Method and system for detecting partial discharge in high-voltage switch cabinet

Abstract

The invention provides a method for detecting partial discharge in a high-voltage switch cabinet, and the method comprises the steps: receiving an ultrasonic signal and an ultraviolet signal which are generated by the partial discharge in the high-voltage switch cabinet and are synchronously collected by an acousto-optic integrated sensor probe in real time, and carrying out the preprocessing; performing phase distribution PRDP spectrum feature analysis on the preprocessed ultrasonic signal and ultraviolet signal to obtain ultraviolet intensity and ultrasonic intensity, and inputting the obtained ultraviolet intensity and ultrasonic intensity into a trained discharge classification model at the same time to obtain a partial discharge type in the high-voltage switch cabinet; wherein the partial discharge types comprise corona discharge, surface discharge and internal discharge. By implementing the method, the defects of a traditional partial discharge detection method can be overcome, and the discharge energy level and the risk degree can be quantitatively evaluated and tracked.

Description

Technical Field

[0001] The present invention relates to the technical field of power detection, and particularly to a method and system for detecting partial discharge in a high-voltage switchgear cabinet. Background Art

[0002] In the operation and maintenance of high-voltage switchgear cabinets, partial discharge phenomena are one of the main causes of insulation faults. Traditional partial discharge detection methods (such as pulse current method, ultra-high frequency method, earth current method, ultrasonic method, and optical measurement method, etc.) have problems such as being easily interfered by on-site noise, low defect detection rate, and high complexity of feature extraction calculation. In addition, limited by the detection efficiency of sensors, the confidence level of partial discharge pattern recognition and diagnosis is poor, the positioning accuracy of the partial discharge source is relatively low, and it is impossible to quantitatively evaluate and track the discharge energy level and risk based on the existing discharge monitoring results. Summary of the Invention

[0003] The technical problem to be solved by the embodiments of the present invention is to provide a method and system for detecting partial discharge in a high-voltage switchgear cabinet, which can solve the defects of traditional partial discharge detection methods and can also quantitatively evaluate and track the discharge energy level and risk.

[0004] To solve the above technical problem, an embodiment of the present invention provides a method for detecting partial discharge in a high-voltage switchgear cabinet, and the method includes the following steps:

[0005] Receiving ultrasonic signals and ultraviolet light signals generated by partial discharge in the high-voltage switchgear cabinet synchronously and in real time by an acousto-optic integrated sensor probe, and performing preprocessing;

[0006] Performing phase distribution PRDP map feature analysis on the ultrasonic signal and ultraviolet light signal after preprocessing to obtain the ultraviolet light intensity and ultrasonic intensity, and simultaneously inputting the obtained ultraviolet light intensity and ultrasonic intensity into a trained discharge classification model to obtain the partial discharge type in the high-voltage switchgear cabinet; wherein, the partial discharge type includes corona discharge, surface discharge, and internal discharge.

[0007] Wherein, the method further includes:

[0008] Obtaining the discharge intensity in the high-voltage switchgear cabinet from a preset two-dimensional feature vector set according to the obtained ultraviolet light intensity and ultrasonic intensity.

[0009] Wherein, the method further includes:

[0010] Predefining a two-dimensional feature vector set; wherein, the two-dimensional feature vector set is constructed in a Cartesian coordinate system with the ultraviolet light intensity as the X-axis and the ultrasonic intensity as the Y-axis, and each discharge intensity is mapped as a point in the two-dimensional space.

[0011] Among them, the discharge intensity inside the high-voltage switchgear automatically marks the corresponding partial discharge type according to its coordinate position and is displayed with a certain brightness value.

[0012] Among them, the method further includes:

[0013] Based on the obtained ultraviolet light intensity and ultrasonic intensity, and in combination with a preset ultrasonic-ultraviolet power correlation model, calculate the apparent discharge power, and further calculate the corresponding energy level index according to a preset energy level exponential function to evaluate the risk level;

[0014] Among them, if the obtained energy level index is greater than or equal to the first threshold, it is determined as a high risk; if the obtained energy level index is less than the first threshold and greater than or equal to the second threshold, it is determined as a medium risk; if the obtained energy level index is less than the second threshold, it is determined as a low risk.

[0015] Among them, the expression of the ultrasonic-ultraviolet power correlation model is P = αU + βS + γU·S; where,

[0016] P is the apparent discharge power; U is the ultraviolet light intensity; S is the ultrasonic intensity; α, β, and γ are all fitting coefficients and are all constants.

[0017] Among them, the expression of the energy level exponential function is E level = w 1 ·U + w 2 ·S + w 3 ·P; where,

[0018] E level is the energy level index; ω 1 , ω 2 , ω 3 are all weight coefficients and are all constants, ω 1 + ω 2 + ω 3 = 1.

[0019] An embodiment of the present invention also provides a partial discharge detection system inside a high-voltage switchgear, including:

[0020] An acoustic-optic signal synchronous acquisition unit, configured to receive the ultrasonic signal and ultraviolet light signal generated by partial discharge inside the high-voltage switchgear synchronously and in real time by an acoustic-optic integrated sensor probe, and perform preprocessing;

[0021] The partial discharge type detection unit is used to analyze the phase distribution PRDP map features of the ultrasonic signal and ultraviolet light signal after preprocessing, so as to obtain the ultraviolet light intensity and ultrasonic intensity, and input the obtained ultraviolet light intensity and ultrasonic intensity into the trained discharge classification model at the same time to obtain the partial discharge type inside the high-voltage switchgear; among them, the partial discharge type includes corona discharge, surface discharge and internal discharge.

[0022] Among them, it also includes:

[0023] The partial discharge intensity detection unit is used to obtain the discharge intensity inside the high-voltage switchgear in the preset two-dimensional feature vector set according to the obtained ultraviolet light intensity and ultrasonic intensity.

[0024] Among them, it also includes:

[0025] The discharge energy level detection unit is used to calculate the apparent discharge power based on the obtained ultraviolet light intensity and ultrasonic intensity, and in combination with the preset ultrasonic-ultraviolet power correlation model, and further calculate the corresponding energy level index according to the preset energy level exponential function to evaluate the risk level;

[0026] Among them, if the obtained energy level index is greater than or equal to the first threshold, it is determined as a high risk; if the obtained energy level index is less than the first threshold and greater than or equal to the second threshold, it is determined as a medium risk; if the obtained energy level index is less than the second threshold, it is determined as a low risk.

[0027] Implementing the embodiments of the present invention has the following beneficial effects:

[0028] 1. After receiving the ultrasonic signal and ultraviolet light signal generated by partial discharge inside the high-voltage switchgear synchronously and in real time by the acousto-optic integrated sensor probe and performing preprocessing, the present invention analyzes the phase distribution PRDP map features to obtain the ultraviolet light intensity and ultrasonic intensity, and further inputs them into the trained discharge classification model at the same time to quickly obtain the partial discharge type inside the high-voltage switchgear, thus avoiding problems such as being interfered by on-site noise, low defect detection rate, and high computational complexity of feature extraction, and solving the defects of traditional partial discharge detection methods being limited by the detection efficiency of sensors, poor confidence in partial discharge pattern recognition and diagnosis, and low accuracy of partial discharge source location;

[0029] 2. The present invention also quickly obtains the discharge intensity inside the high-voltage switchgear in the two-dimensional feature vector set according to the ultrasonic signal and ultraviolet light signal, so as to realize the detection and tracking of the discharge intensity;

[0030] 3. The present invention also calculates the apparent discharge power based on the ultraviolet light intensity and the ultrasonic intensity, and in combination with the ultrasonic ultraviolet power correlation model, and further calculates the corresponding energy level index according to the energy level exponential function to evaluate the risk level, so as to realize the quantitative evaluation and tracking of the discharge energy level and the degree of danger. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, obtaining other drawings based on these drawings still belongs to the scope of the present invention.

[0032] Figure 1 It is a flowchart of a method for detecting partial discharge in a high-voltage switchgear cabinet provided by an embodiment of the present invention;

[0033] Figure 2 It is an application scenario diagram of a method for detecting partial discharge in a high-voltage switchgear cabinet provided by an embodiment of the present invention;

[0034] Figure 3 It is a schematic structural diagram of a system for detecting partial discharge in a high-voltage switchgear cabinet provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings.

[0036] As Figure 1 shown, in an embodiment of the present invention, a method for detecting partial discharge in a high-voltage switchgear cabinet is provided. The method includes the following steps:

[0037] Step S1: Receive the ultrasonic signal and the ultraviolet light signal generated by partial discharge in the high-voltage switchgear cabinet synchronously and real-time by the acousto-optic integrated sensor probe, and perform preprocessing;

[0038] Step S2: Perform phase distribution PRDP map feature analysis on the preprocessed ultrasonic signal and ultraviolet light signal to obtain the ultraviolet light intensity and the ultrasonic intensity, and input the obtained ultraviolet light intensity and ultrasonic intensity into the trained discharge classification model at the same time to obtain the partial discharge type in the high-voltage switchgear cabinet; wherein, the partial discharge type includes corona discharge, surface discharge, and internal discharge.

[0039] The specific process is as follows. In step S1, first, an acoustic-optical integrated sensor probe integrated with an ultrasonic sensor and an ultraviolet sensor is used to synchronously collect the ultrasonic signal and the ultraviolet signal generated by partial discharge in the high-voltage switchgear.

[0040] Among them, the frequency response range of the ultrasonic wave is 20 kHz–200 kHz, the sensitivity is -65 dBV / μPa, and the sampling rate is 1 MS / s; the response wavelength of the ultraviolet light is 240 nm - 280 nm, the photon counting rate is 0 - 10 6 cps, and the dynamic range is 60 dB. At this time, an optical filter (cutoff wavelength 300 nm) is added to the ultraviolet channel to suppress visible light interference.

[0041] Secondly, after receiving the ultrasonic signal and the ultraviolet signal, conditioning is carried out. Specifically, the ultrasonic signal is conditioned by band-pass filtering (center frequency 100 kHz, bandwidth ±20 kHz) and adaptive gain amplification (gain range 20 dB - 60 dB); the ultraviolet signal is conditioned by using photon counting technology and combining with the wavelet threshold denoising algorithm (selecting sym4 wavelet basis, decomposition level 5 layers, soft threshold processing).

[0042] Furthermore, these two conditioned signals are converted into digital signals through analog-to-digital conversion.

[0043] In step S2, first, a discharge classification model is constructed based on the support vector machine (SVM); the input of this discharge classification model is the ultraviolet light intensity and the ultrasonic intensity, and the output is the partial discharge type including corona discharge, surface discharge, and internal discharge.

[0044] Furthermore, after the historical ultrasonic signals and ultraviolet signals are preprocessed in the same way as in step S1, through the phase distribution PRDP map feature analysis, the ultraviolet light intensity and the ultrasonic intensity are obtained, and the ultraviolet light intensity and the ultrasonic intensity are formed into a feature vector to construct a sample set, and each sample is automatically marked with the corresponding partial discharge type; the marked sample set is divided to obtain training samples and test samples; the discharge classification model is trained with the training samples and tested with the test samples, so as to obtain a trained discharge classification model.

[0045] Then, the ultrasonic signal and the ultraviolet signal after preprocessing in step S1 are subjected to phase distribution PRDP map feature analysis to obtain the ultraviolet light intensity and the ultrasonic intensity.

[0046] Finally, the obtained ultraviolet light intensity and ultrasonic intensity are simultaneously input into the trained discharge classification model to obtain one of the partial discharge types in the high-voltage switchgear, namely corona discharge, surface discharge, and internal discharge.

[0047] In an embodiment of the present invention, a two-dimensional feature vector is also defined based on the ultraviolet light intensity and the ultrasonic light intensity as to construct a two-dimensional feature vector set for realizing defect visualization and recognition in the Cartesian coordinate. This form can eliminate the influence of the absolute intensity, focus on the coupling relationship of the photoacoustic signal, and enhance the separability of the defect patterns. Among them, the two-dimensional feature vector set is constructed in the Cartesian coordinate system, with the ultraviolet light intensity U as the X-axis and the ultrasonic wave intensity S as the Y-axis, and each discharge intensity is mapped to a point in the two-dimensional space. At this time, in the Cartesian coordinate system, the discharge intensity will automatically label the corresponding partial discharge type according to its coordinate position and be displayed with a certain brightness value. That is, according to the historical ultraviolet light intensity and ultrasonic wave intensity distribution, the regional boundaries of different defect types are delimited in the Cartesian coordinate system. Once new ultraviolet light intensity and ultrasonic wave intensity are input, the partial discharge type will be automatically identified according to its coordinate position and highlighted in the interface, realizing defect visualization and recognition in the Cartesian coordinate.

[0048] It should be noted that in the Cartesian coordinate system are the values after the ultraviolet light intensity and the ultrasonic wave intensity are normalized respectively, that is, the ultraviolet light intensity ratio and the ultrasonic wave intensity ratio.

[0049] Thus, the method further includes: according to the obtained ultraviolet light intensity and ultrasonic wave intensity, the discharge intensity in the high-voltage switchgear can be quickly obtained in the above two-dimensional feature vector set.

[0050] In an embodiment of the present invention, a multi-dimensional partial discharge energy level development path is established based on the binary spectral radiation power and the apparent discharge power, and corresponding energy level quantization indexes are formulated to evaluate and track risks. Therefore, the method further includes: based on the obtained ultraviolet light intensity and ultrasonic wave intensity, and combined with the preset ultrasonic-ultraviolet power correlation model P = αU + βS + γU·S, calculate the apparent discharge power P, and further according to the preset energy level exponential function E level = w 1 ·U + w 2 ·S + w 3 ·P, calculate the corresponding energy level index E level , to evaluate the risk level; where, if the obtained energy level index E level is greater than or equal to the first threshold (T 2 ), it is determined as a high risk; if the obtained energy level index E level is less than the first threshold and greater than or equal to the second threshold (T 1 ), it is determined as a medium risk; if the obtained energy level index E level is less than the second threshold, it is determined as a low risk. For example, low risk: E level < T 1; Medium risk: T 1 ≤E level <T 2 ; High risk: E level ≥T 2 。

[0051] It should be noted that α, β, and γ are all fitting coefficients and are all constants; ω 1 , ω 2 , ω 3 are all weight coefficients and are all constants, ω 1 +ω 2 +ω 3 = 1.

[0052] In an example, the ultraviolet light intensity U = 0.83; the ultrasonic intensity S = 0.60; the apparent discharge power P = 0.40; the weight coefficients w 1 , w 2 , w 3 are 0.5, 0.3, and 0.2 respectively. Substituting into the energy level quantization formula gives

[0053] E level = 0.5×0.83 + 0.3×0.60 + 0.2×0.40 = 0.72

[0054] At this time, the set thresholds are: low risk: E level <0.5; medium risk: 0.5 ≤ E level <0.7; high risk: E level ≥ 0.7. Among them, the final result: E level = 0.72 → high risk (triggering a red warning).

[0055] As Figure 2 shown, the application scenario of a partial discharge detection method in a high-voltage switchgear in an embodiment of the present invention is further described as follows:

[0056] First step, during partial discharge, synchronously collect ultrasonic signals and ultraviolet light signals;

[0057] Second step, condition the ultrasonic signal through band-pass filtering (center frequency 100 kHz, bandwidth ±20 kHz) and adaptive gain amplification (gain range 20 dB - 60 dB); and, condition the ultraviolet light signal using photon counting technology, combined with a wavelet threshold denoising algorithm (selecting the sym4 wavelet basis, decomposition level 5 layers, soft threshold processing);

[0058] Third step, perform analog-to-digital conversion on both signals adjusted in the second step into digital signals.

[0059] Fourth step, through phase distribution PRDP map feature analysis, extract the ultraviolet light intensity and ultrasonic intensity;

[0060] Step 5: Normalize the ultraviolet light intensity and ultrasonic intensity to obtain the proportion of ultraviolet light intensity and the proportion of ultrasonic intensity for subsequent discharge mode recognition, discharge intensity recognition, and risk assessment;

[0061] Step 6: Identify the discharge intensity based on the two-dimensional feature vector set; or directly import it into the discharge classification model to identify the partial discharge type; or combine the preset ultrasonic-ultraviolet power correlation model to calculate the apparent discharge power, and further calculate the corresponding energy level index according to the preset energy level exponential function to evaluate the risk level.

[0062] Based on the above discharge recognition model, the cross-validation results of the accuracy statistics of the discharge recognition model in Table 1 are obtained based on 1000 groups of measured data (300 groups of corona discharges, 400 groups of surface discharges, and 300 groups of internal discharges). Table 1 is used to quantify and visualize the performance of the classification results and clarify the recognition reliability of each discharge type.

[0063] Table 1

[0064] Discharge type Accuracy rate (%) Recall rate (%) Precision rate (%) F1-Score (%) Corona discharge 98.2 96.5 97.8 97.1 Surface discharge 93.4 91.2 94.1 92.6 Internal discharge 89.7 87.3 88.9 88.1 Overall index 94.3 91.7 93.6 92.6

[0065] Among them, Score refers to the harmonic mean of recall rate and precision rate, which is used to comprehensively evaluate the model performance.

[0066] Typical area examples: Corona discharge: high ultraviolet intensity (U>0.8), low ultrasonic intensity (S<0.3), corresponding to a high accuracy rate (98.2%) in the table. Internal discharge: low ultraviolet intensity (U<0.4), high ultrasonic intensity (S>0.7S>0.7), with a slightly lower recall rate (87.3%) due to signal overlap.

[0067] As Figure 3 shown, in the embodiment of the present invention, a partial discharge detection system in a high-voltage switchgear is provided, including:

[0068] An acoustic-optic signal synchronous acquisition unit 110, configured to receive the ultrasonic signal and ultraviolet light signal generated by partial discharge in the high-voltage switchgear synchronously and in real time collected by the acoustic-optic integrated sensor probe, and perform preprocessing;

[0069] A partial discharge type detection unit 120, configured to perform phase distribution PRDP map feature analysis on the preprocessed ultrasonic signal and ultraviolet light signal to obtain the ultraviolet light intensity and ultrasonic intensity, and input the obtained ultraviolet light intensity and ultrasonic intensity into the trained discharge classification model at the same time to obtain the partial discharge type in the high-voltage switchgear; among them, the partial discharge type includes corona discharge, surface discharge, and internal discharge.

[0070] Among them, it further includes:

[0071] The partial discharge intensity detection unit is used to obtain the discharge intensity inside the high-voltage switchgear in a preset two-dimensional feature vector set according to the obtained ultraviolet light intensity and ultrasonic intensity.

[0072] Among them, it further includes:

[0073] The discharge energy level detection unit is used to calculate the apparent discharge power based on the obtained ultraviolet light intensity and ultrasonic intensity and in combination with a preset ultrasonic-ultraviolet power correlation model, and further calculate the corresponding energy level index according to a preset energy level exponential function to evaluate the risk level;

[0074] Among them, if the obtained energy level index is greater than or equal to the first threshold, it is determined as a high risk; if the obtained energy level index is less than the first threshold and greater than or equal to the second threshold, it is determined as a medium risk; if the obtained energy level index is less than the second threshold, it is determined as a low risk.

[0075] Implementing the embodiments of the present invention has the following beneficial effects:

[0076] 1. After receiving the ultrasonic signal and ultraviolet light signal generated by partial discharge inside the high-voltage switchgear synchronously and in real time by the acousto-optic integrated sensor probe and performing preprocessing, the present invention conducts phase distribution PRDP map feature analysis to obtain the ultraviolet light intensity and ultrasonic intensity, and further inputs them into the trained discharge classification model at the same time to quickly obtain the partial discharge type inside the high-voltage switchgear, thus avoiding problems such as being interfered by on-site noise, low defect detection rate, and high computational complexity of feature extraction, and solving defects such as the traditional partial discharge detection method being limited by the detection efficiency of the sensor, poor confidence in partial discharge pattern recognition and diagnosis, and low accuracy of discharge source positioning;

[0077] 2. The present invention also quickly obtains the discharge intensity inside the high-voltage switchgear in a two-dimensional feature vector set according to the ultrasonic signal and ultraviolet light signal, thereby realizing the detection and tracking of the discharge intensity;

[0078] 3. The present invention also calculates the apparent discharge power based on the ultraviolet light intensity and ultrasonic intensity and in combination with the ultrasonic-ultraviolet power correlation model, and further calculates the corresponding energy level index according to the energy level exponential function to evaluate the risk level, thereby realizing the quantitative evaluation and tracking of the discharge energy level and danger degree.

[0079] It should be noted that in the above system embodiments, the included system modules are only divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, the specific names of the functional modules are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0080] Those of ordinary skill in the art can understand that all or part of the steps in implementing the methods of the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disks, optical discs, etc.

[0081] The above-disclosed are only the preferred embodiments of the present invention, and of course, the scope of the rights of the present invention cannot be limited thereby. Therefore, equivalent changes made according to the claims of the present invention still fall within the scope covered by the present invention.

Claims

1. A method for detecting partial discharge in a high-voltage switch cabinet, characterized in that: The method comprises the following steps: Receive the ultrasonic signal and ultraviolet light signal generated by the local discharge in the high-voltage switch cabinet collected synchronously and in real time by the sound and light integrated sensor probe, and perform pre-processing; The pre-processed ultrasonic signal and ultraviolet light signal are subjected to phase distribution PRDP spectrum feature analysis to obtain ultraviolet light intensity and ultrasonic wave intensity, and the obtained ultraviolet light intensity and ultrasonic wave intensity are simultaneously input into the trained discharge classification model to obtain the local discharge type in the high-voltage switchgear; wherein the local discharge type includes corona discharge, surface discharge and internal discharge.

2. The method for detecting partial discharge in a high-voltage switch cabinet according to claim 1, characterized in that: The method further comprises: According to the obtained ultraviolet light intensity and ultrasonic wave intensity, the discharge intensity in the high-voltage switch cabinet is obtained in a preset two-dimensional feature vector set.

3. The method for detecting partial discharge in a high-voltage switch cabinet according to claim 2, characterized in that: The method further comprises: A two-dimensional feature vector set is predefined; wherein the two-dimensional feature vector set is constructed in a Cartesian coordinate system, with ultraviolet light intensity as the X-axis, ultrasonic intensity as the Y-axis, and each discharge intensity is mapped to a point in the two-dimensional space.

4. The method for detecting partial discharge in a high-voltage switch cabinet according to claim 3, characterized in that: The discharge intensity in the high-voltage switch cabinet is automatically marked with the corresponding partial discharge type according to its coordinate position and displayed with a certain brightness value.

5. The method for detecting partial discharge in a high-voltage switch cabinet according to claim 1, characterized in that: The method further comprises: Based on the obtained ultraviolet light intensity and ultrasonic wave intensity, and in combination with the preset ultrasonic ultraviolet power correlation model, the apparent discharge power is calculated, and further according to the preset energy level index function, the corresponding energy level index is calculated to evaluate the risk level; Among them, if the obtained energy level index is greater than or equal to the first threshold, it is determined to be high risk; if the obtained energy level index is less than the first threshold and greater than or equal to the second threshold, it is determined to be medium risk; if the obtained energy level index is less than the second threshold, it is determined to be low risk.

6. The method for detecting partial discharge in a high-voltage switch cabinet according to claim 5, characterized in that: The expression of the ultrasonic ultraviolet power correlation model is P=αU+βS+γU·S; wherein, P is the apparent discharge power; U is the ultraviolet light intensity; S is the ultrasonic intensity; α, β, and γ are all fitting coefficients and are all constants.

7. The method for detecting partial discharge in a high-voltage switch cabinet according to claim 6, characterized in that: The expression of the energy level exponential function is E level =w1·U+w2·S+w3·P; in, E level is the energy level index; ω1, ω2, ω3 are all weight coefficients and are all constants, ω1+ω2+ω3=1.

8. A partial discharge detection system in a high-voltage switch cabinet, characterized in that: include: The acoustic and optical signal synchronous acquisition unit is used to receive the ultrasonic signal and ultraviolet light signal generated by the local discharge in the high-voltage switch cabinet synchronously and in real time collected by the acoustic and optical integrated sensor probe, and perform pre-processing; The local discharge type detection unit is used to perform phase distribution PRDP spectrum feature analysis on the pre-processed ultrasonic signal and ultraviolet light signal to obtain ultraviolet light intensity and ultrasonic wave intensity, and simultaneously input the obtained ultraviolet light intensity and ultrasonic wave intensity into the trained discharge classification model to obtain the local discharge type in the high-voltage switch cabinet; wherein the local discharge type includes corona discharge, surface discharge and internal discharge.

9. The local discharge detection system in a high-voltage switch cabinet according to claim 8, characterized in that: Also includes: The local discharge intensity detection unit obtains the discharge intensity in the high-voltage switch cabinet in a preset two-dimensional feature vector set according to the obtained ultraviolet light intensity and ultrasonic wave intensity.

10. The local discharge detection system in a high-voltage switch cabinet according to claim 9, characterized in that: Also includes: The discharge energy level detection unit is used to calculate the apparent discharge power based on the obtained ultraviolet light intensity and ultrasonic wave intensity and in combination with a preset ultrasonic ultraviolet power correlation model, and further calculate the corresponding energy level index according to a preset energy level index function to evaluate the risk level; Among them, if the obtained energy level index is greater than or equal to the first threshold, it is determined to be high risk; if the obtained energy level index is less than the first threshold and greater than or equal to the second threshold, it is determined to be medium risk; if the obtained energy level index is less than the second threshold, it is determined to be low risk.

Images