulation fault is determined, determining the faulty phase based on the magnitude relationship between the maximum values ​​of the power loss change rates of each phase and the phase-to-phase differences; and generating early warning, tripping, or blocking control commands based on the winding insulation fault state and the faulty phase. This invention can achieve early warning and accurate phase identification of transformer winding insulation faults.

Description

Technical Field

[0001] This invention belongs to the field of transformer fault identification technology, specifically relating to a method for identifying transformer winding faults and phases based on the power loss change rate. Background Technology

[0002] Transformers typically operate long-term in power grids and distribution systems. They are susceptible to problems such as inter-turn insulation degradation, localized overheating, and inter-turn short circuits due to factors including low-voltage side short-circuit impacts, load rate changes, operating temperature rise, winding electrodynamics, insulation material aging, and manufacturing process differences. Initially, inter-turn insulation degradation may not immediately manifest as a significant change in main circuit current. However, once a short-circuit loop forms on the faulty turn, the short-circuit loop current increases rapidly, leading to a significant rise in localized power loss and heat generation. In severe cases, this can cause winding burnout, insulation breakdown, over-limit protection tripping, or even transformer shutdown.

[0003] like Figure 1 As shown, when inter-turn breakdown occurs between adjacent turns (such as the W8 turn) due to insulation deterioration, a short-circuit loop is formed inside the faulty turn. The short-circuit loop current increases sharply in the local closed loop, resulting in a significant increase in local additional losses and heat generation at the fault point. At the same time, since the short-circuit loop current is confined inside the faulty turn, the main circuit current changes very little. Traditional methods based on the main circuit current or differential current are difficult to identify the fault in the early stage of the fault.

[0004] Existing methods for monitoring the inter-turn insulation condition and identifying faults in transformer windings mostly rely on main circuit current, differential current, overcurrent protection, fuse protection, or offline tests during power outages for judgment. That is, alarms or protection actions are only issued when the current, differential current, or test characteristic quantity reaches a preset threshold. This type of method has at least the following problems: 1) Traditional methods often rely on main circuit current, differential current, or single electrical quantities for triggering, failing to construct a power loss change rate diagnostic chain based on easily obtainable basic information such as voltage, current, and temperature. This is especially problematic when the main circuit current change is not significant in the early stages of an inter-turn short circuit, resulting in a noticeable lag in protection. 2) Insufficient utilization of the deviation between the actual total power loss of the winding and the total theoretical winding loss makes it difficult to promptly reflect the overall loss anomalies and insulation degradation trends caused by the faulty turn short circuit loop. 3) Insufficient utilization of the phase-to-phase differences in the power loss change rate of each phase. Usually, it can only determine whether an anomaly exists, but it is difficult to accurately locate the phase where inter-turn insulation degradation or inter-turn short circuit occurs. 4) Although characteristic quantities such as short-circuit loop current, local magnetic field strength, and winding stress are sensitive to inter-turn short circuits, sensor placement is difficult (or even impossible), alters the winding electric field distribution, endangers insulation, and is costly, making widespread application in the field difficult.

[0005] Therefore, there is an urgent need to propose a transformer winding fault identification method to solve the problems of existing methods such as delayed identification in the early stage of inter-turn insulation deterioration, difficulty in locating faulty phases, and insufficient threshold adaptability. Summary of the Invention

[0006] In view of the shortcomings of the prior art, the purpose of this invention is to provide a method for identifying transformer winding faults and phases based on the power loss change rate. Based on the power loss change rate, early warning and accurate phase identification of transformer winding insulation faults can be achieved.

[0007] To achieve the above objectives, this invention provides a method for identifying transformer winding faults and phases based on the rate of change of power loss, comprising the following steps: S1. Collect the voltage, current and temperature signals of the transformer as basic information, and obtain the transformer nameplate parameters including no-load loss and load loss. S2. During normal operation of the transformer, based on the transformer nameplate parameters and real-time collected basic information, calculate the fault-free calibration data of the transformer under the current load rate, and establish a theoretical power loss model characterizing the normal power loss of each phase winding and the entire transformer under different load rates and temperature conditions. S3. Based on the real-time collected basic information, calculate the actual total power loss of the transformer windings and the actual power loss of each phase winding, and calculate the total power loss change rate and the power loss change rate of each phase respectively. S4. Extract the maximum value of the total power loss change rate and the maximum value of the power loss change rate of each phase within the preset sampling window. Based on the maximum value of the total power loss change rate and the preset loss threshold and temperature threshold, determine the winding insulation fault status. S5. When it is determined that there is a winding insulation fault, the faulty phase is determined according to the relationship between the maximum values ​​of the power loss change rate of each phase and the phase-to-phase differences. S6. Generate early warning, trip, or lockout control commands based on the winding insulation fault status and fault phase, and perform online correction on the theoretical power loss model and the threshold parameters involved in determining the winding insulation fault status and fault phase.

[0008] As a preferred embodiment of the present invention, in S1, the voltage signal and current signal are acquired by current and voltage sensors on the high-voltage side and low-voltage side; the temperature signal is acquired by the transformer body temperature detection device. For oil-immersed transformers, the temperature signal is the oil temperature, and for dry-type transformers, the temperature signal is the winding temperature. The voltage signal and current signal are used to obtain the input and output active power, load rate and current characteristics of each phase, and the temperature signal is used for winding resistance temperature correction and temperature threshold setting. The acquired voltage, current, and temperature signals undergo preprocessing, including time synchronization, outlier removal, missing value completion, and smoothing filtering.

[0009] As a preferred embodiment of the present invention, in S2, the theoretical power loss model includes a load rate calculation model, a winding resistance temperature correction model, a theoretical winding loss model for each phase, and a total theoretical winding loss model. The load factor calculation model includes formulas for calculating the total load factor and the phase load factor: ; ; In the formula, This represents the total load factor of the transformer at time k; Let k be the transformer operating capacity at time k. This refers to the rated capacity of the transformer. Let i represent the load rate of phase i at time k, where i = A, B, C, corresponding to phases A, B, and C respectively; This represents the current in the i-th phase winding at time k; Rated current; The winding resistance temperature correction model is expressed as: ; In the formula, Indicates temperature The winding resistance of the i-th phase; Indicates reference temperature The winding resistance of the i-th phase; K is the temperature coefficient of the winding material; The theoretical winding loss model for each phase is expressed as follows: ; In the formula, This represents the theoretical winding loss of the i-th phase at time k; This represents the rated current of the i-th phase; This represents the additional loss correction term for the i-th phase. This represents the load rate of the i-th phase; The overall theoretical winding loss model is expressed as: ; In the formula, This represents the total theoretical winding loss of the transformer at time k; , , These represent the theoretical winding losses of phases A, B, and C at time k, respectively.

[0010] As a preferred embodiment of the present invention, in S3, the actual total power loss of the winding is obtained by subtracting the core loss from the difference between the active power input on the high-voltage side and the active power output on the low-voltage side, and is expressed as follows: ; In the formula, This represents the actual total power loss of the transformer windings at time k; This represents the active power input to the high-voltage side at time k; This represents the low-voltage side output active power at time k; This represents the core loss at time k; The actual power loss of each phase winding is calculated based on the winding resistance after temperature correction for each phase winding current, and is expressed as: ; In the formula, This represents the actual power loss of the i-th phase winding at time k; This represents the local additional loss of the i-th phase caused by inter-turn insulation degradation or short-circuit circulating current at time k. This represents the additional loss correction term for the i-th phase at time k. This represents the temperature at time k; The rate of change of total power loss is expressed as: ; In the formula, This represents the rate of change of the total power loss of the transformer at time k; The rate of change of power loss in each phase is expressed as: ; In the formula, This represents the rate of change of power loss of the i-th phase winding at time k.

[0011] As a preferred embodiment of the present invention, in step S4, the maximum value of the total power loss change rate is extracted within a preset sampling window, and is expressed as: ; In the formula, This represents the maximum rate of change of total power loss within the preset sampling window at time k; Indicates the preset sampling window length; express The rate of change of the total power loss of the transformer at any given moment; The maximum value of the power loss change rate of each phase is extracted within the preset sampling window and expressed as: ; In the formula, This represents the maximum rate of change of power loss of the i-th phase winding within the preset sampling window at time k. express The rate of change of power loss of the i-th phase winding at time i.

[0012] In a preferred embodiment of the present invention, in step S4, before determining the winding insulation fault state, the voltage change rate and the main circuit current change rate are first calculated, expressed as follows: ; ; In the formula, Indicates the rate of change of voltage; , Let these represent the voltages of the i-th phase at times k and k-1, respectively. Indicates the rate of change of the main circuit current; , These represent the main circuit current of the i-th phase at times k and k-1, respectively. When the voltage, current, and temperature signals are all in a valid sampling state, and the voltage change rate is less than the voltage blocking threshold and the main circuit current change rate is less than the current blocking threshold, winding insulation fault diagnosis is allowed. If the above conditions are not met, it is determined that there is an external disturbance, temperature abnormality, or sampling abnormality in the current state, and the output of winding insulation fault diagnosis results is suspended.

[0013] As a preferred embodiment of the present invention, in S4, the method for determining the winding insulation fault state is based on the current total load rate of the transformer. Set the first loss threshold Second loss threshold First loss threshold The second loss threshold corresponds to the critical degradation state of the inter-turn insulation of the winding. The corresponding winding inter-turn insulation is in a critical collapse state, and Greater than ; Set the first temperature threshold according to the transformer type. Second temperature threshold ,and Greater than ; When the maximum rate of change of total power loss is less than the first loss threshold When the temperature signal is greater than or equal to the first temperature threshold, the winding inter-turn insulation is determined to be in a normal state. If so, an abnormal temperature alarm signal will be output or the process will begin. When the maximum rate of change of total power loss is greater than or equal to the first loss threshold And less than the second loss threshold When the winding inter-turn insulation is in a deteriorated state, a warning signal is output; if the temperature signal is greater than or equal to the second temperature threshold at this time... If so, the warning level will be raised or the review period will be shortened; When the maximum rate of change of total power loss is greater than or equal to the second loss threshold When the winding inter-turn insulation is in a state of collapse or near collapse, a trip protection signal is output.

[0014] As a preferred embodiment of the present invention, in S5, the phase with the largest value among the maximum values ​​of the power loss change rate of each phase in the three phases is selected as the candidate fault phase. When the total power loss change rate has met the winding insulation fault judgment condition, and the maximum power loss change rate of each phase of the candidate fault phase is greater than the phase identification threshold under the current load rate, if the proportion of the power loss change rate of the candidate fault phase is greater than the preset proportion threshold or the difference in power loss change rate between the candidate fault phase and other phases is greater than the preset phase difference threshold, the candidate fault phase is determined to be the fault phase. If two or three phases simultaneously meet the above fault phase discrimination criteria, then a multi-phase winding inter-turn insulation abnormality signal will be output.

[0015] As a preferred embodiment of the present invention, in S6, when the winding insulation is in a normal state, a normal operation signal is output; when the winding insulation is in a deteriorated state, a warning signal is output; when the winding insulation is in a collapse or near collapse state, a trip protection signal is output; when there is external disturbance, abnormal temperature, abnormal sampling, or the fault phase identification result does not meet the criteria, a blocking signal is output or a review process is entered. Online correction involves correcting the winding material temperature coefficient and additional loss correction term based on the deviation between the actual total power loss and the total theoretical winding loss after the control cycle ends, when the winding insulation is in a normal state; and correcting the first loss threshold, second loss threshold, phase identification threshold, preset proportion threshold, and preset phase difference threshold based on the statistical distribution characteristics of the total power loss change rate under normal operating conditions, and setting parameter value range limits, correction step size limits, and parameter anti-drift processing mechanisms.

[0016] In a preferred embodiment of the present invention, in S6, the deviation between the actual total power loss of the winding and the total theoretical winding loss at time k is defined as... : ; In the formula, This represents the actual total power loss of the transformer windings at time k; This represents the total theoretical winding loss of the transformer at time k; The winding material temperature coefficient and the additional loss correction term are used to construct the parameter vector to be corrected. The correction method is expressed as follows: ; In the formula, , These represent the parameter vectors to be corrected at time k+1 and time k, respectively; Indicates the correction step size; , These represent the minimum and maximum allowable values ​​of the parameter to be corrected, respectively. This represents a truncation function used to restrict the parameter correction result to within an allowed range.

[0017] The beneficial effects of this invention are: This invention uses the rate of change of total power loss as a sensitive state quantity for diagnosing inter-turn insulation faults in transformer windings. By establishing a theoretical power loss model and comparing the actual total power loss of the windings calculated in real time with the total theoretical winding loss, it can reflect the overall additional loss change caused by the short-circuit circulating current of the faulty turn. Compared with traditional methods that rely solely on main circuit current, differential current, or overcurrent protection, this invention can identify anomalies in the early stages of inter-turn insulation degradation and when the change in main circuit current is not significant, thus improving the early detection capability and protection judgment sensitivity of inter-turn short-circuit faults.

[0018] This invention uses voltage, current, and temperature signals as basic information. For oil-immersed transformers, oil temperature is used as the temperature signal, while for dry-type transformers, winding temperature is used. It combines voltage change rate and main circuit current change rate for blocking and auxiliary judgment, effectively distinguishing between power loss anomalies caused by external disturbances, load fluctuations, temperature abnormalities, and winding inter-turn insulation degradation, reducing the probability of misjudgments and malfunctions. Furthermore, this invention determines the faulty phase based on the maximum power loss change rate, its proportion, and phase-to-phase differences. It implements graded early warning and trip protection based on insulation status using a first and second loss threshold. It supports online correction based on the deviation between measured and theoretical data, thereby improving the method's adaptability to different capacities, structures, and long-term operating parameter drift. Attached Figure Description

[0019] Figure 1 This is a physical schematic diagram of a single-turn short-circuit fault in the low-voltage winding of a transformer. Figure 2 This is a flowchart illustrating the principle of this invention. Detailed Implementation

[0020] The embodiments of the present invention will be further described below with reference to the accompanying drawings: Example 1: As Figure 2 As shown, the method for identifying transformer winding faults and phases based on the power loss change rate includes the following steps: S1. Collect the voltage, current and temperature signals of the transformer as basic information, and obtain the transformer nameplate parameters including no-load loss and load loss. S2. During normal operation of the transformer, based on the transformer nameplate parameters and real-time collected basic information, calculate the fault-free calibration data of the transformer under the current load rate, and establish a theoretical power loss model characterizing the normal power loss of each phase winding and the entire transformer under different load rates and temperature conditions. S3. Based on the real-time collected basic information, calculate the actual total power loss of the transformer windings and the actual power loss of each phase winding, and calculate the total power loss change rate and the power loss change rate of each phase respectively. S4. Extract the maximum value of the total power loss change rate and the maximum value of the power loss change rate of each phase within the preset sampling window. Based on the maximum value of the total power loss change rate and the preset loss threshold and temperature threshold, determine the winding insulation fault status. S5. When it is determined that there is a winding insulation fault, the faulty phase is determined according to the relationship between the maximum values ​​of the power loss change rate of each phase and the phase-to-phase differences. S6. Generate early warning, trip, or lockout control commands based on the winding insulation fault status and fault phase, and perform online correction on the theoretical power loss model and the threshold parameters involved in determining the winding insulation fault status and fault phase.

[0021] The total power loss change rate reflects the overall deviation of the entire transformer from its theoretical loss and serves as a basis for fault diagnosis. The power loss change rate of each phase reflects the local deviation of each phase winding from its corresponding theoretical loss and is used for fault phase identification.

[0022] In S1, voltage and current signals are acquired by current and voltage sensors on the high-voltage and low-voltage sides; temperature signals are acquired by the transformer body temperature detection device. For oil-immersed transformers, the temperature signal is the oil temperature, and for dry-type transformers, the temperature signal is the winding temperature. Voltage and current signals are used to obtain input and output active power, load rate, and current characteristics of each phase, while temperature signals are used for winding resistance temperature correction and temperature threshold setting. The acquired voltage, current, and temperature signals undergo preprocessing, including time synchronization, outlier removal, missing value completion, and smoothing filtering.

[0023] In S2, the theoretical power loss model includes the load rate calculation model, the winding resistance temperature correction model, the theoretical winding loss model for each phase, and the total theoretical winding loss model. The load factor calculation model includes formulas for calculating the total load factor and the phase load factor: ; ; In the formula, This represents the total load factor of the transformer at time k; Let k be the transformer operating capacity at time k. This refers to the rated capacity of the transformer. Let i represent the load rate of phase i at time k, where i = A, B, C, corresponding to phases A, B, and C respectively; This represents the current in the i-th phase winding at time k; Rated current; The winding resistance temperature correction model is expressed as: ; In the formula, Indicates temperature The winding resistance of the i-th phase; Indicates reference temperature The winding resistance of the i-th phase; K is the temperature coefficient of the winding material; when the winding material is copper, K can be initially set to 235, and when the winding material is aluminum, K can be initially set to 225.

[0024] The theoretical winding loss model for each phase is expressed as follows: ; In the formula, This represents the theoretical winding loss of the i-th phase at time k; This represents the rated current of the i-th phase; Indicates and , The relevant i-th phase additional loss correction term, This represents the load rate of the i-th phase; It can be determined based on one or more of the following: factory test data, field calibration data, simulation data, or long-term trouble-free operation data. Similarly.

[0025] The overall theoretical winding loss model is expressed as: ; In the formula, This represents the total theoretical winding loss of the transformer at time k; , , These represent the theoretical winding losses of phases A, B, and C at time k, respectively.

[0026] In S3, the actual total power loss of the winding is obtained by subtracting the core loss from the difference between the active power input on the high-voltage side and the active power output on the low-voltage side, and is expressed as: ; In the formula, This represents the actual total power loss of the transformer windings at time k; This represents the active power input to the high-voltage side at time k; This represents the low-voltage side output active power at time k; This represents the core loss at time k, which can be taken as the no-load loss marked on the transformer nameplate. The actual power loss of each phase winding is calculated based on the winding resistance after temperature correction for each phase winding current, and is expressed as: ; In the formula, This represents the actual power loss of the i-th phase winding at time k; This represents the local additional loss of the i-th phase caused by inter-turn insulation degradation or short-circuit circulating current at time k. This represents the additional loss correction term for the i-th phase at time k. This represents the temperature at time k; The actual total power loss of the winding can be fused and corrected with the actual power loss of each phase winding using a weighted average, so as to improve the accuracy of the actual power loss calculation results.

[0027] The rate of change of total power loss is expressed as: ; In the formula, This represents the rate of change of the total power loss of the transformer at time k; The rate of change of power loss in each phase is expressed as: ; In the formula, This represents the rate of change of power loss of the i-th phase winding at time k.

[0028] In S4, the maximum value of the total power loss change rate is extracted within a preset sampling window (preferably 20ms~100ms), and is expressed as: ; In the formula, This represents the maximum rate of change of total power loss within the preset sampling window at time k; Indicates the preset sampling window length; express The rate of change of the total power loss of the transformer at any given moment; The maximum value of the power loss change rate of each phase is extracted within the preset sampling window and expressed as: ; In the formula, This represents the maximum rate of change of power loss of the i-th phase winding within the preset sampling window at time k. express The rate of change of power loss of the i-th phase winding at time i.

[0029] Before determining the winding insulation fault state, first calculate the voltage change rate and the main circuit current change rate, expressed as: ; ; In the formula, Indicates the rate of change of voltage; , Let these represent the voltages of the i-th phase at times k and k-1, respectively. Indicates the rate of change of the main circuit current; , These represent the main circuit current of the i-th phase at times k and k-1, respectively. and The current signals are collected at different measurement locations for the same electrical quantity. For a two-winding transformer, the values ​​are the same, and they are used for winding loss calculation and external disturbance identification, respectively.

[0030] Winding insulation fault diagnosis is permitted when voltage, current, and temperature signals are all in valid sampling condition, and the voltage change rate is less than the voltage blocking threshold and the main circuit current change rate is less than the current blocking threshold. If the above conditions are not met, it is determined that there is an external disturbance, temperature abnormality, or sampling abnormality in the current state, and the output of winding insulation fault diagnosis results is suspended. The voltage blocking threshold can be set to 2%~5%, and the current blocking threshold can be set to 10%~20%.

[0031] The method for determining the winding insulation fault status is based on the current transformer total load factor. Set the first loss threshold Second loss threshold First loss threshold The second loss threshold corresponds to the critical degradation state of the inter-turn insulation of the winding. The corresponding winding inter-turn insulation is in a critical collapse state, and Greater than , The preferred value is 5% to 15%. The preferred value is 30% to 50%; Set the first temperature threshold according to the transformer type. Second temperature threshold ,and Greater than For oil-immersed transformers, The preferred temperature is 85℃. The preferred temperature is 95℃; for dry-type transformers, The preferred temperature is 110℃. The preferred temperature is 130℃.

[0032] When the maximum rate of change of total power loss is less than the first loss threshold When the temperature signal is greater than or equal to the first temperature threshold, the winding inter-turn insulation is determined to be in a normal state. If so, an abnormal temperature alarm signal will be output or the process will begin. When the maximum rate of change of total power loss is greater than or equal to the first loss threshold And less than the second loss threshold When the winding inter-turn insulation is in a deteriorated state, a warning signal is output; if the temperature signal is greater than or equal to the second temperature threshold at this time... If so, the warning level will be raised or the review period will be shortened; When the maximum rate of change of total power loss is greater than or equal to the second loss threshold When the winding inter-turn insulation is in a state of collapse or near collapse, a trip protection signal is output.

[0033] The above fault diagnosis process does not directly trigger protection action based on the single-phase power loss change rate. Instead, it uses the maximum value of the total power loss change rate as the primary criterion for protection action, thereby avoiding false tripping caused by single-phase sampling noise, local temperature fluctuations, or short-term load imbalance. Once the total power loss change rate meets the fault diagnosis or protection conditions, the power loss change rate of each phase participates in phase identification to guide subsequent maintenance location and phase alarm.

[0034] In S5, the phase with the largest value among the maximum power loss change rates of each of the three phases is selected as the candidate fault phase. When the total power loss change rate meets the winding insulation fault judgment condition, and the maximum power loss change rate of each phase of the candidate fault phase is greater than the phase identification threshold under the current load rate, if the proportion of the power loss change rate of the candidate fault phase is greater than the preset proportion threshold or the difference in power loss change rate between the candidate fault phase and other phases is greater than the preset phase difference threshold, the candidate fault phase is determined to be the fault phase; where the proportion of the power loss change rate of the candidate fault phase refers to the ratio of the maximum power loss change rate of the candidate fault phase to the sum of the maximum power loss change rates of the three phases.

[0035] If two or three phases simultaneously meet the above fault phase discrimination criteria, then a multi-phase winding inter-turn insulation abnormality signal will be output.

[0036] The phase recognition threshold can be taken as follows: The percentage can be set to 50%~80%, the preset percentage threshold can be set to 50%~60%, and the preset difference threshold can be set to... 30% to 50%.

[0037] In S6, when the winding insulation is in a normal state, a normal operation signal is output; when the winding insulation is in a deteriorated state, a warning signal is output; when the winding insulation is in a collapse or near collapse state, a trip protection signal is output; when there is external disturbance, abnormal temperature, abnormal sampling, or the fault phase identification result does not meet the criteria, a blocking signal is output or a review process is initiated. Online correction involves correcting the winding material temperature coefficient and additional loss correction term based on the deviation between the actual total power loss and the total theoretical winding loss after the control cycle ends, when the winding insulation is in a normal state; and correcting the first loss threshold, second loss threshold, phase identification threshold, preset proportion threshold, and preset phase difference threshold based on the statistical distribution characteristics of the total power loss change rate under normal operating conditions, and setting parameter value range limits, correction step size limits, and parameter anti-drift processing mechanisms.

[0038] The control cycle refers to the time interval between the start of one complete online correction and the start of the next online correction, typically encompassing multiple sampling windows for fault diagnosis and data acquisition. Parameter value range limitation refers to setting physically reasonable upper and lower bounds for each parameter to be corrected, preventing the correction results from exceeding the range allowed by the actual characteristics of the transformer or engineering experience. Correction step size limitation refers to imposing an upper limit constraint on the increment of a single correction, avoiding drastic jumps in parameters due to measurement noise or transient disturbances. Parameter anti-drift processing mechanisms involve introducing forgetting factors, moving averages, or long-term trend monitoring mechanisms to suppress the slow shift of parameters in a single direction over time, ensuring that the corrected model parameters and threshold parameters always fluctuate around the true values ​​rather than continuously deviating.

[0039] The statistical distribution characteristics of the total power loss change rate under normal operating conditions are as follows: taking the most recent N control cycles under normal operating conditions... mean and standard deviation ,according to The principle is to adjust the threshold.

[0040] For the first loss threshold Second loss threshold The parameters include phase identification threshold, preset proportion threshold, and preset phase difference threshold. An example of correction is as follows: for a 1000KVA oil-immersed transformer, before correction... , The set values ​​are 10% and 40% respectively; the phase recognition threshold is set to... 60%, with a preset percentage threshold of 55%, and a preset difference threshold of [missing information]. 40%.

[0041] Data is continuously collected for N control cycles under normal operating conditions, and statistical analysis is performed. The distribution characteristic is the mean. Standard deviation ,according to In principle, the benchmark value is 2.0% + 3 × 0.7% = 4.1%.

[0042] Correction: Because 4.1% is lower than The preferred value range has a lower limit of 5%, therefore the actual corrected value is... ; Correction: The second loss threshold should maintain a multiple relationship of 3 to 4 times with the first loss threshold. Calculated at 4 times, 5% × 4 = 20%. Considering the correction step size limitation mechanism requires that the single correction magnitude should not be too large, and... The lower limit of the preferred value range is 30%, and this time it has only been lowered by 10%. The percentage has been revised from 40% to 30%.

[0043] Correction of phase identification threshold: The phase identification threshold is calculated as 60% of the first loss threshold, i.e., 5% × 60% = 3.0%. This value is within the allowable range of 50% to 80% of the first loss threshold, so the corrected value of 3.0% is directly taken.

[0044] Correction of the preset percentage threshold: Under normal operating conditions, the three-phase power loss change rate is evenly distributed and the statistical distribution has no significant deviation. Therefore, the preset percentage threshold remains unchanged at the original value of 55%.

[0045] Correction of the preset phase difference threshold: The preset phase difference threshold is calculated as 40% of the first loss threshold, i.e., 5% × 40% = 2.0%. This value is within the allowable range of 30% to 50% of the first loss threshold, so the corrected value of 2.0% is directly taken.

[0046] After this correction, the first loss threshold The percentage was reduced from 10% to 5%, which is closer to the statistical boundary under actual normal operating conditions of the transformer, thus improving the sensitivity of early fault identification. Second loss threshold. The threshold is adjusted synchronously to maintain a reasonable multiple relationship with the first loss threshold, but due to limitations in the correction step size and parameter value range, it is only reduced to 30% this time. The phase identification threshold is adjusted in conjunction with the first loss threshold to ensure the consistency of the fault phase identification criteria. At the same time, the parameter value range limitation, correction step size limitation, and anti-drift processing mechanism jointly ensure the robustness of the correction process, avoiding drastic changes in parameters under measurement noise or transient disturbances.

[0047] The deviation between the actual total power loss of the winding and the total theoretical winding loss at time k is defined as... : ; In the formula, This represents the actual total power loss of the transformer windings at time k; This represents the total theoretical winding loss of the transformer at time k; The winding material temperature coefficient and the additional loss correction term are used to construct the parameter vector to be corrected. The correction method is expressed as follows: ; In the formula, , These represent the parameter vectors to be corrected at time k+1 and time k, respectively; Indicates the correction step size; , These represent the minimum and maximum allowable values ​​of the parameter to be corrected, respectively. This represents a truncation function used to restrict the parameter correction result to within an allowed range.

[0048] Example 2: A transformer winding fault and phase identification device based on power loss change rate, comprising: One or more processors; Memory, used to store one or more computer programs; When one or more programs are executed by one or more processors, the one or more processors execute the method in Example 1.

[0049] Example 3: A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, cause the processor to perform the method in Example 1.

Claims

1. A method for identifying transformer winding faults and phases based on the rate of change of power loss, characterized in that... Includes the following steps: S1. Collect the voltage, current and temperature signals of the transformer as basic information, and obtain the transformer nameplate parameters including no-load loss and load loss. S2. During normal operation of the transformer, based on the transformer nameplate parameters and real-time collected basic information, calculate the fault-free calibration data of the transformer under the current load rate, and establish a theoretical power loss model characterizing the normal power loss of each phase winding and the entire transformer under different load rates and temperature conditions. The theoretical power loss model includes a load rate calculation model, a winding resistance temperature correction model, a theoretical winding loss model for each phase, and a total theoretical winding loss model. The load factor calculation model includes formulas for calculating the total load factor and the phase load factor: ； ； In the formula, This represents the total load factor of the transformer at time k; Let k be the transformer operating capacity at time k; This refers to the rated capacity of the transformer. Let i represent the load rate of phase i at time k, where i = A, B, C, corresponding to phases A, B, and C respectively; This represents the current in the i-th phase winding at time k; Rated current; The winding resistance temperature correction model is expressed as: ； In the formula, Indicates temperature The winding resistance of the i-th phase; Indicates reference temperature The winding resistance of the i-th phase; K is the temperature coefficient of the winding material; The theoretical winding loss model for each phase is expressed as follows: ； In the formula, This represents the theoretical winding loss of the i-th phase at time k; This represents the rated current of the i-th phase; This represents the additional loss correction term for the i-th phase. This represents the load rate of the i-th phase; The overall theoretical winding loss model is expressed as: ； In the formula, This represents the total theoretical winding loss of the transformer at time k; , , These represent the theoretical winding losses of phases A, B, and C at time k, respectively. S3. Based on the real-time collected basic information, calculate the actual total power loss of the transformer windings and the actual power loss of each phase winding, and calculate the total power loss change rate and the power loss change rate of each phase respectively. The actual total power loss of the winding is obtained by subtracting core losses from the difference between the active power input on the high-voltage side and the active power output on the low-voltage side, and is expressed as: ； In the formula, This represents the actual total power loss of the transformer windings at time k; This represents the active power input to the high-voltage side at time k; This represents the low-voltage side output active power at time k; This represents the core loss at time k; The actual power loss of each phase winding is calculated based on the winding resistance after temperature correction for each phase winding current, and is expressed as: ； In the formula, This represents the actual power loss of the i-th phase winding at time k; This represents the local additional loss of the i-th phase caused by inter-turn insulation degradation or short-circuit circulating current at time k. This represents the additional loss correction term for the i-th phase at time k. This represents the temperature at time k; The rate of change of total power loss is expressed as: ； In the formula, This represents the rate of change of the total power loss of the transformer at time k; The rate of change of power loss in each phase is expressed as: ； In the formula, This represents the rate of change of power loss of the i-th phase winding at time k; S4. Extract the maximum value of the total power loss change rate and the maximum value of the power loss change rate of each phase within the preset sampling window. Based on the maximum value of the total power loss change rate and the preset loss threshold and temperature threshold, determine the winding insulation fault status. S5. When it is determined that there is a winding insulation fault, the faulty phase is determined according to the relationship between the maximum values ​​of the power loss change rate of each phase and the phase-to-phase differences. S6. Generate early warning, trip, or lockout control commands based on the winding insulation fault status and fault phase, and perform online correction on the theoretical power loss model and the threshold parameters involved in determining the winding insulation fault status and fault phase.

2. The method for identifying transformer winding faults and phases based on power loss change rate according to claim 1, characterized in that, In S1, voltage and current signals are acquired by current and voltage sensors on the high-voltage and low-voltage sides, respectively; temperature signals are acquired by a transformer body temperature detection device. For oil-immersed transformers, the temperature signal is the oil temperature, and for dry-type transformers, the temperature signal is the winding temperature. Voltage and current signals are used to obtain input and output active power, load rate, and current characteristics of each phase, while temperature signals are used for winding resistance temperature correction and temperature threshold setting. The acquired voltage, current, and temperature signals undergo preprocessing, including time synchronization, outlier removal, missing value completion, and smoothing filtering.

3. The method for identifying transformer winding faults and phases based on power loss variation rate according to claim 1, characterized in that, In step S4, the maximum value of the total power loss change rate is extracted within the preset sampling window, and is expressed as: ； In the formula, This represents the maximum rate of change of total power loss within the preset sampling window at time k; Indicates the preset sampling window length; express The rate of change of the total power loss of the transformer at any given moment; The maximum value of the power loss change rate of each phase is extracted within the preset sampling window and expressed as: ； In the formula, This represents the maximum rate of change of power loss of the i-th phase winding within the preset sampling window at time k. express The rate of change of power loss of the i-th phase winding at time i.

4. The method for identifying transformer winding faults and phases based on the power loss change rate according to claim 1, characterized in that, In S4, before determining the winding insulation fault state, the voltage change rate and the main circuit current change rate are first calculated, expressed as: ； ； In the formula, Indicates the rate of change of voltage; , Let these represent the voltages of the i-th phase at times k and k-1, respectively. Indicates the rate of change of the main circuit current; , These represent the main circuit current of the i-th phase at times k and k-1, respectively. When the voltage, current, and temperature signals are all in a valid sampling state, and the voltage change rate is less than the voltage blocking threshold and the main circuit current change rate is less than the current blocking threshold, winding insulation fault diagnosis is allowed. If the above conditions are not met, it is determined that there is an external disturbance, temperature abnormality, or sampling abnormality in the current state, and the output of winding insulation fault diagnosis results is suspended.

5. The method for identifying transformer winding faults and phases based on power loss variation rate according to claim 1, characterized in that, In S4, the method for determining the winding insulation fault state is based on the current total load factor of the transformer. Set the first loss threshold Second loss threshold First loss threshold The second loss threshold corresponds to the critical degradation state of the inter-turn insulation of the winding. The corresponding winding inter-turn insulation is in a critical collapse state, and Greater than ; Set the first temperature threshold according to the transformer type. Second temperature threshold ,and Greater than ; When the maximum rate of change of total power loss is less than the first loss threshold When this is the case, it is determined that the inter-turn insulation of the winding is in a normal state; If the temperature signal is greater than or equal to the first temperature threshold at this time If so, an abnormal temperature alarm signal will be output or the process will begin. When the maximum rate of change of total power loss is greater than or equal to the first loss threshold And less than the second loss threshold When this occurs, the winding inter-turn insulation is determined to be in a deteriorated state, and a warning signal is output; If the temperature signal is greater than or equal to the second temperature threshold at this time If so, the warning level will be raised or the review period will be shortened; When the maximum rate of change of total power loss is greater than or equal to the second loss threshold When the winding inter-turn insulation is in a state of collapse or near collapse, a trip protection signal is output.

6. The method for identifying transformer winding faults and phases based on power loss change rate according to claim 5, characterized in that, In S5, the phase with the largest value among the maximum power loss change rates of the three phases is selected as the candidate fault phase. When the total power loss change rate has met the winding insulation fault judgment condition, and the maximum power loss change rate of each phase of the candidate fault phase is greater than the phase identification threshold under the current load rate, if the proportion of the power loss change rate of the candidate fault phase is greater than the preset proportion threshold or the difference in power loss change rate between the candidate fault phase and other phases is greater than the preset phase difference threshold, the candidate fault phase is determined to be the fault phase. If two or three phases simultaneously meet the above fault phase discrimination criteria, then a multi-phase winding inter-turn insulation abnormality signal will be output.

7. The method for identifying transformer winding faults and phases based on the power loss change rate according to claim 6, characterized in that, In S6, when the winding insulation is in a normal state, a normal operation signal is output; when the winding insulation is in a deteriorated state, a warning signal is output; when the winding insulation is in a collapse or near collapse state, a trip protection signal is output; when there is external disturbance, abnormal temperature, abnormal sampling, or the fault phase identification result does not meet the criteria, a blocking signal is output or the review process is entered. Online correction refers to correcting the winding material temperature coefficient and additional loss correction item based on the deviation between the actual total power loss and the total theoretical winding loss after the control cycle ends when the winding insulation is in a normal state. Based on the statistical distribution characteristics of the total power loss change rate under normal operating conditions, the first loss threshold, the second loss threshold, the phase identification threshold, the preset proportion threshold, and the preset phase difference threshold are corrected, and parameter value range restrictions, correction step size restrictions, and parameter anti-drift processing mechanisms are set.

8. The method for identifying transformer winding faults and phases based on power loss variation rate according to claim 7, characterized in that, In S6, the deviation between the actual total power loss of the winding and the total theoretical winding loss at time k is defined as... : ； In the formula, This represents the actual total power loss of the transformer windings at time k; This represents the total theoretical winding loss of the transformer at time k; The winding material temperature coefficient and the additional loss correction term are used to construct the parameter vector to be corrected. The correction method is expressed as follows: ； In the formula, , These represent the parameter vectors to be corrected at time k+1 and time k, respectively; Indicates the correction step size; , These represent the minimum and maximum allowable values ​​of the parameter to be corrected, respectively. This represents a truncation function used to restrict the parameter correction result to within an allowed range.

Images