An online fault detection method and system applicable to UHVDC DC filters

Abstract

The present invention provides an online fault detection method and system applicable to UHVDC DC filters. The method includes: when the relay protection detects that the DC filter is detuned, the fault recorder will perform fault recording and generate a recording file. According to the recording data related to the DC filter in the recording file, calculate the harmonic current ratio of each parallel branch after detuning, and by determining whether it exceeds the corresponding threshold, the parallel branch where the detuned element is located can be determined. Finally, the detuned element can be located by calculating the error between the theoretical and actual harmonic impedance offset amounts. The present invention solves the technical problems of low fault detection efficiency and low accuracy for the detuned fault elements in the damping type double-tuned DC filter in the prior art.

Description

Technical Field

[0001] The present invention relates to the technical field of control and protection of high voltage direct current transmission systems, and in particular to an online fault detection method and system suitable for UHVDC direct current filters. Background Art

[0002] In recent years, my country's ultra-high voltage direct current (UHVDC) transmission projects have developed rapidly. In the UHVDC system, the converter station is a high-power, nonlinear power conversion equipment. The nonlinear switching action of its high-power thyristor generates a large number of harmonics in the DC system and the AC system to which it is connected. At present, the main method to suppress the harmonics of the DC system is to install passive DC filters. Common passive filters include single-tuned filters, double-tuned filters, triple-tuned filters, and high-pass filters. Among them, double-tuned filters have good filtering characteristics and economic characteristics and are widely used in UHVDC systems.

[0003] The damped double tuned filter includes a high voltage capacitor C in series 1 , connected in parallel with R 1 and L 1 , connected in parallel with R 2 , L 2 and C 2 ,like Figure 1 As shown, the filter presents low impedance at the tuning frequency, and the harmonic current is filtered out through the filter channel. During the operation of UHVDC, changes in filter component parameters due to temperature changes or aging of DC filter components will cause changes in the tuning characteristics of the filter, causing the actual resonant frequency to deviate from the set value, thereby affecting the filter effect of the filter and further affecting the normal operation of the system. When the filter component parameters exceed the allowable range of variation, the detuning protection will be activated. In order to locate the components with slight changes in parameters in the filter, all components in the filter still need to be tested after the detuning protection is activated, which is labor-intensive and cumbersome. Since the probability of slight changes in the parameters of a single component of the filter is the highest in actual engineering, the gradual change of the parameters of a single component in the DC filter causes a detuning fault. If the faulty component can be detected quickly and accurately while the filter detuning protection is activated, the repair time of the filter will be greatly shortened and its availability will be improved.

[0004] However, most of the current research focuses only on high-voltage capacitors C 1The existing fault location based on rough sets depends on the selection of sample sets and the division of attribute space. The algorithm is complex and not easy to implement in engineering. The detection method based on impedance offset ratio is suitable for three-tuned filters, but because the difference in impedance offset at the tuning point of the double-tuned filter is not large, the ratio will overlap and cannot be accurately identified. The document "A method for detecting detuned fault elements in three-tuned DC filters" with patent publication number CN102401866B mainly includes the following steps: first, calculate the harmonic impedance offset ratio of the three-tuned filter; then judge the detuned element according to the set offset ratio range and the harmonic impedance characteristics after detuning. Therefore, it is necessary to propose a double-tuned filter fault location method with a simple algorithm and reliable application. Summary of the invention

[0005] The technical problem to be solved by the present invention is the problem of low fault detection efficiency and low accuracy of a detuned fault element in a damped double-tuned DC filter.

[0006] The present invention adopts the following technical solution to solve the above technical problem: an online fault detection method applicable to UHVDC DC filter comprises:

[0007] S1. Compare the tail harmonic currents of two adjacent poles to determine whether the DC filter is detuned, generate a COMTRADE fault recording file based on the comparison, and obtain recording data from the COMTRADE fault recording file;

[0008] S2, perform abnormal detection and repair on the recorded wave data, and calculate the parallel element R according to the recorded wave data 1 / L 1 , R 2 / L 2 , R 2 / C 2 The harmonic current ratio of the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Whether the current ratio exceeds a preset threshold, step S2 comprises:

[0009] S21, collects the detuned flow through R 1 , R 2 , L 1 , L 2 , C 2 The instantaneous value data of the current is selected, n different time points are selected, and the mth harmonic current value is obtained by Fourier decomposition, and the harmonic current ratio at the n time points is calculated by using a preset logic according to the harmonic current value;

[0010] S22, determine the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Harmonic current ratio threshold;

[0011] S23, determining the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Whether the current ratio exceeds the harmonic current ratio threshold;

[0012] S3, if R 1 / L 1 The current exceeds the preset threshold, according to the R 1 / L 1 Harmonic current ratio calculation R 1 , L 1 The corresponding change is calculated as R 1 , L 1 The harmonic impedance offset error under the corresponding change is used to determine R 1 / L 1 The step S3 comprises:

[0013] S31, obtain a harmonic impedance ratio according to the harmonic current ratio, and calculate the harmonic impedance ratio to obtain L 1 , R 1 The amount of change;

[0014] S32. Calculate R 1 , L 1 The mth order theoretical harmonic impedance value corresponding to the corresponding change amount is subtracted from the theoretical nominal value by using preset logic, and then the current actual mth order harmonic impedance value is obtained according to the harmonic voltage-current ratio, and the actual normal value is subtracted to obtain the mth order harmonic impedance value;

[0015] S4. If the current ratio of R1 / L1 does not exceed the preset threshold, the faulty element among C1 / R2 / L2 / C2 is obtained according to the preset fault determination logic.

[0016] After the relay protection detects the detuning of the DC filter, the present invention performs fault recording through a fault recorder and generates a recording file, calculates the harmonic current ratio of each parallel branch after the detuning according to the recording data related to the DC filter in the recording file, and determines whether it exceeds the corresponding threshold value, so as to determine the parallel branch where the detuning element is located, and finally locates the detuning element by calculating the theoretical and actual harmonic impedance offset errors.

[0017] The present invention is based on R 1 / L 1 Harmonic current ratio calculation R 1 , L 1 The corresponding change is used to obtain the harmonic impedance offset error under the corresponding change, and quickly determine R 1 / L 1 The faulty element in the double-tuned filter is located, which avoids the ratio overlap caused by the small difference in impedance offset at the tuning point of the double-tuned filter in the traditional fault location technology, thereby improving the fault identification accuracy.

[0018] The invention is based on on-site wave recording data, has a simple detection process, and has clear physical meaning. It can quickly and accurately detect detuned fault components in a damped double-tuned DC filter online, and has certain engineering application value.

[0019] In a more specific technical solution, step S1 includes:

[0020] S11. Compare the tail harmonic currents of two adjacent poles with the following relay protection action equation to determine whether the DC filter is detuned:

[0021] IZxT2_OP_m>k*IZxT2_m and IZxT2_OP>Iset

[0022] Wherein, x represents the DC filter group, IZxT2_m is the mth harmonic current amplitude of the DC filter tail current of this pole, IZxT2_OP_m is the mth harmonic current amplitude of the DC filter tail current of the adjacent other pole, IZxT2_OP is the adjacent other pole DC filter tail current amplitude, Iset is the set value, and k is the detuning monitoring coefficient;

[0023] S12, detecting and reading a COMTRADE fault recording file, and transferring the recording information in the COMTRADE fault recording file to a preset program operation variable;

[0024] S13, intercepting the voltage and current waveforms of the DC filter in a specific period after the fault from the COMTRADE fault recording file;

[0025] S14. Generate the waveform recording data according to the voltage and current waveform recording waveforms.

[0026] The present invention detects whether a COMTRADE recording file is generated on a file server, specifically including a .cfg file and a .data file. By using the above method, information stored in the COMTRADE text in binary format is transferred to variables that are convenient for program operation, thereby improving the online detection efficiency of detuned fault components in a damped double-tuned DC filter.

[0027] In a more specific technical solution, step S21 includes:

[0028] The instantaneous current value data is collected to obtain the harmonic current ratios at the n time points, and the n harmonic current ratios are averaged using the following logic:

[0029]

[0030] in, They represent the final values ​​of the mth harmonic current ratios of the R1 / L1, R2 / L2, and R2 / C2 branches of this pole, They respectively represent the ratio of the effective values ​​of the mth harmonic currents of the R1 / L1, R2 / L2, and R2 / C2 branches of the local pole obtained by Fourier decomposition at the time point ti of the recorded data, i = 1, 2, ..., n.

[0031] In a more specific technical solution, step S22 includes:

[0032] S221, obtaining component nominal values;

[0033] S222, R calculated based on the nominal value of the component 1 / L 1 , R 2 / L 2 , R 2 / C 2 Standard value of mth harmonic impedance ratio of parallel branch;

[0034] S223. According to the standard value, obtain the harmonic current ratio threshold value by the following logic processing.

[0035] In a more specific technical solution, in step S222, the R calculated according to the nominal value of the component by the following logic 1 / L 1 , R 2 / L 2 , R 2 / C 2 Standard value of the mth harmonic impedance ratio of the parallel branch:

[0036]

[0037] Where Z R1 , Z R2 , They are the corresponding components at frequency f m The AC impedance value at m is the frequency of the mth harmonic, R 1 , R 2 , L1 , L 2 , C 2 are the nominal values ​​of the components respectively.

[0038] The present invention is not only aimed at high voltage capacitor C 1 The fault is detected and R 1 , R 2 , L 1 , L 2 , C 2 The present invention does not rely on the selection of sample sets and the division of attribute space, reduces the complexity of the algorithm, is easy to implement in engineering, and improves the applicability of the fault detection technology.

[0039] In a more specific technical solution, in step S223, the standard value is processed by the following logic to obtain the harmonic current ratio threshold value:

[0040]

[0041]

[0042] Wherein, bottom_x and top_x are the lower and upper boundaries of the harmonic current ratio threshold, respectively, x={RL1, RL2, RC2}, k 1 =min{k R1 , k L1}, k 2 =min{k R2 , k L2 , k C2}, ε i (i=1,2,3) is a small positive constant

[0043] In a more specific technical solution, in step S31, the harmonic impedance ratio is calculated using the following logic to obtain L 1 , R 1 The amount of change:

[0044]

[0045] Among them, k R1 , k L1 They are the R corresponding to the harmonic current ratio after detuning 1 and L 1 The amount of change;

[0046] In a more specific technical solution, in step S32, the following logic is used to make a difference between the mth order theoretical harmonic impedance value and the actual normal value to obtain the mth order harmonic impedance value:

[0047]

[0048] in, and R 1 , L 1 In the corresponding change k R1 , k L1 The value of the mth harmonic impedance obtained by theoretical calculation deviates from the nominal value. is the nominal value of the mth harmonic impedance of the DC filter, and R 1 and L 1 When the change is k R1 , k L1 The corresponding mth harmonic impedance theoretical value is: is the deviation of the actual calculated mth harmonic impedance value from the normal value, and are the mth harmonic voltage effective values ​​at both ends of the DC filter at the local pole and the relative pole, and are the effective values ​​of the mth harmonic current flowing through the first end of the DC filter at the local pole and the relative pole, respectively, and error R1 and error L1 R 1 , L 1 Corresponding theoretical and actual harmonic impedance offset errors.

[0049] In a more specific technical solution, step S4 includes:

[0050] S41, in R 2 / L 2 , R 2 / C 2 When the current ratio of the harmonic current ratio exceeds the harmonic current ratio threshold at the same time, it is determined that the faulty element is R 2 ;

[0051] S42, in R 2 / L 2 The current ratio exceeds the harmonic current ratio threshold and R 2 / C 2 When the current ratio does not exceed the harmonic current ratio threshold, the fault element is determined to be L 2 ;

[0052] S43, in R 2 / C 2 The current ratio exceeds the harmonic current ratio threshold and R 2 / L 2 When the current ratio does not exceed the harmonic current ratio threshold, the fault element is C 2 ;

[0053] S44, in the R 2 / L 2 The current ratio, the R 2 / C 2 When the current ratios are both within the threshold, it is determined that the faulty element is C 1 .

[0054] After the detuning protection is activated, the present invention detects the components in the filter to locate the components whose parameters have slightly changed. When a detuning fault is caused by a gradual change in the parameters of a single component in the DC filter, the present invention can quickly and accurately detect and determine the faulty component while the filter detuning protection is activated, greatly shortening the filter repair time and improving its availability.

[0055] In a more specific technical solution, an online fault detection system for a UHVDC filter includes:

[0056] A recording data acquisition module is used to compare the tail end harmonic currents of two adjacent poles to determine whether the DC filter is detuned, generate a COMTRADE fault recording file based on the comparison, and obtain recording data from the COMTRADE fault recording file;

[0057] The parallel element current ratio judgment module is used to detect and repair the abnormality of the recorded wave data, and calculate the parallel element R according to the recorded wave data. 1 / L 1 , R 2 / L 2 , R 2 / C 2 The harmonic current ratio of the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Whether the current ratio of the parallel element exceeds a preset threshold, the parallel element current ratio judgment module is connected to the wave recording data acquisition module, and the parallel element current ratio judgment module includes:

[0058] The harmonic current ratio module is used to collect the detuned current flowing through R 1 , R 2 , L 1 , L 2 , C 2 The instantaneous value data of the current is selected, n different time points are selected, and the mth harmonic current value is obtained by Fourier decomposition, and the harmonic current ratio at the n time points is calculated by using a preset logic according to the harmonic current value;

[0059] The harmonic current ratio threshold module is used to determine the parallel element R1 / L 1 , R 2 / L 2 , R 2 / C 2 Harmonic current ratio threshold;

[0060] A threshold judgment module is used to judge the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Whether the current ratio exceeds the harmonic current ratio threshold, the current ratio judgment module is connected to the harmonic current ratio threshold module and the harmonic current ratio module;

[0061] R 1 / L 1 Fault component determination module, used to 1 / L 1 When the current ratio exceeds the preset threshold, the R 1 / L 1 Harmonic current ratio calculation R 1 , L 1 The corresponding change is calculated as R 1 , L 1 The harmonic impedance offset error under the corresponding change is used to determine R 1 / L 1 The fault element in the parallel element, the fault element determination module is connected to the parallel element current ratio judgment module, and the fault element determination module includes:

[0062] A harmonic impedance ratio processing module is used to obtain a harmonic impedance ratio according to the harmonic current ratio processing, and calculate the harmonic impedance ratio to obtain L 1 , R 1 The amount of change;

[0063] Harmonic impedance value acquisition module, used to calculate R 1 , L 1 The mth order theoretical harmonic impedance value corresponding to the corresponding variation is subtracted from the theoretical nominal value by using preset logic, and then the current actual mth order harmonic impedance value is obtained according to the harmonic voltage-current ratio, and the actual normal value is subtracted to obtain the mth order harmonic impedance value, and the harmonic impedance value acquisition module is connected to the harmonic impedance ratio processing module;

[0064] The C1 / R2 / L2 / C2 fault component determination module is used to obtain the fault component in C1 / R2 / L2 / C2 according to the preset fault determination logic processing when the R1 / L1 current ratio does not exceed the preset threshold value. The C1 / R2 / L2 / C2 fault component determination module is connected to the parallel element current ratio judgment module.

[0065] Compared with the prior art, the present invention has the following advantages: after the relay protection detects that the DC filter is detuned, the present invention uses a fault recorder to record the fault and generate a recording file, calculates the harmonic current ratio of each parallel branch after the detuning according to the recording data related to the DC filter in the recording file, and determines whether it exceeds the corresponding threshold value, so as to determine the parallel branch where the detuned element is located, and finally locates the detuned element by calculating the theoretical and actual harmonic impedance offset errors.

[0066] The present invention is based on R 1 / L 1 Harmonic current ratio calculation R 1 , L 1 The corresponding change is used to obtain the harmonic impedance offset error under the corresponding change, and quickly determine R 1 / L 1 The faulty element in the double-tuned filter is located, which avoids the ratio overlap caused by the small difference in impedance offset at the tuning point of the double-tuned filter in the traditional fault location technology, thereby improving the fault identification accuracy.

[0067] The invention is based on on-site wave recording data, has a simple detection process, and has clear physical meaning. It can quickly and accurately detect detuned fault components in a damped double-tuned DC filter online, and has certain engineering application value.

[0068] The present invention detects whether a COMTRADE recording file is generated on a file server, specifically including a .cfg file and a .data file. By using the above method, information stored in the COMTRADE text in binary format is transferred to variables that are convenient for program operation, thereby improving the online detection efficiency of detuned fault components in a damped double-tuned DC filter.

[0069] The present invention is not only aimed at high voltage capacitor C 1 The fault is detected and R 1 , R 2 , L 1 , L 2 , C 2 The present invention does not rely on the selection of sample sets and the division of attribute space, reduces the complexity of the algorithm, is easy to implement in engineering, and improves the applicability of the fault detection technology.

[0070] After the detuning protection is activated, the present invention detects the components in the filter to locate the components whose parameters have slightly changed in the filter. When a detuning fault is caused by a gradual change in the parameters of a single component in the DC filter, the present invention can quickly and accurately detect and determine the faulty component while the filter detuning protection is activated, greatly shortening the filter repair time and improving the filter availability. The present invention solves the technical problem of low fault detection efficiency and low accuracy for detuning faulty components in damped double-tuned DC filters in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a schematic diagram of parallel resistor connection of a damped double tuned filter in the background technology;

[0072] Figure 2 This is a schematic diagram of the steps of an online fault detection method applicable to a UHVDC direct current filter according to Embodiment 1 of the present invention;

[0073] FIG. 3( a ) is a diagram of R in Example 2 of the present invention. 1 / L 1 Schematic diagram of the corresponding first harmonic current ratio change;

[0074] FIG. 3( b ) is a diagram of R in Example 2 of the present invention. 2 / L 2 Schematic diagram of the corresponding second harmonic current ratio change;

[0075] FIG. 3( c ) is a diagram of R in Example 2 of the present invention. 2 / C 2 Schematic diagram of the corresponding third harmonic current ratio change;

[0076] FIG. 4( a ) is a diagram of R in Example 2 of the present invention. 1 / L 1 Schematic diagram of the corresponding first harmonic current ratio change;

[0077] FIG. 4( b ) is a diagram of R in Example 2 of the present invention. 2 / L 2 Schematic diagram of the corresponding second harmonic current ratio change;

[0078] FIG. 4( c ) is a diagram of R in Example 2 of the present invention. 2 / C 2 Schematic diagram of the corresponding third harmonic current ratio change;

[0079] FIG4(d) is a schematic diagram showing the difference between the theoretical calculation and the actual calculation of the 12th harmonic impedance offset according to Embodiment 2 of the present invention under different variation amounts;

[0080] FIG. 5( a ) is a diagram of R in Example 2 of the present invention. 1 / L1 Schematic diagram of the corresponding first harmonic current ratio change;

[0081] FIG. 5( b ) is a diagram of R in Example 2 of the present invention. 2 / L 2 Schematic diagram of the corresponding second harmonic current ratio change;

[0082] FIG. 5( c ) is a diagram of R in Example 2 of the present invention. 2 / C 2 Schematic diagram of the corresponding third harmonic current ratio change;

[0083] FIG5(d) is a schematic diagram showing the difference between the error of the 12th harmonic impedance offset obtained by theoretical calculation and actual calculation in Example 2 of the present invention under different variation amounts;

[0084] FIG. 6( a ) is a diagram of R in Example 2 of the present invention. 1 / L 1 Schematic diagram of the corresponding first harmonic current ratio change;

[0085] FIG. 6( b ) is a diagram of R in Example 2 of the present invention. 2 / L 2 Schematic diagram of the corresponding second harmonic current ratio change;

[0086] FIG. 6( c ) is a diagram of R in Example 2 of the present invention. 2 / C 2 Schematic diagram of the corresponding third harmonic current ratio change;

[0087] FIG. 7( a ) is a diagram of R in Example 2 of the present invention. 1 / L 1 Schematic diagram of the corresponding first harmonic current ratio change;

[0088] FIG. 7( b ) is a diagram of R in Example 2 of the present invention. 2 / L 2 Schematic diagram of the corresponding second harmonic current ratio change;

[0089] FIG. 7( c ) is a diagram of R in Example 2 of the present invention. 2 / C 2 Schematic diagram of the corresponding third harmonic current ratio change;

[0090] FIG8(a) is a diagram of R in Example 2 of the present invention. 1 / L 1 Schematic diagram of the corresponding first harmonic current ratio change;

[0091] FIG8(b) is a diagram of R in Example 2 of the present invention. 2 / L 2 Schematic diagram of the corresponding second harmonic current ratio change;

[0092] FIG8(c) is a diagram of R in Example 2 of the present invention. 2 / C 2 Schematic diagram of the corresponding third harmonic current ratio change. DETAILED DESCRIPTION

[0093] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0094] Example 1

[0095] like Figure 1 As shown, an online fault detection method applicable to a UHVDC DC filter comprises the following steps:

[0096] S1. Determine whether the DC filter is detuned. In this embodiment, determine whether the DC filter is detuned according to the relay protection action equation. The relay protection of the DC filter identifies the slight changes of the DC filter by comparing the tail harmonic currents of the two adjacent poles. The action equation is as follows: IZxT2_OP_m>k*IZxT2_m and IZxT2_OP>Iset, where x is 1 or 2, indicating DC filter group 1 and group 2, IZxT2_m is the mth harmonic current amplitude of the tail current of the DC filter of this pole, IZxT2_OP_m is the mth harmonic current amplitude of the tail current of the adjacent other-pole DC filter, IZxT2_OP is the tail current amplitude of the adjacent other-pole DC filter, Iset is the set value, k is the detuning monitoring coefficient, and when it is detected that the action equation is satisfied, it is considered that the DC filter has a detuning fault, and this is used as the starting condition for the subsequent steps;

[0097] S2. Read the recording file. In this embodiment, the COMTRADE fault recording file is detected and read. In this embodiment, it is necessary to detect whether the file server has a COMTRADE recording file generated, which specifically includes a .cfg file and a .data file. The purpose of this step is to transfer the information stored in the COMTRADE text in binary to a variable that is convenient for the program to operate. However, the specific storage variable used should be adapted to the software language for subsequent analysis and is not limited to a specific format.

[0098] S3, intercepting the waveform data sample. In this embodiment, the waveforms of the voltage and current waveforms of the DC filter for a period of time after the fault are intercepted as the waveform data sample. The waveform data related to the DC filter are collected, including: the voltage UDL and UDL_OP at both ends of the DC filter at the current pole and the opposite pole, the current R 1 , R 2 , L 1 , L 2 The branch currents IZ1R1, IZ1R2, IZ1L1, and IZ1L2, the DC filter tail currents IZ1T2 and IZ1T2_OP flowing through the pole and the opposite pole, a total of 8 recording channel data, the currents flowing through the pole C 2 The branch current is calculated by the node current law, which is recorded as IZ1C2 = IZ1T2-IZ1R2-IZ1L2. The recorded data of a certain power grid cycle after the fault is intercepted and used as the data sample of the DC filter detuning. The specific fault start time is based on the protection action equation of step (1);

[0099] S4. Abnormal data detection and repair. In this embodiment, abnormal data detection and repair are performed on the recorded data. Abnormal recorded data detection is achieved through the upper and lower limits of the analog channel of the recorded .cfg file. The analog channel line information format is as follows:

[0100] An,ch_id,ph,ccbm,uu,a,b,skew,min,max,primary,secondary,PS <cr lf>

[0101] Among them, a and b are the gain coefficient and offset mentioned above; min and max are the lower and upper limits of the analog channel storage value in the .dat data file. Therefore, the maximum and minimum limits of a certain analog channel are a*max+b and a*min+b respectively. If the recorded data point exceeds the limit, it needs to be repaired by regression fitting or other methods;

[0102] S5, calculate the harmonic current ratios of R1 / L1, R2 / L2, and R2 / C2. In this embodiment, the parallel element R is calculated based on the recorded data. 1 / L 1 , R 2 / L 2 , R 2 / C 2 The harmonic current ratio is collected and flows through R after detuning 1 , R 2 , L 1 , L 2 , C 2 The instantaneous current value data of the current is obtained by selecting n different time points and Fourier decomposition to obtain the mth harmonic current value. The harmonic current ratio at the n time points is calculated, and finally the obtained n harmonic current ratios are averaged. The calculation formula of the harmonic current ratio is expressed as follows:

[0103]

[0104] in, Respectively represent the pole R 1 / L 1 , R 2 / L 2 , R 2 / C 2 The final value of the branch mth harmonic current ratio is: The respective poles represent R 1 / L 1 , R 2 / L 2 , R 2 / C 2 The branch is at the recording data time point t i The ratio of the effective value of the mth harmonic current is obtained by Fourier decomposition, i = 1, 2, ..., n;

[0105] S6. Determine the current ratio threshold of the parallel element. In this embodiment, determine the current ratio threshold of the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Harmonic current ratio threshold, harmonic current ratio is equal to harmonic impedance ratio, R calculated from the nominal value of the component 1 / L 1 , R 2 / L 2 , R 2 / C 2 The nominal value of the mth harmonic impedance ratio of the parallel branch is:

[0106]

[0107] Where Z R1 , Z R2 , They are the corresponding components at frequency f m The AC impedance value at m is the frequency of the mth harmonic, R 1 , R 2 , L 1 , L 2 , C 2 are the nominal values ​​of the components respectively. Under normal circumstances, the harmonic current ratio is equal to the above-mentioned nominal value of the harmonic impedance ratio. If a component changes, the harmonic current ratio related to the component will deviate from the nominal value as the harmonic impedance ratio changes, while the remaining harmonic current ratios are still equal to the nominal value. Therefore, the detuned parallel branch can be preliminarily judged. However, since the process of obtaining harmonic currents by measurement and Fourier decomposition will cause errors in the size of the obtained harmonic current ratio, the size of the harmonic current ratio is not strictly equal to the harmonic impedance ratio, and a threshold needs to be determined to avoid this error. Obviously, the larger the threshold, the stronger the ability of the harmonic current ratio of the normal parallel branch to resist error interference, but the lower the detection sensitivity of the harmonic current ratio of the faulty branch; the smaller the threshold, the smaller the ability to resist error interference, but the higher the detection sensitivity.

[0108] A threshold determination method that takes into account both anti-interference ability and sensitivity and facilitates subsequent judgment is as follows: When a DC filter component undergoes a small micro-change, the DC filter performance does not change much. At this time, the action equation in step S1 is not satisfied, and the DC filter is still considered normal. The micro-change range corresponding to the component is the allowable error of the component, which is determined by the action equation in step S1. Assume that the allowable errors corresponding to the components are: ±k R1 ±k R2 ±k L1 ±k L2 ±k C2 , then the method for determining the mth harmonic current ratio threshold is as follows:

[0109]

[0110] Where bottom_x and top_x are the lower and upper boundaries of the harmonic current ratio threshold, respectively, x = {RL1, RL2, RC2}, k 1 =min{k R1 , k L1 }, k 2 =min{k R2 , k L2 , k C2 }, ε i (i=1,2,3) is a small normal number, which is adjusted according to the actual situation;

[0111] S7, determine whether the current ratio of R1 / L1 exceeds the threshold. In this embodiment, determine whether the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 Whether the current ratio exceeds the threshold value, the judgment method of whether it exceeds the threshold value is as follows:

[0112] like Then R 1 / L 1 The current ratio of the branch exceeds the threshold, otherwise it is normal.

[0113] like Then R 2 / L 2 The current ratio of the branch exceeds the threshold, otherwise it is normal.

[0114] like Then R 2 / C 2 The current ratio of the branch exceeds the threshold, otherwise it is normal;

[0115] S8. If the current ratio of R1 / L1 exceeds the threshold, the corresponding changes of R1 and L1 are calculated. 1 / L 1 The current ratio of exceeds the threshold, indicating that the parameters of one of the components have changed. If the harmonic current ratio is known, the harmonic impedance ratio can be inferred from its size. Since the harmonic impedance ratio is related to the parameters of the components, it can be assumed that R 1 , L 1 Normal, calculate L 1 , R 1 In this embodiment, R 1 and L 1 The formula for calculating the change in is as follows:

[0116]

[0117] where k R1 , k L1 They are the R corresponding to the harmonic current ratio after detuning. 1 and L 1 The amount of change;

[0118] S9, calculate the theoretical and actual harmonic impedance offset errors of R1 and L1. In this embodiment, the calculation formula of the theoretical and actual harmonic impedance offset errors is as follows:

[0119]

[0120] and R 1 , L 1 In the corresponding change k R1 , k L1 The value of the mth harmonic impedance obtained by theoretical calculation deviates from the nominal value. is the nominal value of the mth harmonic impedance of the DC filter, and R 1 and L 1 When the change is k R1 , k L1 The corresponding mth harmonic impedance theoretical value is: is the deviation of the actual calculated mth harmonic impedance value from the normal value, and are the mth harmonic voltage effective values ​​at both ends of the DC filter at the local pole and the relative pole, and are the mth harmonic current effective value flowing through the end of the DC filter at the local pole and the relative pole, respectively, and error R1 and error L1 R 1 , L 1 The corresponding theoretical and actual harmonic impedance offset errors. In this embodiment, because R 1 and L 1 The impact on the harmonic impedance of the DC filter is not exactly the same, so the harmonic impedance value at this time can be used to determine that the faulty component is R 1 or L 1 Considering that the actual situation will not be completely consistent with the theoretical analysis, the difference is considered to reduce the error. The specific method is: calculate the 1 , L 1 The mth order theoretical harmonic impedance value corresponding to the change in step (9) is obtained, and the difference is made between the value and the theoretical nominal value; then, the actual mth order harmonic impedance value at this time is obtained according to the harmonic voltage-current ratio, and the difference is made between the value and the actual normal value, and the actual normal value is the mth order harmonic impedance value obtained by relative pole calculation;

[0121] S10, determine whether the R1 error is less than the L1 error. In this embodiment, determine whether the R1 error is less than the L1 error. 1 / L 1 The judgment logic of the faulty component is:

[0122] If error R1 ≤error L1 , then R 1 is a faulty component, otherwise L 1 For faulty components;

[0123] S11, if the R1 error is less than the L1 error, it is determined that R1 is faulty;

[0124] S12, if the R1 error is not less than the L1 error, L1 is determined to be faulty;

[0125] S13, if the R1 / L1 current ratio does not exceed the threshold, determine whether the R2 / L2 current ratio exceeds the threshold;

[0126] S14, if the R2 / L2 current ratio exceeds the threshold, determine whether the R2 / C2 current ratio exceeds the threshold;

[0127] S15, if the current ratio of R2 / C2 exceeds the threshold, it is determined that R2 is faulty;

[0128] S16, if the R2 / C2 current ratio does not exceed the threshold, L2 is determined to be faulty;

[0129] S17, if it is determined that the R2 / L2 current ratio does not exceed the threshold, determine whether the R2 / C2 current ratio exceeds the threshold;

[0130] S18, if the current ratio of R2 / C2 exceeds the threshold, it is determined that C2 is faulty;

[0131] S19. If the R2 / C2 current ratio does not exceed the threshold, it is determined that C1 is faulty.

[0132] In this embodiment, if R 1 / L 1 If the current ratio exceeds the threshold, the process goes to step S9, otherwise it goes to step S12; 1 / L 1 Harmonic current ratio calculation R 1 , L 1 The corresponding change; calculate R 1 , L 1 Theoretical and actual harmonic impedance offset errors under corresponding changes; R is determined according to the error size. 1 / L 1 The faulty component in the process ends; according to the logical operation, determine C 1 / R 2 / L 2 / C 2 The faulty component in the process ends.

[0133] In this embodiment, R 1 / L 1 The current is higher than normal:

[0134] If R 2 / L 2 , R 2 / C 2 If the current ratio of both exceeds the threshold, the faulty component is R 2 ;

[0135] If only R 2 / L 2 If the current ratio exceeds the threshold, the faulty element is L 2 ;

[0136] If only R 2 / C 2 If the current ratio exceeds the threshold, the faulty element is C 2 ;

[0137] If R 2 / L 2 , R 2 / C 2 The current ratio of is within the threshold at the same time, then the faulty component is C 1 .

[0138] Example 2

[0139] In this embodiment, simulation verification is used to further explain and illustrate the method proposed in this patent. The simulation model is built using Matlab / Simulink, and the specific parameters of its primary circuit and equipment are consistent with a domestic ±1100kV ultra-high voltage direct current transmission system.

[0140] Table 1 gives the parameters of the various components of the DC filter.

[0141] Table 1 DC filter component parameters

[0142]

[0143]

[0144] This DC filter is a 2 / 12 type, and the 12th harmonic content is higher, so m is selected as 12.

[0145] The allowable error values ​​are as follows: C 1 ±0.5%, and the remaining components are ±1%, so k 1 =k 2 =1%,ε i (i=1,2,3) are all set to 0.1%, the number of Fourier decompositions n is selected to be 3, and the time point t 1 =0.2s, t 2 =0.3s, t 3 =0.4s.

[0146] Based on the Simulink model, the data of the parameter changes of each component in the DC filter device of the UHV DC transmission system within ±50% are collected. The parameter changes of the components when the detuning fault occurs are generally within this range. 1 The specific changes are: from -50% to -2%, the increment is 1%; from -2% to -0.5%, the increment is 0.1%; from 0.5% to 2%, the increment is 0.1%; from 2% to 50%, the increment is 1%. 1 , L 1 , R 2 , L 2 , C 2 The changes are: from -50% to -2%, the increment is 1%; from -2% to -1%, the increment is 0.1%; from 1% to 2%, the increment is 0.1%; from 2% to 50%, the increment is 1%. A total of 718 sets of data were collected, and all the faulty components were correctly identified.

[0147] like Figure 3a to Figure 3c As shown, C 1 The harmonic current ratio when a fault occurs. It can be seen that the current ratios at the three locations are all within the threshold value, and the faulty component is C 1 .

[0148] like Figures 4a to 4d As shown, R 1 Harmonic current ratio and impedance error when a fault occurs. It can be seen that only R 1 / L 1 The current ratio of the branch is always outside the threshold, and R 1 The corresponding theoretical and actual harmonic impedance errors are always lower than L 1 , the faulty component is R 1 .

[0149] like Figure 5a to Figure 5d As shown, L 1 Harmonic current ratio and impedance error when a fault occurs. It can be seen that only R 1 / L 1 The current ratio of the branch is always outside the threshold, and L 1 The corresponding theoretical and actual harmonic impedance errors are always lower than R 1 , the faulty component is L 1 .

[0150] like Figures 6a to 6c As shown, R 2 The harmonic current ratio when a fault occurs can be seen from the R 2 / L 2 and R 2 / C 2 The current ratio of the branch is always outside the threshold at the same time, and the faulty element is R 2 .

[0151] like Figures 7a to 7c As shown, L 2 Harmonic current ratio when a fault occurs. It can be seen that only R 2 / L 2 The current ratio of the branch is always outside the threshold, and the faulty element is L 2 .

[0152] like Figures 8a to 8c As shown, C 2 Harmonic current ratio when a fault occurs. It can be seen that only R 2 / L 2 The current ratio of the branch is always outside the threshold, and the faulty element is C 2 .

[0153] In summary, after the relay protection detects that the DC filter is detuned, the present invention uses a fault recorder to record the fault and generate a recording file. According to the recording data related to the DC filter in the recording file, the harmonic current ratio of each parallel branch after detuning is calculated, and by judging whether it exceeds the corresponding threshold, the parallel branch where the detuned element is located can be determined, and finally the detuned element can be located by calculating the theoretical and actual harmonic impedance offset errors.

[0154] The present invention is based on R 1 / L 1 Harmonic current ratio calculation R 1 , L 1 The corresponding change is used to obtain the harmonic impedance offset error under the corresponding change, and quickly determine R 1 / L 1 The faulty element in the double-tuned filter is located, which avoids the ratio overlap caused by the small difference in impedance offset at the tuning point of the double-tuned filter in the traditional fault location technology, thereby improving the fault identification accuracy.

[0155] The invention is based on on-site wave recording data, has a simple detection process, and has clear physical meaning. It can quickly and accurately detect detuned fault components in a damped double-tuned DC filter online, and has certain engineering application value.

[0156] The present invention detects whether a COMTRADE recording file is generated on a file server, specifically including a .cfg file and a .data file. By using the above method, information stored in the COMTRADE text in binary format is transferred to variables that are convenient for program operation, thereby improving the online detection efficiency of detuned fault components in a damped double-tuned DC filter.

[0157] The present invention is not only aimed at high voltage capacitor C 1 The fault is detected and R 1 , R 2 , L 1 , L 2 , C 2 The present invention does not rely on the selection of sample sets and the division of attribute space, reduces the complexity of the algorithm, is easy to implement in engineering, and improves the applicability of the fault detection technology.

[0158] After the detuning protection is activated, the present invention detects the components in the filter to locate the components whose parameters have slightly changed in the filter. When a detuning fault is caused by a gradual change in the parameters of a single component in the DC filter, the present invention can quickly and accurately detect and determine the faulty component while the filter detuning protection is activated, greatly shortening the filter repair time and improving the filter availability. The present invention solves the technical problem of low fault detection efficiency and low accuracy for detuning faulty components in damped double-tuned DC filters in the prior art.

[0159] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.< / cr>

Claims

1. An online fault detection method applicable to UHVDC DC filters, characterized in that, the method includes: S1. Compare the magnitudes of the harmonic currents at the ends of adjacent poles to determine whether the DC filter is detuned, generate a COMTRADE fault recording file, and intercept the recording data from the COMTRADE fault recording file; S2. Perform anomaly detection and repair on the oscillographic data, and calculate the parallel elements R 1 / L 1 、R 2 / L 2 、R 2 / C 2 's harmonic current ratios, and determine whether the current ratios of the parallel elements R 1 / L 1 、R 2 / L 2 、R 2 / C 2 exceed the preset thresholds. Step S2 includes: S21. Collect the instantaneous current value data flowing through R 1 , R 2 , L 1 , L 2 , C 2 . Select n different time points, obtain the m-th harmonic current value through Fourier decomposition, and calculate the harmonic current ratio at n time points using the preset logic based on the harmonic current value; S22. Determine the parallel element R 1 / L 1 、R 2 / L 2 、R 2 / C 2 Harmonic current ratio threshold value S23. Determine whether the current ratios of the parallel components R 1 / L 1 、R 2 / L 2 、R 2 / C 2 exceed the harmonic current ratio threshold; S3. If R 1 / L 1 the current ratio exceeds the preset threshold, calculate the change amount corresponding to R 1 / L 1 according to the harmonic current ratio, calculate the harmonic impedance offset error of R 1 and L 1 under the corresponding change amount, and determine the faulty component in R 1 and L 1 accordingly. S3 includes: 1 / L 1 ​ S31. Process the harmonic current ratio to obtain the harmonic impedance ratio, calculate the harmonic impedance ratio to obtain the change in L 1 , R 1 . Among them, the harmonic impedance ratio is calculated according to the following logic to obtain the change in L 1 , R 1 : where k R1 and k L1 are the change amounts of R 1 and L 1 corresponding to the harmonic current ratio after detuning, respectively; S32. Calculate R 1 , L 1 respectively corresponding to the m - th theoretical harmonic impedance values under the corresponding change amounts, subtract the preset logic from the theoretical nominal value, obtain the current actual m - th harmonic impedance value according to the harmonic voltage - current ratio, and subtract the actual normal value to obtain the m - th harmonic impedance value; Among them, the following logic is used to calculate the difference between the m - th theoretical harmonic impedance value and the actual normal value to obtain the m - th harmonic impedance value: Among them, and are the magnitudes of the deviation of the m - th harmonic impedance values theoretically calculated for R 1 and L 1 at the corresponding change amounts k R1 and k L1 from the nominal value, is the nominal value of the m - th harmonic impedance of the DC filter, and are the theoretical values of the m - th harmonic impedance for R 1 and L 1 at the change amounts of k R1 and k L1 respectively, is the magnitude of the deviation of the actually calculated m - th harmonic impedance value from the normal value, and are the effective values of the m - th harmonic voltages at both ends of the DC filters of the positive and opposite poles respectively, and are the effective values of the m - th harmonic currents flowing through the first ends of the DC filters of the positive and opposite poles respectively, error R1 and error L1 are the theoretical and actual harmonic impedance offset errors corresponding to R 1 and L 1 respectively; S4. If the R1 / L1 current ratio does not exceed the preset threshold, determine the faulty component in C1 / R2 / L2 / C2 according to the preset fault determination logic.

2. The online fault detection method applicable to UHVDC DC filters according to claim 1, characterized in that, the step S1 includes: S11. Compare the magnitudes of the harmonic currents at the ends of adjacent poles with the following relay protection action equation to determine whether the DC filter is detuned: IZxT2_OP_m > k * IZxT2_m and IZxT2_OP > Iset where x represents the DC filter group, IZxT2_m is the amplitude of the m - th harmonic current of the current at the end of the DC filter of this pole, IZxT2_OP_m is the amplitude of the m - th harmonic current of the current at the end of the DC filter of the adjacent other pole, IZxT2_OP is the amplitude of the current at the end of the DC filter of the adjacent other pole, Iset is the set value, and k is the detuning monitoring coefficient; S12. Detect and read the COMTRADE fault recording file, and transfer the recording information in the COMTRADE fault recording file to the preset program operation variable; S13. Intercept the recorded waveforms of the voltages and currents of each DC filter in a specific period after the fault from the COMTRADE fault recording file; S14. Generate the recording data according to the voltage and current recorded waveforms.

3. The online fault detection method applicable to UHVDC DC filters according to claim 1, characterized in that, the step S21 includes: Collect the instantaneous current value data to process and obtain the harmonic current ratios at the n time points, and use the following logic to average the n harmonic current ratios: Among them, respectively represent the final values of the m - th harmonic current ratios of the branches R1 / L1, R2 / L2, and R2 / C2 of this pole, respectively represent the ratios of the effective values of the m - th harmonic currents obtained by Fourier decomposition at the recording data time point ti for the branches R1 / L1, R2 / L2, and R2 / C2 of this pole, where i = 1, 2, …, n.

4. The online fault detection method applicable to UHVDC DC filters according to claim 1, characterized in that, the step S22 includes: S221. Obtain the nominal value of the component; S222, R calculated according to the nominal value of the component 1 / L 1 , R 2 / L 2 , R 2 / C 2 The standard value of the m - th harmonic impedance ratio of the parallel branch S223. According to the standard value, process with the following logic to obtain the harmonic current ratio threshold.

5. The online fault detection method applicable to UHVDC DC filters according to claim 4, characterized in that, In the step S222, R calculated according to the nominal value of the component by the following logic 1 / L 1 , R 2 / L 2 , R 2 / C 2 Standard value of the impedance ratio of the parallel branch for the mth harmonic: where Z R1 、 Z R2 、 are the AC impedance values of the corresponding components at a frequency of f m respectively, f m is the frequency magnitude of the m-th harmonic, R 1 、R 2 、L 1 、L 2 、C 2 are the nominal values of the components respectively.

6. The online fault detection method applicable to UHVDC DC filters according to claim 4, characterized in that, In the step S223, process the standard value with the following logic to obtain the harmonic current ratio threshold: Among them, bottom_x and top_x are the lower and upper boundaries of the harmonic current ratio threshold respectively, where x = {RL1, RL2, RC2}, k 1 = min{k R1 , k L1}, k 2 = min{k R2 , k L2 , k C2}, and ε i (i = 1, 2, 3) are small positive constants.

7. The online fault detection method applicable to UHVDC DC filters according to claim 1, characterized in that, the step S4 includes: S41. At R 2 / L 2 and the current ratio of R 2 / C 2 both exceed the harmonic current ratio threshold at the same time, determine that the faulty component is R 2 ; S42. At R 2 / L 2 When the current ratio exceeds the harmonic current ratio threshold and R 2 / C 2 When the current ratio does not exceed the harmonic current ratio threshold, determine that the faulty component is L 2 ; S43. At R 2 / C 2 When the current ratio exceeds the harmonic current ratio threshold and R 2 / L 2 When the current ratio does not exceed the harmonic current ratio threshold, the faulty component is C 2 ; S44. When the 2 R / L 2 current ratio and the 2 R / C 2 current ratio are both within the threshold, it is determined that the faulty component is C 1 .

8. An on-line fault detection system applicable to UHVDC DC filters, which is used to execute the on-line fault detection method applicable to UHVDC DC filters described in any one of the foregoing claims 1 to 7, Characterized in that, The system includes: A waveform recording data acquisition module, which is used to compare the magnitudes of the harmonic currents at the tails of two adjacent poles to determine whether the DC filter is detuned, generate a COMTRADE fault waveform recording file accordingly, and intercept waveform recording data from the COMTRADE fault waveform recording file; Parallel component current ratio judgment module, which is used to perform anomaly detection and repair on the recorded wave data, and calculate the parallel components R 1 / L 1 、R 2 / L 2 、R 2 / C 2 of the harmonic current ratio, and judge whether the current ratios of the parallel components R 1 / L 1 、R 2 / L 2 、R 2 / C 2 exceed the preset thresholds. The parallel component current ratio judgment module is connected to the recorded wave data acquisition module. The parallel component current ratio judgment module includes: Harmonic current ratio module, used to collect the instantaneous current value data flowing through R after detuning 1 , R 2 , L 1 , L 2 , C 2 , select n different time points, obtain the mth harmonic current value through Fourier decomposition, and calculate the harmonic current ratio at the n time points using the preset logic according to the harmonic current value; Harmonic current ratio threshold module for determining the parallel element R 1 / L 1 、R 2 / L 2 、R 2 / C 2 harmonic current ratio thresholds; A threshold judgment module for judging the parallel element R 1 / L 1 , R 2 / L 2 , R 2 / C 2 whether the current ratio exceeds the harmonic current ratio threshold. The current ratio judgment module is connected to the harmonic current ratio threshold module and the harmonic current ratio module; R 1 / L 1 Faulty component determination module, used to, when the current ratio exceeds the preset threshold, calculate the change amount corresponding to R 1 / L 1 and L according to the harmonic current ratio, calculate the harmonic impedance offset error of R 1 / L 1 under the corresponding change amount, and determine the faulty component in R 1 、L 1 accordingly. The faulty component determination module is connected to the parallel component current ratio judgment module. The faulty component determination module includes: 1 、L 1 When the corresponding change amount occurs, calculate the harmonic impedance offset error of R 1 / L 1 and L, and determine the faulty component in R A harmonic impedance ratio processing module is used to process the harmonic current ratio to obtain a harmonic impedance ratio, calculate the harmonic impedance ratio to obtain the changes in L 1 , R 1 ; Harmonic impedance value acquisition module for calculating R 1 , L 1 The m - th theoretical harmonic impedance values corresponding respectively under the corresponding change amounts, based on which the difference is made between the preset logic and the theoretical nominal value, and then the current actual m - th harmonic impedance value is obtained according to the harmonic voltage - current ratio, and based on which the difference is made with the actual normal value to obtain the m - th harmonic impedance value. The harmonic impedance value acquisition module is connected to the harmonic impedance ratio processing module; A C1 / R2 / L2 / C2 fault element determination module, which is used to determine the fault element in C1 / R2 / L2 / C2 according to the preset fault determination logic when the R1 / L1 current ratio does not exceed the preset threshold, and the C1 / R2 / L2 / C2 fault element determination module is connected to the parallel element current ratio judgment module.

Images