A method and system for extracting time difference of ultra-high frequency signal for local discharge positioning

Abstract

The present invention discloses a method and system for extracting time difference of ultra-high frequency signals for localizing partial discharge, and belongs to the technical field of localizing partial discharge of power equipment. It includes: obtaining partial discharge signals collected by two ultra-high frequency sensors at N different times for the same discharge source, and taking the partial discharge signals collected at the same time as a group of signals to obtain N groups of ultra-high frequency partial discharge signals; calculating the correlation function within the optimal time window for each group of ultra-high frequency partial discharge signals; calculating the cumulative results of the correlation functions of ultra-high frequency partial discharge signals within the N groups of optimal time windows; determining the maximum value of the cumulative result within the range of [‑Δd / c, Δd / c], and the horizontal coordinate corresponding to the maximum value is the time difference of the N groups of ultra-high frequency partial discharge signals. The present invention performs correlation analysis on the partial discharge signals measured by the two ultra-high frequency sensors within the optimal small window range, and superimposes the correlation results of the N groups of signals measured at different times, so as to minimize the influence of noise on the correlation, thereby obtaining a time difference with higher accuracy.

Description

Technical Field

[0001] The present invention belongs to the technical field of partial discharge positioning of electric power equipment, and more specifically, relates to a method and system for extracting time difference of ultra-high frequency signals for partial discharge positioning. Background Art

[0002] Partial discharge (PD) is a discharge phenomenon that occurs locally when insulation defects occur inside electrical equipment. Most faults in power equipment are caused by partial discharge caused by internal insulation defects, which continues to intensify and develop under operating voltage, eventually causing the loss of overall insulation performance of the equipment.

[0003] Ultra High Frequency (UHF) detection technology is a method for PD fault diagnosis of power equipment that has developed rapidly in recent years. It has the advantages of strong anti-interference ability, wide detection bandwidth and high sensitivity. In view of the many advantages of UHF detection technology, UHF method is widely used in the positioning of PD in high-voltage electrical equipment such as GIS and transformers.

[0004] The basic principle of the UHF positioning method is to use a sensor array composed of four or more UHF sensors to detect and obtain PD information, and to locate the PD source by calculating the signal delay relationship between each sensor and the reference sensor (select any one of them) and calculating through the positioning algorithm. When obtaining time difference information, since the propagation speed of electromagnetic waves is close to the speed of light, the time difference obtained is in the nanosecond level. There are extremely stringent accuracy requirements for the extracted time difference. A very small time difference error will also cause a large error in the subsequent positioning results or even make it impossible to solve the positioning result. However, traditional time difference extraction algorithms such as the initial peak method have the disadvantage that the peak point is difficult to judge; the generalized correlation method has the disadvantage that the time difference extraction effect is very poor when the signal-to-noise ratio is low. Summary of the invention

[0005] In view of the defects of the prior art, the purpose of the present invention is to provide a method and system for extracting the time difference of ultra-high frequency signals for localizing partial discharge, aiming to solve the problem of low accuracy of the existing time difference extraction methods.

[0006] To achieve the above object, in a first aspect, the present invention provides a method for extracting a time difference of a partial discharge positioning ultra-high frequency signal, comprising:

[0007] S1. Get two UHF sensors for the same discharge source The partial discharge signals collected at different times are taken as a group of signals, and the partial discharge signals collected by two UHF sensors at the same time are taken as a group of signals. Group UHF partial discharge signal;

[0008] S2. For each group of UHF partial discharge signals, calculate the correlation function of the partial discharge signals collected by two UHF sensors within the optimal time window;

[0009] S3. Calculation The cumulative result of the correlation function of the UHF PD signal within the group's optimal time window;

[0010] S4. Determine the cumulative result in The maximum value in the range, the horizontal coordinate corresponding to the maximum value is The time difference of the UHF partial discharge signal of the group;

[0011] in, Indicates the distance between two UHF sensors; It represents the propagation speed of electromagnetic waves in the medium. .

[0012] Preferably, in step S2, the optimal time window is determined by:

[0013] (1) Region As a time window, the cross-correlation between the PD signals collected by the two UHF sensors is calculated;

[0014] (2) Calculation Stability of the cross-correlation function at different values ;

[0015] (3) Fitting and The relationship curve of

[0016] (4) Determination The inflection point of the curve is determined The best value of and the best time window length for correlation analysis;

[0017] in, express The horizontal axis corresponding to the peak value is Indicates UHF sensor The collected Group UHF signal, , used to control the time window length, Indicates the minimum time scale for collecting UHF partial discharge signals.

[0018] It should be noted that the present invention calculates the corresponding stability function for the cross-correlation function under different time window lengths, and determines the optimal time window through the inflection point of the stability function. The optimal time window can minimize the influence of random disturbance on the oscillation of the cross-correlation function and the influence of the signal-to-noise ratio on the correlation, thereby solving the shortcomings of the generalized correlation method in the traditional time difference algorithm that is greatly affected by the signal-to-noise ratio and the initial peak method that is difficult to judge, and achieving high-precision time difference calculation accuracy.

[0019] Preferably, the calculation formula of the correlation function of the partial discharge signals collected by two UHF sensors within the time window is as follows:

[0020]

[0021] in, The time window length is , time shift is The correlation between the two UHF PD signals at express The horizontal axis corresponding to the peak value is , UHF sensor , The collected The group sends a signal. , used to control the time window length, Indicates the minimum time scale for collecting UHF partial discharge signals.

[0022] Preferably, the stability of the cross-correlation function The calculation formula is as follows:

[0023]

[0024] in, The time window length is , time shift is Correlation of two UHF partial discharge signals.

[0025] It should be noted that the present invention defines △R² and determines the stability of the cross-correlation function according to the amplitude of △R². This variable can quantitatively describe the influence of random oscillation on the cross-correlation function. The optimal time window length is determined by determining the inflection point of the △R² curve, thereby avoiding the situation where the time window length is too long, resulting in the correlation result being greatly affected by noise, or the time window length is too short, resulting in the correlation result being affected by random oscillation and the time difference cannot be determined.

[0026] Preferably, The value range is .

[0027] It should be noted that, considering that the frequency range of the UHF signal is 300MHz~1000MHz, and according to Shannon sampling theorem, the sampling rate should be more than twice the highest frequency of the signal, the present invention preferably uses 0.5ns as Taking into account the upper limit of the sampling frequency of the signal acquisition device, the present invention preferably uses 0.05ns as The lower limit of .

[0028] Preferably, .

[0029] It should be noted that, the larger the value of N is, the more stable the correlation results will be. However, in actual measurement, the value will not be too large in order to reduce the measurement time. The present invention preferably uses 10 as the upper limit of N.

[0030] To achieve the above-mentioned object, in a second aspect, the present invention provides a partial discharge positioning ultra-high frequency signal time difference extraction system, comprising: a processor and a memory;

[0031] The memory is used to store computer programs or instructions;

[0032] The processor is used to execute the computer program or instruction in the memory, so that the time difference extraction method described in the first aspect is executed.

[0033] In general, the above technical solution conceived by the present invention has the following beneficial effects compared with the prior art:

[0034] The present invention performs correlation analysis on the waveforms of partial discharge signals measured by two UHF sensors within a specific small window range, and The correlation results of the UHF PD signals measured at different times are analyzed. The superposition is repeated to minimize the influence of noise on the correlation, thereby obtaining a highly accurate UHF PD signal time difference. This method can effectively extract the time difference of noisy UHF PD signals in actual projects with high accuracy and little influence from the signal-to-noise ratio. It is of great significance to solve the problem of difficulty in extracting the time difference of noisy UHF PD signals in actual projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of a method for extracting time difference of ultra-high frequency signal for local discharge positioning provided by the present invention;

[0036] Figure 2 It is a waveform diagram of a UHF partial discharge signal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0038] Figure 1 This is a flow chart of a method for extracting the time difference of a partial discharge positioning ultra-high frequency signal provided by the present invention. Figure 1 As shown, the method comprises the following steps:

[0039] Step S1. Obtain two UHF sensors for the same discharge source. The partial discharge signals collected at different times are taken as a group of signals, and the partial discharge signals collected by two UHF sensors at the same time are taken as a group of signals. Group of UHF partial discharge signals.

[0040] By making typical insulation defect specimens commonly found in large-scale electrical equipment in actual projects, a partial discharge platform that can simulate actual large-scale electrical equipment is built in the laboratory. Under test conditions that simulate the field environment as much as possible, ultra-high frequency partial discharge signals that can simulate the actual project site are obtained, and N groups of ultra-high frequency partial discharge signals are measured at different times.

[0041] Step S2: For each group of ultra-high frequency partial discharge signals, a correlation function of the partial discharge signals collected by two ultra-high frequency sensors within an optimal time window is calculated.

[0042] In order to minimize the influence of noise on the cross-correlation of UHF PD signals, the correlation analysis is performed in the area near the peak of UHF PD signals. UHF partial discharge signal , ,set up The horizontal coordinates corresponding to the peak values ​​are , in nanoseconds), then select The area is used as a small window for cross-correlation analysis, where is the minimum time scale for the device to collect UHF PD signals, n=0,1,2,…; Group signal, a discrete sequence of partial discharge signals received by two UHF sensors , , and its cross-correlation function can be expressed as:

[0043] (1)

[0044] In space, the placement of UHF sensors S1, S2 and the discharge source S forms a triangle or a straight line. According to the principle that the difference between the two sides of the triangle is less than the third side, the time difference can be obtained. Should meet:

[0045] (2)

[0046] in, is the distance between the two UHF sensors; is the propagation speed of electromagnetic waves in the medium, then τ∈ Conduct relevant analysis within the scope.

[0047] In order to quantitatively describe the relevant function description , The correlation is disturbed by random oscillations, and the stability of the cross-correlation function is defined as:

[0048] (3)

[0049] along with As the time window length increases, the influence of random oscillation on the cross-correlation function gradually offsets, and the stability of the cross-correlation function gradually increases. However, the increase in the time window length will also be accompanied by the introduction of intervals with low signal-to-noise ratio, which will affect the accuracy of the cross-correlation function. and The relationship curve of The inflection point of the curve is determined The best value of and the best time window length for correlation analysis. The inflection point can be determined by any method in the prior art, which will not be expanded here.

[0050] In τ∈ Calculate within range The value of The maximum value of determines the corresponding , get the The calculated value of the time difference corresponding to the group of UHF partial discharge signals.

[0051] Step S3. Calculation Cumulative results of the correlation function of the UHF PD signal within the group's optimal time window.

[0052] Since the time difference of calculating a single group of UHF PD signals may cause errors in the correlation results of UHF PD signals due to random correlation results of noise, when τ= The amplitude of the correlation function at will also be affected by the signal-to-noise ratio. The UHF PD signals measured at different times are superimposed by correlation functions, so that The group signal is at τ = The amplitude of the correlation curve at The accumulation of times becomes prominent, and The amplitude of random noise with group correlation will be reduced after superposition. The cross-correlation function of the group signal ,have:

[0053] (4)

[0054] The cumulative result of the correlation function of the group data is:

[0055] (5)

[0056] Step S4. Determine the cumulative result in The maximum value in the range, the horizontal coordinate corresponding to the maximum value is The time difference of a group of UHF partial discharge signals.

[0057] By determining In τ∈ The horizontal coordinate corresponding to the maximum value in the range is calculated The time difference of a group of UHF partial discharge signals.

[0058] It can be seen from this that when the distance between the two UHF sensors and different time numbers When the The time difference of the UHF partial discharge signals of the two groups is the same.

[0059] Example

[0060] Using the built partial discharge measurement signal test platform, two UHF partial discharge sample waveform data were collected. The distance difference between the two UHF sensors was set to =88cm (theoretical time delay is 2.933ns), collect 10 groups of sample data at different times. Use energy accumulation method, initial peak method, generalized correlation method, and small window cross-correlation accumulation algorithm proposed by the present invention to calculate the 10 groups of sample data measured at different times to obtain the measured partial discharge ultra-high frequency signal time difference. Among them, the ten groups of time difference results obtained by energy accumulation method, initial peak method, and generalized correlation method are averaged to obtain the average calculated time difference. The calculation results of each algorithm are shown in the following table:

[0061]

[0062] It can be seen from Table 1 that the small window cross-correlation accumulation method has the best time difference extraction effect, while the initial peak method, generalized correlation method, and energy accumulation method have poorer effects than the small window cross-correlation accumulation method. The main reasons are:

[0063] Figure 2Figure 1 is a waveform diagram of a UHF PD signal provided by an embodiment of the present invention, (a) is a complete UHF PD signal, and (b) is an amplified image of the area near the peak of the signal. Figure 2 The high-frequency oscillation shown makes it difficult to determine the true peak point of the signal;

[0064] When the signal-to-noise ratio of the UHF PD signal is low, the generalized correlation method is prone to cause the signal correlation to be weakened by the noise due to the irrelevance of the noise;

[0065] The energy accumulation method actually only relies on the first-order derivative or second-order derivative of the energy function between two adjacent points to determine the inflection point of the energy signal. The amount of information used to determine the arrival time is too little and it is easy to introduce errors.

[0066] The small window cross-correlation accumulation method proposed in the present invention selects an area near the peak of the UHF partial discharge signal for correlation analysis, which reduces the influence of noise on the correlation to the greatest extent. At the same time, selecting an interval of a certain length also avoids random oscillation errors caused by too little information in a single correlation analysis. Through multiple correlation accumulations, the peak of the correlation function at τ=△t is made more prominent.

[0067] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for extracting time difference of ultra-high frequency signal for local discharge positioning, characterized in that: include: S1. Get two UHF sensors for the same discharge source The partial discharge signals collected at different times are taken as a group of signals, and the partial discharge signals collected by two UHF sensors at the same time are taken as a group of signals. Group UHF partial discharge signal; S2. For each group of UHF partial discharge signals, calculate the correlation function of the partial discharge signals collected by two UHF sensors within the optimal time window; S3. Calculation The cumulative result of the correlation function of the UHF PD signal within the group's optimal time window; S4. Determine the cumulative result in The maximum value in the range, the horizontal coordinate corresponding to the maximum value is The time difference of the UHF partial discharge signal of the group; in, Indicates the distance between two UHF sensors; It represents the propagation speed of electromagnetic waves in the medium. ; In step S2, the optimal time window is determined by: (1) Region As a time window, the cross-correlation between the PD signals acquired by the two UHF sensors is calculated; (2) Calculation Stability function of cross-correlation function at different values ; (3) Fitting and The relationship curve of (4) Determination The inflection point of the curve is determined The best value of and the best time window length for correlation analysis; in, Respectively The horizontal axis corresponding to the peak value is Indicates UHF sensor The collected Group UHF signal, , used to control the time window length, Indicates the minimum time scale for collecting UHF partial discharge signals.

2. The method according to claim 1, characterized in that The calculation formula of the correlation function of two UHF PD signals in the time window is as follows: in, The time window length is , time shift is The correlation between the two UHF PD signals at , UHF sensor , The collected The group sends out signals.

3. The method according to claim 1, characterized in that Stability of the cross-correlation function The calculation formula is as follows: in, The time window length is , time shift is Correlation of two UHF partial discharge signals.

4. The method according to any one of claims 1 to 3, characterized in that The value range is .

5. The method according to claim 1, characterized in that 。 6. A partial discharge positioning ultra-high frequency signal time difference extraction system, characterized in that: include: Processor and memory; The memory is used to store computer programs or instructions; The processor is used to execute the computer program or instruction in the memory, so that the time difference extraction method according to any one of claims 1 to 5 is executed.

Images