CN-121995138-A - Loss diagnosis and health management system and method for transformer

CN121995138ACN 121995138 ACN121995138 ACN 121995138ACN-121995138-A

Abstract

The invention discloses a loss diagnosis and health management system and method of a transformer, which belong to the technical field of intelligent transformers and comprise a thermal fingerprint module, a theoretical temperature module and a net temperature rise module, wherein the thermal fingerprint module is used for recording a temperature drop curve of each part of the transformer after planned power failure, obtaining a thermal inertia fingerprint of the transformer based on multiple temperature drop curve fitting, the theoretical temperature module is used for calculating the theoretical cold body temperature of each part of the transformer based on real-time environment temperature and the thermal inertia fingerprint when the transformer operates, the net temperature rise module is used for obtaining the actual temperature value of each part of the transformer and subtracting the corresponding theoretical cold body temperature from the actual temperature value, and the weak thermal characteristic signals caused by insulation aging, loss abnormality and the like of the transformer can be accurately extracted from strong background noise such as load fluctuation, environment temperature and the like by the technologies such as thermal inertia fingerprint construction, net temperature rise separation and the like, so that early diagnosis of gradual health degradation is realized, support is provided for intervention of potential faults in advance, and a passive situation of post-operation treatment is changed.

Inventors

SHEN JIANKUN
LIU XUSHENG
LING JIE

Assignees

西安万硕电子科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260120

Claims (10)

1. A loss diagnosis and health management system for a transformer, comprising: the thermal fingerprint module is used for recording temperature drop curves of all parts of the transformer after planned power failure and obtaining thermal inertia fingerprints of the transformer based on multiple temperature drop curve fitting; The theoretical temperature module is used for calculating the theoretical cold body temperature of each part of the transformer based on the real-time environment temperature and the thermal inertia fingerprint when the transformer is in operation; the net temperature rise module is used for obtaining actual temperature values of all parts of the transformer, and subtracting the corresponding theoretical cold body temperature from the actual temperature values to obtain net heating body temperature rise; The heat loss module is used for establishing a corresponding relation between the temperature rise of the net heating element and the load current, obtaining a heat loss response curve and calculating the curvature variation of the heat loss response curve; the early warning judging module is used for generating a healthy stress coefficient according to the curvature variation and triggering early warning according to the variation trend of the healthy stress coefficient; and the degradation positioning module is used for injecting current pulses into the transformer winding after the early warning triggering, recording the temperature rise response curve of each part of the transformer to the current pulses, and determining the degradation zone position by analyzing the difference between the temperature rise response curve and the thermal inertia fingerprint.
2. The transformer loss diagnosis and health management system of claim 1, wherein the construction of the thermal inertia fingerprint comprises: reversely deducting a dynamic heat dissipation coefficient set according to a temperature drop curve reflected by the temperature record of the secondary environment in the cooling process after power failure is planned each time; clustering and normalizing the multiple groups of dynamic heat dissipation coefficients to obtain a reference heat dissipation structure vector; Based on the reference heat radiation structure vector, generating a standard cooling expected curve cluster through thermal network model simulation; and extracting a feature set from the standard cooling expected curve cluster to form a thermal inertia fingerprint.
3. The system for diagnosing and managing loss of a transformer according to claim 2, wherein calculating the theoretical cold body temperature of each part of the transformer comprises: According to the real-time environment temperature, carrying out parameter remapping on a reference heat dissipation structure vector in the thermal inertia fingerprint according to the temperature dependence relation of the thermal conductivity of the material to generate a dynamic thermal conductivity matrix; Based on the dynamic thermal conductivity matrix and the initial state of the residual heat distribution of the transformer, constructing and solving a heat flow balance equation, and deducing heat dissipation dynamics; and extracting initial temperature and initial cooling rate characteristics from the heat radiation dynamic, matching with standard cooling expected curve clusters in the thermal inertia fingerprint, and calculating and outputting theoretical cold body temperature through interpolation.
4. A transformer loss diagnosis and health management system according to claim 3, wherein the construction of the loss thermal response curve comprises: Based on the load current time sequence, performing hysteresis compensation based on a thermal time constant on the net heating element temperature rise sequence to generate a quasi-steady-state temperature rise sequence and an equivalent steady-state current value; Fitting the steady-state temperature rise sequence segments according to the equivalent steady-state current value alignment to obtain a plurality of local linear relation segments, and coupling the plurality of local linear relation segments based on a physical model of copper loss and iron loss to form an initial response curve; And calculating confidence coefficient according to the density and the dispersion degree of the data points in each current section, smoothing the initial response curve based on the confidence coefficient, and carrying out weighted fusion on the initial response curve and the historical loss thermal response curve to output the loss thermal response curve.
5. The system for diagnosing and managing the loss of a transformer according to claim 4, wherein calculating the curvature variation of the thermal response curve comprises: Selecting a reference current point, and calculating the form deviation between a loss thermal response curve and a theoretical secondary function substrate in a preset current window of the reference current point to obtain a primary curvature characteristic; According to the primary curvature characteristic, selecting N detection sections on the loss thermal response curve, and extracting the change rate of the slope in each detection section along with the current to form a multidimensional deformation characteristic vector; mapping the multidimensional deformation feature vector through a pre-constructed nonlinear feature interaction network to generate a comprehensive curvature degradation index; and carrying out trend test and persistence discrimination on the comprehensive curvature degradation index sequence generated in a time window with preset duration, and outputting the curvature variation.
6. The transformer loss diagnosis and health management system of claim 5, wherein the generation of the health stress coefficient comprises: calculating a standard fraction of the current curvature variation based on the current curvature variation and the historical curvature variation sequence; Mapping the standard fraction through a nonlinear conversion function based on an Arrhenius model to obtain an equivalent aging acceleration factor; and multiplying the equivalent aging acceleration factor by the current load rate and the running time derivative element and integrating to generate an accumulated healthy stress value serving as a healthy stress coefficient.
7. The system of claim 6, wherein the early warning is triggered according to a trend of change in the healthy stress coefficient, comprising: Carrying out local linear fitting based on a self-adaptive sliding window on the healthy stress coefficient sequence, and extracting the slope and the confidence interval of the current trend; combining the slope of the current trend, the confidence interval and the absolute level of the healthy stress coefficient, and carrying out joint state evaluation through a multi-stage early warning rule table to generate a preliminary early warning grade; Dynamically adjusting the subsequent data acquisition and calculation frequency according to the preliminary early warning level, and starting multi-period backtracking verification and consistency check when the preliminary early warning level reaches a preset level so as to confirm the early warning level; and generating gradient early warning instructions comprising the early warning level, the potential fault mode, the suggested response window and the maintenance priority based on the confirmed early warning level.
8. The transformer loss diagnosis and health management system of claim 7, wherein the determining of the degradation location comprises: Extracting a multi-time constant spectrum from the temperature rise response curve; Differentiating the multi-time constant spectrum with a health reference time constant spectrum stored in the thermal inertia fingerprint to generate an abnormal time constant offset vector; Projecting the abnormal time constant offset vector to a preset fault location mapping matrix, and calculating the confidence score of each candidate degradation location; and integrating the confidence score, the early warning level and the historical diagnosis record of the transformer, and generating a zone bit judgment conclusion and a recommended verification sequence through a decision tree.
9. The transformer loss diagnosis and health management system according to claim 2, wherein the thermal inertia fingerprint construction is further performed synchronously: under the healthy state of the transformer, current pulses which are the same as those used in fault diagnosis are injected into the winding of the transformer, and the temperature rise response curves of all parts are recorded; Extracting a multi-time constant spectrum from the temperature rise response curve; the multi-time constant spectrum is stored in the thermal inertia fingerprint as a health reference time constant spectrum.
10. A loss diagnosis and health management method for a transformer, applied to the loss diagnosis and health management system of a transformer according to any one of claims 1 to 9, comprising the steps of: Step 1, recording a temperature drop curve of each part of the transformer after planned power failure, and obtaining a thermal inertia fingerprint of the transformer based on multiple temperature drop curve fitting; Step 2, calculating theoretical cold body temperature of each part of the transformer based on real-time environment temperature and thermal inertia fingerprint when the transformer is in operation; step 3, obtaining actual temperature values of all parts of the transformer, and subtracting the corresponding theoretical cold body temperature from the actual temperature values to obtain net heating element temperature rise; step 4, establishing a corresponding relation between the temperature rise of the net heating element and the load current, obtaining a loss thermal response curve, and calculating the curvature variation of the loss thermal response curve; Step 5, generating a healthy stress coefficient according to the curvature variation, and triggering early warning according to the variation trend of the healthy stress coefficient; And 6, after the early warning triggering, injecting current pulses into the transformer winding, recording a temperature rise response curve of each part of the transformer to the current pulses, and determining a degradation zone position by analyzing the difference between the temperature rise response curve and the thermal inertia fingerprint.

Description

Loss diagnosis and health management system and method for transformer Technical Field The invention relates to the technical field of intelligent transformers, in particular to a loss diagnosis and health management system and method for a transformer. Background The transformer is used as main equipment of an electric power system, the operation stability of the transformer directly determines the safety and reliability of electric power transmission, in the current electric power operation and maintenance, the operation state monitoring of the transformer is generally realized by adopting equipment such as a WS-IDTM temperature controller, a TCM monitor and the like, the functions of the equipment are concentrated in collecting temperature data of all parts of the transformer in real time, when the temperature exceeds a preset threshold value, an overtemperature alarm is triggered, and a notification is pushed to operation staff in a telephone mode after a fault occurs, so that a conventional technical scheme of the temperature monitoring and basic alarm of the transformer is formed, however, the existing monitoring technology based on the equipment only can realize passive monitoring and threshold value alarm of the operation state of the transformer, lacks active judging capability of the health state of the equipment, and in an actual operation and maintenance scene, the technology can trigger the alarm only after the fault of the transformer appears, for example, when the temperature rises sharply due to the overheat of internal parts, the signal can be sent out, the progressive health degradation process generated inside the transformer cannot be identified early, the predictive maintenance work is more difficult to support, the operation and maintenance work is always in a passive potential face of post-treatment, the early warning is not carried out, the fault is not possible, and the fault and the operation loss is greatly increased. The deep technical reason for the limitation is that the parameter cognition of the transformer health state is deviated, namely, the comprehensive characterization parameter of temperature is directly equivalent to the equipment health state, the health problem of the transformer, such as aging of an insulating layer, abnormal iron core loss, excessive local winding loss and the like, cannot cause obvious temperature change at the early stage of degradation, only extremely weak extra temperature rise can be generated or fine distortion is caused to local heat distribution, the weak heat characteristic signals are extremely easily completely submerged by strong background noise such as load fluctuation, environmental temperature change and the like in the long-term operation process of the transformer, and are difficult to be effectively identified, the prior monitoring technology relies on absolute temperature value or simple temperature difference calculation to realize alarm judgment, the capability of separating and quantifying the weak heat characteristic signals from the complex background noise is lacked, and in the actual working condition that the transformer operates for a long time and the load and the environmental temperature change, the operation staff pursues accuracy and reduces the false alarm rate, the high threshold can always set a high temperature early warning threshold, the false alarm caused by the load fluctuation and the environmental temperature change can be effectively avoided, the important early stage of the important equipment failure characteristic signals are also taken as the false alarm signal caused by the load fluctuation, the serious early stage of the serious early stage failure can be greatly influenced by the severe power failure, the early stage failure of the early stage failure can be greatly, the power failure is stable, and the system is greatly influenced, and the power failure is greatly poor, and the early stage is stable, and the system can be stably developed, and the power failure is greatly bad, and the power failure is greatly caused. Disclosure of Invention Aiming at the problems existing in the prior art, the invention aims to provide a loss diagnosis and health management system and method for a transformer, which can accurately extract weak thermal characteristic signals caused by insulation aging and loss abnormality of the transformer from strong background noise such as load fluctuation, environment temperature and the like through technologies such as thermal inertia fingerprint construction, net temperature rise separation and the like, realize early diagnosis of progressive health degradation, provide support for intervention potential faults in advance and change the passive situation of post-operation and maintenance treatment. In order to solve the problems, the invention adopts the following technical scheme: In a first aspect, a loss diagnosis and health management system for a transformer, compris