A prediction and early warning device for power equipment

Abstract

The present invention relates to the technical field of power equipment detection and discloses a power equipment prediction and early warning device, including: a data acquisition module for obtaining the number of turns voltage of the power equipment transformer coil and the temperature of the transformer coil; a two-port network for constructing a voltage-current two-port impedance matrix through the main magnetic circuit inductance, leakage magnetic capacitance and core temperature, and predicting power equipment faults through the power equipment impedance characteristics; a fault analysis module for analyzing power equipment faults through the voltage-current two-port impedance matrix, performing node-by-node fault processing on the power equipment, clustering multiple nodes, and processing cluster processing conflicts, thereby realizing intelligent detection of power equipment transformers and intelligent fault inspection, and greatly reducing the complexity of the inspection system.

Description

Technical Field

[0001] The invention relates to the technical field of power equipment detection and discloses a power equipment prediction and early warning device. Background Art

[0002] In power systems, transformers are core equipment, and their operating status directly impacts the reliability and safety of the power grid. Traditional transformer fault monitoring technologies rely primarily on regular maintenance, partial discharge detection, or threshold alarms based on a single parameter. Existing technologies typically make judgments based on independent parameters such as temperature, oil chromatography, or vibration. This single-valued test indicator for power equipment transformers results in a single evaluation metric, making it impossible to provide targeted predictions and early warnings based on the device's inherent characteristics. Traditional methods use fixed equivalent circuit models, resulting in large deviations in resonant frequency calculations. Cluster power equipment fault handling strategies have numerous shortcomings, including uncoordinated redundant switching that can lead to local overloads. Transformers from different manufacturers lack unified identification coding and data interfaces, making cluster-level monitoring system integration difficult. Summary of the Invention

[0003] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0004] In order to solve the above technical problems, the main purpose of the present invention is to provide a power equipment prediction and early warning device, comprising:

[0005] The data acquisition module includes a voltage acquisition unit for setting multiple parameter acquisition nodes of the power equipment transformer and acquiring parameters, and a temperature acquisition unit for acquiring the transformer coil temperature;

[0006] Two-port network, used to construct a voltage-current two-port impedance matrix using the main magnetic circuit inductance, leakage capacitance, and core temperature, and to predict power equipment faults using the impedance characteristics of the power equipment;

[0007] The fault analysis module is used to analyze power equipment faults through the voltage-current two-port impedance matrix, handle power equipment faults by node, and perform cluster processing on multiple nodes, and to build a power equipment arbitration model and handle cluster processing conflicts.

[0008] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0009] Setting multiple transformers of the power equipment as multiple nodes, wherein the multiple parameter collection nodes include voltage, magnetic flux density, equivalent impedance, main magnetic circuit inductance and leakage capacitance;

[0010] The voltage acquisition unit acquires the turn-to-turn voltages of multiple nodes of the power equipment;

[0011] Collecting magnetic flux densities of multiple nodes of the power equipment through a Hall sensor array, and inputting the magnetic flux densities of the multiple nodes into a resonance compensation unit;

[0012] The resonance compensation unit is provided with a multi-stage filter, dynamically adjusts the filtering parameters, and performs filtering processing on the collected magnetic flux density signal and inter-turn voltage.

[0013] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0014] n physical ports are provided at the n node winding endpoints and the core yoke of the power equipment, the n physical ports are connected to the Hall sensor array and the voltage acquisition unit, the port magnetic flux density is acquired through the Hall sensor array, and the port voltage is acquired through the voltage acquisition unit, and the acquisition is kept synchronous;

[0015] Extract magnetic flux density features including peak value, kurtosis and spectrum;

[0016] The signal characteristic values ​​of the extracted voltage include the effective value of the voltage.

[0017] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0018] Establishing a theoretical table of electromagnetic characteristics of the iron core magnetic circuit, and optimizing the theoretical table of electromagnetic characteristics by setting an open circuit test on the first and second ports of the power equipment transformer;

[0019] Perform an open-circuit test on the first port, measure and obtain the theoretical first-port voltage and the theoretical second-port voltage under the first-port open-circuit test, calculate the first-port impedance of the transformer, perform an open-circuit test on the second port, measure and obtain the theoretical first-port voltage and the theoretical second-port voltage under the second-port open-circuit test, calculate the second-port impedance, apply voltage to the first port, connect the impedance Z_load to the second port, measure the dual-port current, and optimize the transformer equivalent inductance, magnetic flux density, temperature, magnetic permeability, leakage capacitance and inter-turn current in the electromagnetic characteristics theoretical table.

[0020] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0021] The actual open-circuit voltage of the first port and the open-circuit voltage of the second port of the power equipment transformer are obtained, and the temperature of the iron core is collected simultaneously. By collecting the total magnetomotive force, magnetic field strength and iron core temperature of the transformer in real time, the dynamic magnetic permeability in the optimized electromagnetic characteristics theoretical table is retrieved, and the main magnetic circuit inductance and leakage magnetic capacitance are obtained through the dynamic magnetic permeability.

[0022] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0023] The influence of the transformer's nonlinear voltage distortion on the magnetic circuit resonant frequency is obtained through the main magnetic circuit inductance and leakage magnetic capacitance. If the transformer input voltage is equal to the transformer network resonance, resonance compensation is triggered.

[0024] The resonance compensation includes voltage nonlinear compensation, reverse compensation and PID adjustable capacitance compensation.

[0025] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0026] Based on the data of main magnetic circuit inductance, leakage magnetic capacitance and core temperature, a two-port impedance matrix of voltage and current at the transformer port of the power equipment is mapped. The impedance elements in the two-port impedance matrix of voltage and current are determined by the main magnetic circuit inductance, main magnetic circuit impedance, leakage magnetic capacitance and real impedance.

[0027] Obtaining the main magnetic circuit inductance change rate through real-time main magnetic circuit inductance, and triggering a saturation warning if the main magnetic circuit inductance change rate is greater than a threshold;

[0028] Obtaining a resonant frequency offset through real-time magnetic leakage capacitance, wherein the resonant frequency offset is calculated through the resonant frequency, and determining that the magnetic leakage capacitance is aged if the resonant frequency offset is greater than an offset threshold;

[0029] The temperature change rate of the eddy current loss resistance is obtained through the core temperature. If the temperature change rate of the eddy current loss resistance exceeds the resistance change threshold, it is determined that the eddy current loss is abnormal;

[0030] The fault analysis module extracts the main magnetic circuit inductance change rate, resonant frequency offset and eddy current loss resistance characteristic values ​​of the transformer port through the two-port impedance matrix of the power equipment transformer port voltage and current to predict the fault probability distribution.

[0031] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0032] By defining the identification of the power equipment transformer, obtaining the voltage and current two-port impedance matrix with the identification, partitioning the transformer, and then identifying the nodes of the partitioned transformer, the transformer fault analysis is performed by using the voltage and current two-port impedance matrix with the identification;

[0033] Performing fault prediction on the transformer with the node identification according to the result of the fault analysis;

[0034] Node fault redundancy processing is performed through the fault prediction.

[0035] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0036] Transformer cluster fault processing is performed through transformer fault prediction results identified by multiple nodes;

[0037] A cluster evaluation is performed on the transformer cluster fault processing results. If no fault processing conflict occurs in the cluster evaluation, the cluster fault processing is completed. If a fault processing conflict occurs in the cluster evaluation, the fault processing conflicts occurring in the cluster evaluation are layered and the layered faults are processed.

[0038] As a preferred solution of the power equipment prediction and early warning device of the present invention, wherein:

[0039] Adjust fault conflict handling strategies in different areas through the power equipment arbitration model;

[0040] The power equipment arbitration model includes a propagation coefficient of a transformer fault node, a regional load standard deviation, and a cluster conflict index, and regional hierarchical optimization is dispatched through the cluster conflict index.

[0041] Beneficial effects of the present invention:

[0042] This application integrates multi-dimensional parameters such as temperature, magnetic permeability, and leakage capacitance to construct a two-port network impedance matrix, map the electromagnetic characteristics of the iron core in real time, switch from node-level redundancy to cluster-level conflict arbitration, and achieve dual optimization of fault isolation and resource coordination. Combined with resonant adaptive compensation, it improves fault prediction accuracy and harmonic suppression capabilities, defines a unified equipment identification and zoning strategy, and quickly deploys transformer cluster parameter detection and maintenance, reducing fault downtime and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0044] Figure 1 This is a working diagram of a power equipment prediction and early warning device of the present invention;

[0045] Figure 2 This is a system composition diagram of a power equipment prediction and early warning device according to the present invention;

[0046] Figure 3 This is a topological diagram of a voltage-current two-port impedance matrix circuit of a power equipment prediction and early warning device of the present invention;

[0047] Figure 4 This is a flow chart of a fault analysis method for a power equipment prediction and early warning device of the present invention;

[0048] Figure 5 This is a multi-node and partition diagram of a power equipment prediction and early warning device of the present invention. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0051] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0052] like Figure 2 As shown, a prediction and early warning device for electric power equipment includes:

[0053] The data acquisition module includes a voltage acquisition unit for acquiring the voltage of the turns of the transformer coil of the power equipment and a temperature acquisition unit for acquiring the temperature of the transformer coil.

[0054] like Figure 1 As shown, the workflow of a power equipment prediction and early warning device includes:

[0055] A specific implementation method of a data acquisition module includes:

[0056] The voltage acquisition unit is used to collect the transformer turn-to-turn voltage.

[0057] The main sensor of the temperature acquisition unit is connected to the A / D converter to convert the AC signal into a DC signal. A backup temperature sensor is also provided. When the main sensor fails or is abnormal, the backup temperature sensor is used.

[0058] Setting multiple transformers of the power equipment as multiple nodes, setting n physical ports at the winding endpoints and core yokes of the n nodes of the power equipment, connecting the n physical ports to a Hall sensor array and a voltage acquisition unit, acquiring port magnetic flux density through the Hall sensor array, and acquiring port voltage through the voltage acquisition unit, and maintaining synchronous acquisition;

[0059] Extract magnetic flux density features including peak value, kurtosis and spectrum;

[0060] The signal characteristic values ​​of the extracted voltage include the effective value of the voltage.

[0061] The voltage acquisition unit obtains the turn-to-turn voltages of multiple nodes of the power equipment;

[0062] Collecting magnetic flux densities of multiple nodes of the power equipment through a Hall sensor array, and inputting the magnetic flux densities of the multiple nodes into a resonance compensation unit;

[0063] The resonance compensation unit sets a multi-stage filter, dynamically adjusts the filtering parameters, and performs filtering on the collected magnetic flux density signal and inter-turn voltage;

[0064] Dynamically adjusting the filtering parameters is actually adaptively adjusting the damping coefficient.

[0065] The two-port network is used to construct a voltage-current two-port impedance matrix through the main magnetic circuit inductance, leakage capacitance and core temperature, and predict power equipment faults through the impedance characteristics of power equipment.

[0066] Establishing a theoretical table of electromagnetic characteristics of the iron core magnetic circuit, and optimizing the theoretical table of electromagnetic characteristics by setting an open circuit test on the first and second ports of the power equipment transformer;

[0067] An open-circuit test is performed on the first port to obtain the theoretical first-port voltage and the theoretical second-port voltage under the first-port open-circuit test, and the impedance of the first port of the transformer is calculated. An open-circuit test is performed on the second port to obtain the theoretical first-port voltage and the theoretical second-port voltage under the second-port open-circuit test, and the impedance of the second port is calculated. Voltage is applied to the first port, and the second port is connected to the adjustable impedance Z_load. The dual-port current is measured, and the transformer equivalent inductance, magnetic flux density, temperature, magnetic permeability, leakage capacitance and inter-turn current of the electromagnetic characteristics theoretical table are optimized.

[0068] Specifically, a transformer test platform with two standard test ports is built, equipped with a programmable signal source, a precision voltmeter, an ammeter and an impedance analyzer.

[0069] The second port is physically short-circuited, a sinusoidal voltage is applied to the first port, and the relationship between the port voltage and the excitation current is measured.

[0070] A DC current is injected into the first port, the second port is short-circuited with a low-resistance wire, and the port voltage changes are recorded.

[0071] An AC voltage is applied to the first port, and the second port is connected to an adjustable impedance Z_load to monitor the current and voltage waveforms of the two ports.

[0072] Furthermore, the signal conditioning process includes using a voltage sensor to obtain the port voltage and transmitting it to the data acquisition module through a shielded cable.

[0073] The current is measured by placing a Hall effect current sensor around the winding to be measured, and the terminal current is monitored in real time.

[0074] Furthermore, optimizing the electromagnetic characteristic theoretical table includes load testing, parameter optimization, temperature compensation and dynamic updating of the electromagnetic characteristic theoretical table.

[0075] The load test involves applying a sinusoidal voltage to the first port, connecting an adjustable impedance Z_load to the second port, measuring the dual-port voltage V1 / V2 and current I1 / I2, and extracting the amplitude ratio and phase difference through vector analysis.

[0076] Furthermore, parameter optimization measures the capacitance between ports through high-frequency resonance, compensates for the stray capacitance in the electromagnetic characteristics theoretical table, and extracts Im=V1 / (2πf×Rm) from the short-circuit test data, where Rm is the port equivalent resistance. The relationship between Im and the magnetic circuit saturation is analyzed, and the magnetic circuit saturation content in the electromagnetic characteristics theoretical table is further updated.

[0077] Where V1 represents the voltage at the first port, V2 represents the voltage at the second port, I1 represents the current at the first port, I2 represents the current at the second port, Z_load is the load impedance connected to the second port, Im is a current value extracted from the short-circuit test data, Rm is the port equivalent resistance, f is the frequency, Lm is the excitation inductance, and Cλ is the inter-port capacitance of the two ports.

[0078] Based on the principle of measuring and obtaining the theoretical first-port voltage and the theoretical second-port voltage under the first-port open-circuit test, the theoretical first-port voltage and the second-port voltage under the second-port open-circuit test can be obtained.

[0079] The actual open-circuit voltage of the first port and the open-circuit voltage of the second port of the power equipment transformer are obtained, and the temperature of the iron core is collected simultaneously. By collecting the total magnetomotive force, magnetic field strength and iron core temperature of the transformer in real time, the dynamic magnetic permeability in the revised electromagnetic characteristics theoretical table is retrieved, and the main magnetic circuit inductance and leakage magnetic capacitance are obtained through the dynamic magnetic permeability.

[0080] The influence of the transformer's nonlinear voltage distortion on the magnetic circuit resonant frequency is obtained through the main magnetic circuit inductance and leakage magnetic capacitance. If the transformer input voltage is equal to the transformer network resonance, resonance compensation is triggered.

[0081] Specifically, the evolution direction of the main magnetic circuit inductance, leakage capacitance and resonant frequency is analyzed through coupling analysis.

[0082] Furthermore, the main magnetic circuit inductance includes the degree of core saturation, which directly affects the Lm value. When the input voltage increases and the magnetic flux density approaches the saturation point, Lm shows a nonlinear decrease.

[0083] The leakage capacitance C in the insulating medium between the windings is affected by temperature and voltage gradients, and Cλ changes nonlinearly with the voltage amplitude.

[0084] Resonant frequency evolution including the theoretical resonant frequency of the system , nonlinear parameter changes cause f_res to deviate from the design value.

[0085] A specific implementation method for dynamic monitoring of resonant frequency includes: indirectly reflecting Lm changes by detecting the phase difference change between the port voltage and current. The phase difference increases when saturated. Using an oscilloscope to capture 100Hz high-frequency oscillations, analyzing its amplitude and frequency relationship, locating the resonant point, automatically adjusting the Z_load impedance, and measuring the amplitude-phase characteristic curve of the port impedance. The extreme point of the curve corresponds to the resonant frequency.

[0086] The resonance compensation includes voltage nonlinear compensation, reverse compensation and PID adjustable capacitance compensation;

[0087] Specifically, voltage nonlinear compensation targets voltage spikes or troughs caused by core saturation by injecting a reverse compensation current to offset the nonlinear effect, and collects the input voltage waveform in real time, compares it with an ideal sine wave template, and generates a compensation signal.

[0088] Furthermore, reverse compensation is achieved by connecting an additional resonant branch in parallel in the transformer network. Its resonant frequency is designed to be the same as the main line resonant frequency. The main line resonant amplitude is reduced by energy absorption, and the additional branch parameters are adjusted in real time according to the changes in Lm and Cλ to maintain the stability of the total resonant frequency.

[0089] PID adjustable capacitor compensation monitors the resonant frequency deviation in real time and uses the PID algorithm to dynamically adjust the adjustable capacitor value, so that the system resonant frequency is always far away from the input voltage frequency. It also automatically optimizes the capacitor value according to load changes to compensate for parameter drift.

[0090] Specifically, the PID algorithm sets the integral, differential, and proportional constants to take the system disturbance as input, dynamically eliminates the system disturbance, and adjusts the capacitance value.

[0091] Furthermore, the adjustable capacitor adopts a MOS capacitor array, and the resolution adjustment is achieved through the SPI interface, and the PID parameters are set to the proportional coefficient Kp, integral time Ti, and differential time Td.

[0092] Based on the data of main magnetic circuit inductance, leakage magnetic capacitance and core temperature, a two-port impedance matrix of voltage and current at the transformer port of the power equipment is mapped. The impedance elements in the two-port impedance matrix of voltage and current are determined by the main magnetic circuit inductance, main magnetic circuit impedance, leakage magnetic capacitance and real impedance.

[0093] A specific implementation method of a voltage-current two-port impedance matrix includes:

[0094] like Figure 3 As shown, the basic expression of the voltage and current two-port impedance matrix is ​​as follows:

[0095]

[0096] Where Z11 is the ratio of input port voltage to input current;

[0097] Z22 is the ratio of output port voltage to output current;

[0098] Z12 and Z21 are the electromagnetic coupling strengths between ports;

[0099] Z is the voltage and current two-port impedance matrix.

[0100] Based on the basic principle of voltage-current two-port impedance matrix, the loop of the power equipment transformer is mapped into a two-port network, and the physical parameters of the transformer are quantified by establishing a two-port matrix.

[0101] Furthermore, the inductive reactance Xm of the main magnetic circuit is determined by the inductance Lm of the main magnetic circuit: Xm=2πf×Lm, which changes nonlinearly with the saturation degree of the core. When saturated, Lm drops sharply, resulting in a decrease in Xm and an increase in the port voltage.

[0102] The leakage magnetic reactance Xλ is determined by the leakage magnetic capacitance Cλ: Xλ=1 / (2πf×Cλ). As the temperature rises and the voltage distortion decays, when Cλ decays, Xλ increases, causing the resonant frequency to shift downward.

[0103] The main magnetic circuit impedance Rm is used to reflect the eddy current loss of the iron core and is positively correlated with the temperature. Rm=Kt×T, Kt is the temperature coefficient, T is the temperature. When the temperature rises, Rm increases.

[0104] Obtaining the main magnetic circuit inductance change rate through real-time main magnetic circuit inductance, and triggering a saturation warning if the main magnetic circuit inductance change rate is greater than a threshold;

[0105] By comparing the Lm values ​​under no-load and rated load conditions, the main magnetic circuit inductance change rate ΔLm / Δt is calculated. For example, if the inductance drops from 120μH to 100μH within 10 seconds, the change rate is 2μH / s, where ΔLm is the inductance change and Δt is the time change.

[0106] Furthermore, a safety threshold is set. When ΔLm / Δt is less than the safety threshold, it is a normal aging rate. An alarm threshold is set. When ΔLm / Δ is greater than or equal to the alarm threshold, a core saturation warning is triggered. If the warning is triggered, the input voltage is automatically reduced and the event is recorded in the health file. If it continues to deteriorate, oil chromatography analysis is started.

[0107] The resonant frequency offset is obtained through the real-time magnetic leakage capacitance, and the resonant frequency offset is calculated through the resonant frequency. If the resonant frequency offset is greater than an offset threshold, it is determined that the magnetic leakage capacitance is aged.

[0108] Specifically, the method for obtaining the resonant frequency offset includes the following steps: , real-time measurement of the extreme point frequency. The impedance analyzer is used to scan the port impedance curve to find the extreme point frequency. The absolute value of the difference between the actual resonant frequency f_measured and the theoretical resonant frequency f_res is calculated, and the ratio of the absolute value of the difference to the theoretical resonant frequency f_res is multiplied by 100% to obtain the resonant frequency offset.

[0109] Specifically, the aging judgment method judges by the resonant frequency offset. If the resonant frequency offset is less than the offset threshold, no aging occurs. If the resonant frequency offset is greater than the offset threshold, it is determined that the leakage magnetic capacitor is aged.

[0110] Specifically, the threshold is set according to the actual production standard of the transformer.

[0111] The eddy current loss resistance temperature change rate is obtained through the core temperature. If the eddy current loss resistance temperature change rate exceeds the resistance change threshold, it is determined that the eddy current loss is abnormal.

[0112] Specifically, the temperature data is periodically collected through sensors arranged on the surface of the core, and the change in the main magnetic circuit resistance ΔRm / Δt = (actual resistance - historical resistance) / (actual time - historical time) is calculated.

[0113] The fault analysis module extracts the main magnetic circuit inductance change rate, resonant frequency offset and eddy current loss resistance characteristic values ​​of the transformer port through the two-port impedance matrix of the voltage and current of the power equipment transformer port to predict the fault probability distribution.

[0114] like Figure 4 As shown, a specific implementation method of a fault analysis module includes:

[0115] Fault analysis models include fault mode library and power equipment fault analysis model;

[0116] Specifically, the fault mode library includes core saturation, leakage capacitance aging, temperature anomaly, and winding turn-to-turn short circuit.

[0117] Furthermore, the core saturation includes a sudden drop in the imaginary part of Z11 and a decrease in the real part Xm, which leads to an increase in the port voltage.

[0118] The aging of leakage magnetic capacitance includes the increase of leakage magnetic capacitance reactance Xλ and the downward shift of resonant frequency.

[0119] The short circuit between winding turns includes abnormal amplitude of Z12 and Z21, and the coupling degree between ports decreases.

[0120] Specifically, the power equipment fault analysis model includes power equipment parameter feature extraction, establishing real-time monitoring parameters and parameter feature prediction through the extracted feature values, dynamically adjusting fault problems through parameter feature prediction, and updating the fault mode library.

[0121] Furthermore, the method for establishing a power equipment fault analysis model includes:

[0122] Determine the depth of the power equipment fault analysis model and set it according to different power equipment models to prevent overfitting. If the depth of the power equipment fault analysis model is too large, the training data will be sensitive to noise. If the depth of the power equipment fault analysis model is too small, it will not be able to capture the complex changes in parameters.

[0123] A random data selection function is defined. Through the random data selection function, the voltage and current two-port impedance matrix data set of the power equipment is determined to be a random sample, and the probability of inconsistency between the theoretical value of the label and the actual value of the label is reduced, making the data selection more accurate.

[0124] Fault warning categories include green health, yellow warning, and red warning. Fault modes include core saturation, leakage capacitor aging, winding inter-turn short circuit, and temperature abnormality caused by partial discharge. A fault probability function is also determined, which is used to obtain the failure probability of a certain fault type.

[0125] Furthermore, the failure probability function is used to map the probability distribution of multiple failure modes, and outputs the probability of occurrence of different failure modes through the failure probability function, rather than a yes or no judgment.

[0126] Through probability distribution, we can identify the causes of primary and secondary faults, formulate fault handling strategies, give priority to high-probability faults, and avoid ineffective maintenance. The fault probability changes dynamically over time, reflecting the progress of the fault development. If a yellow warning is triggered, preventive tests are arranged. If a red warning is triggered, an emergency shutdown process is triggered to avoid unplanned power outages. Through the standardized mapping from numerical values ​​to probabilities of the power equipment fault analysis model, the dimensional differences between multiple fault modes are resolved. The fault mode-driven warning classification converts the abstract model output into actionable operation and maintenance decisions and explainable fault tracing. The primary and secondary fault causes are identified through probability distribution.

[0127] The real resistance and imaginary inductance of the parameters in the voltage-current two-port impedance matrix Z are separated and normalized to between 0 and 1. The temperature is standardized using the Z-Score to obtain the normalized results of the inductance change rate and the main magnetic circuit impedance change rate. The impedance, temperature, inductance change rate, and main magnetic circuit impedance change rate data change trends are statistically analyzed by setting a time window, and the maximum fluctuation coefficient and trend slope are obtained.

[0128] It should be noted that the core saturation affects the change of the main magnetic circuit inductance, resulting in the change of the core magnetic permeability, and the core magnetic permeability affects the change of the magnetic flux density. The main magnetic circuit impedance is divided into real resistance and imaginary impedance, and the imaginary impedance is inductive reactance or capacitive reactance.

[0129] Furthermore, the effect of temperature on magnetic permeability is mapped using piecewise function or polynomial fitting, and a temperature compensation model is established to compensate for the nonlinear effect of temperature on magnetic permeability.

[0130] To train the power equipment fault analysis model, the data set is divided into M groups. The model is trained with M-1 groups in sequence, and the model is verified with the Mth group. The average accuracy is calculated to verify the generalization ability of the model and avoid overfitting the training data.

[0131] The verified power equipment fault analysis model outputs the conditional probability of each fault type, such as P (core saturation) and P (leakage capacitance aging), and sets safety thresholds and alarm thresholds based on requirements.

[0132] Upgrade fault diagnosis from "yes / no" judgment to probability distribution prediction, achieve quantitative ranking of main causes of faults, dynamically adjust model depth and data selection strategy, balance accuracy and robustness, retrain the model after fault repair, and realize automatic system update.

[0133] like Figure 2 As shown, a prediction and early warning device for electric power equipment includes:

[0134] The fault analysis module includes a module for analyzing power equipment faults through a voltage-current two-port impedance matrix, performing node-by-node fault processing on power equipment, performing cluster processing on multiple nodes, and handling cluster processing conflicts.

[0135] Specifically, by defining the identification of the power equipment transformer, obtaining the voltage and current two-port impedance matrix with the identification, partitioning the transformer, and performing node identification on the partitioned transformer, the transformer fault analysis is performed by using the voltage and current two-port impedance matrix with the identification;

[0136] Performing fault prediction on the transformer with the node identification according to the result of the fault analysis;

[0137] Node fault redundancy processing is performed through the fault prediction.

[0138] The specific implementation method of transformer zoning includes: core zoning, the lamination group divides the core into multiple physical segments, each segment is equipped with a temperature sensor, the clamp is used to monitor the core tightening force, the high-voltage winding is divided into multiple segments, each segment is equipped with a partial discharge sensor, the low-voltage winding is divided into multiple segments, and a pressure sensor is installed. The upper oil temperature in the oil tank partition is measured by multi-point thermometers, and the bottom oil temperature is measured by an ultrasonic flowmeter. In the electrical partition, the input port of the port partition is the voltage measurement point and the secondary winding of the current transformer, and the output port is the load side voltage and current sampling point.

[0139] Furthermore, the equivalent node partitioning includes port nodes and internal nodes;

[0140] The port nodes include input terminal Node1, output terminal Node2, and neutral point Node3;

[0141] Internal nodes include core node Node4 and winding node Node5;

[0142] In the voltage-current two-port impedance matrix, Z11 is the ratio of the input port voltage to the current, which is between Node1 and Node2;

[0143] Z22 is the output port voltage to current ratio, between Node2 and Node3;

[0144] Z12 and Z21 are the impedances of the inter-port coupling between Node1 and Node3, and between Node2 and Node3.

[0145] Transformer cluster fault processing is performed through multi-node identification of transformer fault prediction results;

[0146] A cluster evaluation is performed on the transformer cluster fault processing results. If no fault processing conflict occurs in the cluster evaluation, the cluster fault processing is completed. If a fault processing conflict occurs in the cluster evaluation, the fault processing conflicts occurring in the cluster evaluation are layered and the layered faults are processed.

[0147] like Figure 5As shown in the figure, S1, S2, S3, and S4 are multi-node identified transformer clusters. Furthermore, the power equipment network in the actual production process is divided into multiple areas, and multiple areas are used as multiple clusters S1, S2, S3, and S4. The black dots in the figure are power equipment transformers. If a problem occurs with a power equipment transformer with an identifier in a cluster, the power equipment transformer fault is first processed. After the processing is completed, the real-time operation data of the transformers in all regions are obtained, and the fault processing records of multiple labeled transformers in each cluster are recorded, including processing time, resource consumption, whether chain failures are caused, etc. For example, if a transformer in S1 fails, after completing the fault processing of the transformer, the processing time, resource consumption, and whether it is a chain failure are recorded. Whether it causes chain failures or fault handling conflicts. If it causes chain failures, the fault will be reprocessed and an early warning will be triggered. If a conflict occurs in the fault handling, the conflict level will be divided to determine whether the fault handling of a transformer in the S1 area will cause overload in the S2, S3 and S4 areas. The cluster fault will be handled according to the conflict level and category to prevent the transformer of a certain node from affecting one area or one area from affecting the entire group of related power equipment. Whether it will affect other power equipment will be designed and judged according to the actual situation, and further overall judgment is needed. This embodiment only provides a cluster conflict handling for power equipment transformers. Other power equipment can make conflict judgments based on load voltage abnormalities, load changes, etc.

[0148] A specific implementation method for handling conflicts in transformer cluster faults includes:

[0149] Obtain real-time operating data for all regional transformers, record fault handling records, calculate average fault repair time, calculate the occupancy rate of key equipment such as mobile SVGs and maintenance robots, compare regional load fluctuations before and after fault handling, count the number of abnormal voltage fluctuations, and accumulate power outage losses, equipment maintenance costs, and labor costs.

[0150] The health of the region is judged, including whether various indicators such as processing time, cascading failure rate, and economic loss meet the green health standard. If so, it is green health. If any indicator exceeds the threshold but does not reach the red warning standard, it is judged as a yellow warning. when A red alert is issued when there is a risk of cascading failures or huge economic losses.

[0151] Conflict resolution is initiated when multiple regions request mobile SVG simultaneously or when handling a failure would overload other regions.

[0152] The specific conflict levels include urgent conflict Level 1, important conflict Level 2, observed conflict Level 3 and ignored conflict Level 4.

[0153] If a Level 1 emergency conflict causes a voltage collapse in an adjacent area during fault handling, the adjacent area's SVG will be immediately called to isolate the faulty area and initiate a manual emergency response.

[0154] Important conflict Level 2: When abnormal load fluctuation occurs during fault handling, the local backup regional SVG is dispatched first, load transfer is initiated, and transformer taps are adjusted online.

[0155] Observe the conflict. Level 3 fault handling may cause intermittent load fluctuations. In this case, it is necessary to increase the inspection frequency and deploy SVGs in hot spots in advance.

[0156] Ignore conflicts. Level 4 has no significant impact on fault handling and is therefore included in the monthly maintenance plan and scheduled for maintenance during off-peak hours.

[0157] Furthermore, priority is given to the urgent conflict areas, shared resources are allocated on demand, and resources are released and marked as available after the conflict is resolved to avoid repeated contention.

[0158] After the conflict is resolved, SVG is automatically deployed and the fault area is isolated based on the conflict level.

[0159] Automatically adjust the conflict level based on real-time operation data to avoid relying on static thresholds, pre-configure backup resources for emergency conflicts and important conflict areas, shorten response time, display conflict heat maps and resource occupancy status, support manual forced intervention, and encapsulate high-frequency fault handling steps into a knowledge base to reduce the burden of manual decision-making.

[0160] Transformer cluster fault processing is performed through transformer fault prediction results identified by multiple nodes;

[0161] A cluster evaluation is performed on the transformer cluster fault processing results. If no fault processing conflict occurs in the cluster evaluation, the cluster fault processing is completed. If a fault processing conflict occurs in the cluster evaluation, the fault processing conflicts occurring in the cluster evaluation are layered and the layered faults are processed.

[0162] Through the identification system, zoning, node-based fault analysis and intelligent cluster processing mechanism, a full-process closed loop from accurate diagnosis to efficient handling of power equipment failures has been achieved.

[0163] Adjust fault conflict handling strategies in different areas through the power equipment arbitration model;

[0164] The power equipment arbitration model includes the propagation coefficient of the transformer fault node, the regional load standard deviation and the cluster conflict index, and the regional hierarchical optimization is dispatched through the cluster conflict index.

[0165] A specific implementation method of a power equipment arbitration model includes:

[0166] The power equipment arbitration model includes propagation coefficient, regional load standard deviation, cluster conflict index and hierarchical optimization unit.

[0167] The propagation coefficient reflects the impact of fault handling in a certain area on other areas. A larger value indicates a wider impact. The measurement dimension assesses the closeness of the connection between regions through indicators such as the short-circuit current ratio and the power flow change rate, and calculates the load fluctuation amplitude of adjacent areas after fault handling.

[0168] Specific applications include: if the fault handling in area A causes load fluctuations in area B, the propagation coefficient of area A is determined to be high impact; if the fault handling in area C has no significant impact on other areas, the propagation coefficient is determined to be low impact.

[0169] The regional load standard deviation is used to count the fluctuations in transformer load within the region. A larger value indicates a more unstable system.

[0170] The cluster conflict index is used to weight and fuse indicators such as the propagation coefficient, regional load standard deviation, and historical processing time to generate a unique conflict index.

[0171] The scheduling mechanism of the hierarchical optimization unit includes four-layer processing strategy, dynamic resource allocation strategy and conflict resolution.

[0172] Furthermore, the four-layer processing strategy includes the emergency layer, the important layer, the observation layer, and the ignore layer.

[0173] Specifically, at the emergency level, immediate action is required to ensure the safety of the power grid, automatically engage backup power, isolate the faulty area, and mobilize surrounding resources for emergency support.

[0174] At the critical layer, queues need to be processed, local resources are used first, regional backup transformers are dispatched, and preventive maintenance procedures are initiated.

[0175] At the observation level, continuous monitoring is required, timely processing is required, the frequency of fault prediction is enhanced, and maintenance plans are optimized.

[0176] At the neglect level, it is included in the regular maintenance plan and does not require priority treatment.

[0177] Dynamic resource allocation includes mobile SVG, emergency power generation vehicles, regional spare parts warehouses, and multi-functional maintenance robots.

[0178] The emergency layer monopolizes resources and gives priority to calling the most recently available resources; the important layer competes based on the ratio of conflict index to resource demand; the observation layer polls and allocates shared resources to ensure fairness.

[0179] Conflict resolution methods include:

[0180] When the conflict index of two or more regions exceeds the standard, arbitration is triggered. The marginal benefit of the conflict is calculated to avoid the potential losses of the N-1 failure of the power grid, and the direct cost of the failure is processed. The region with the highest marginal benefit to cost ratio is selected for priority processing.

[0181] If resources are insufficient, initiate cross-regional collaboration.

[0182] Furthermore, the power equipment arbitration model identifies chain risk through the propagation coefficient, evaluates system vulnerability through the load standard deviation, and quantifies comprehensive contradictions through the conflict index. Combined with the four-layer scheduling strategy and dynamic resource allocation, it reduces the transformer chain failure rate, shortens the emergency response time, improves the resource utilization rate of transformer maintenance, and greatly reduces the average repair time, annual economic losses and operation and maintenance costs.

[0183] It is important to note that the configuration and arrangement of the present application, as illustrated in various exemplary embodiments, are exemplary only. Although only two embodiments are described in detail in this disclosure, those reading this disclosure should readily appreciate that numerous modifications are possible without materially departing from the novel teachings and advantages of the subject matter described herein, including, for example, variations in the size, dimensions, structure, shape, and proportions of various components, as well as parameter values ​​(e.g., temperature, pressure, etc.), mounting arrangements, use of materials, color, orientation, and the like. For example, components shown as integrally formed may be constructed from multiple parts or components, the positions of components may be inverted or otherwise altered, and the nature, number, or position of discrete components may be modified or changed. Therefore, all such modifications are intended to be encompassed within the scope of this invention. The order or sequence of any process or method steps may be altered or reordered according to alternative embodiments. Any "means-plus-function" clause is intended to cover structures that perform the functions described herein, and not only structural equivalence but also structural equivalents. Other substitutions, modifications, changes, and omissions may be made in the design, operating conditions, and arrangement of the exemplary embodiments without departing from the scope of this invention. Therefore, the present invention is not limited to a specific embodiment, but extends to various modifications.

[0184] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment may not be described (i.e., those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention).

[0185] It should be understood that in the development of any actual embodiment, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, the development effort will be a routine task of design, fabrication, and production without undue experimentation.

[0186] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, and all of these should be included in the scope of the present invention.

Claims

1. A prediction and early warning device for electric power equipment, characterized in that: include: The data acquisition module includes a voltage acquisition unit for setting multiple parameter acquisition nodes of the power equipment transformer and acquiring parameters, and a temperature acquisition unit for acquiring the transformer coil temperature; Two-port network, used to construct a voltage-current two-port impedance matrix using the main magnetic circuit inductance, leakage capacitance, and core temperature, and to predict power equipment faults using the impedance characteristics of the power equipment; The fault analysis module is used to analyze power equipment faults through the voltage and current two-port impedance matrix, handle power equipment faults by node, and perform cluster processing on multiple nodes, by building a power equipment arbitration model and handling cluster processing conflicts; Based on the data of main magnetic circuit inductance, leakage magnetic capacitance and core temperature, a two-port impedance matrix of voltage and current at the transformer port of the power equipment is mapped. The impedance elements in the two-port impedance matrix of voltage and current are determined by the main magnetic circuit inductance, main magnetic circuit impedance, leakage magnetic capacitance and real impedance. Obtaining the main magnetic circuit inductance change rate through real-time main magnetic circuit inductance, and triggering a saturation warning if the main magnetic circuit inductance change rate is greater than a threshold; Obtaining a resonant frequency offset through real-time magnetic leakage capacitance, wherein the resonant frequency offset is calculated through the resonant frequency, and determining that the magnetic leakage capacitance is aged if the resonant frequency offset is greater than an offset threshold; The temperature change rate of the eddy current loss resistance is obtained through the core temperature. If the temperature change rate of the eddy current loss resistance exceeds the resistance change threshold, it is determined that the eddy current loss is abnormal. The fault analysis module extracts the main magnetic circuit inductance change rate, resonant frequency offset and eddy current loss resistance characteristic values ​​of the transformer port through the two-port impedance matrix of the power equipment transformer port voltage and current to predict the fault probability distribution.

2. The power equipment prediction and early warning device according to claim 1, characterized in that: Setting multiple transformers of the power equipment as multiple nodes, wherein the multiple parameter collection nodes include voltage, magnetic flux density, equivalent impedance, main magnetic circuit inductance and leakage magnetic capacitance; The voltage acquisition unit acquires the turn-to-turn voltages of multiple nodes of the power equipment; Collecting magnetic flux densities of multiple nodes of the power equipment through a Hall sensor array, and inputting the magnetic flux densities of the multiple nodes into a resonance compensation unit; The resonance compensation unit is provided with a multi-stage filter, dynamically adjusts the filtering parameters, and performs filtering processing on the collected magnetic flux density signal and inter-turn voltage.

3. The power equipment prediction and early warning device according to claim 1, characterized in that: n physical ports are provided at the n node winding endpoints and the core yoke of the power equipment, the n physical ports are connected to the Hall sensor array and the voltage acquisition unit, the port magnetic flux density is acquired through the Hall sensor array, and the port voltage is acquired through the voltage acquisition unit, and the acquisition is kept synchronous; Extract magnetic flux density features including peak value, kurtosis and spectrum; The signal characteristic values ​​of the extracted voltage include the effective value of the voltage.

4. The power equipment prediction and early warning device according to claim 3, characterized in that: Establishing a theoretical table of electromagnetic characteristics of the iron core magnetic circuit, and optimizing the theoretical table of electromagnetic characteristics by setting an open circuit test on the first and second ports of the power equipment transformer; Perform an open-circuit test on the first port, measure and obtain the theoretical first-port voltage and the theoretical second-port voltage under the first-port open-circuit test, calculate the first-port impedance of the transformer, perform an open-circuit test on the second port, measure and obtain the theoretical first-port voltage and the theoretical second-port voltage under the second-port open-circuit test, calculate the second-port impedance, apply voltage to the first port, connect the impedance Z_load to the second port, measure the dual-port current, and optimize the transformer equivalent inductance, magnetic flux density, temperature, magnetic permeability, leakage capacitance and inter-turn current in the electromagnetic characteristics theoretical table.

5. The power equipment prediction and early warning device according to claim 4, characterized in that: The actual open-circuit voltage of the first port and the open-circuit voltage of the second port of the power equipment transformer are obtained, and the temperature of the iron core is collected simultaneously. By collecting the total magnetomotive force, magnetic field strength and iron core temperature of the transformer in real time, the dynamic magnetic permeability in the optimized electromagnetic characteristics theoretical table is retrieved, and the main magnetic circuit inductance and leakage magnetic capacitance are obtained through the dynamic magnetic permeability.

6. The power equipment prediction and early warning device according to claim 5, characterized in that: The influence of the transformer's nonlinear voltage distortion on the magnetic circuit resonant frequency is obtained through the main magnetic circuit inductance and leakage magnetic capacitance. If the transformer input voltage is equal to the transformer network resonance, resonance compensation is triggered. The resonance compensation includes voltage nonlinear compensation, reverse compensation and PID adjustable capacitance compensation.

7. The power equipment prediction and early warning device according to claim 6, characterized in that: By defining the identification of the power equipment transformer, obtaining the voltage and current two-port impedance matrix with the identification, partitioning the transformer, and then identifying the nodes of the partitioned transformer, the transformer fault analysis is performed by using the voltage and current two-port impedance matrix with the identification; Performing fault prediction on the transformer with the node identification according to the result of the fault analysis; Node fault redundancy processing is performed through the fault prediction.

8. The power equipment prediction and early warning device according to claim 7, characterized in that: Transformer cluster fault processing is performed through transformer fault prediction results identified by multiple nodes; A cluster evaluation is performed on the transformer cluster fault processing results. If no fault processing conflict occurs in the cluster evaluation, the cluster fault processing is completed. If a fault processing conflict occurs in the cluster evaluation, the fault processing conflicts occurring in the cluster evaluation are layered and the layered faults are processed.

9. The power equipment prediction and early warning device according to claim 8, characterized in that: Adjust fault conflict handling strategies in different areas through the power equipment arbitration model; The power equipment arbitration model includes a propagation coefficient of a transformer fault node, a regional load standard deviation, and a cluster conflict index, and regional hierarchical optimization is dispatched through the cluster conflict index.

Images