Generator excitation rectification loop fault identification method based on information fingerprints

Abstract

The invention discloses a generator excitation rectification loop fault identification method based on information fingerprints, and belongs to the technical field of power system fault diagnosis. According to the method, multi-signal acquisition is utilized, fault fingerprint information is calculated, fault detection is carried out, a set rough fingerprint is used for judging a damaged rectifier cabinet, and a fine fingerprint is used for identifying a damaged thyristor in the rectifier cabinet; in order to improve the recognition precision, a coarse bit error rate (GBER) and a fine bit error rate (ABER) are defined as fingerprint similarity measurement indexes, and an energy correction coefficient is introduced for correction; by calculating the bit error rate, the system can accurately match the fault fingerprints with fingerprints in the I-level and II-level fault recognition libraries to determine the fault type; in addition, the self-expansion capability of the system is improved, the fault recognition library can be continuously updated during operation, and more data are collected by establishing a supplementary fault recognition library, so that the fault recognition function is enhanced. According to the method, the fault detection speed and accuracy of the excitation system are improved, and then the stability and safety of the power system are enhanced.

Description

Technical Field

[0001] The invention relates to the technical field of power system fault diagnosis, and in particular to a generator excitation rectifier circuit fault identification method based on information fingerprint. Background Art

[0002] With the transformation of the global energy structure, the large-scale construction of new energy power generation systems and the operation of large-capacity thermal power units have put forward higher requirements for the reliability, safety and stability of modern power grids. As a key component to ensure the stable operation of the power grid, the synchronous generator excitation system can dynamically adjust the excitation current of the generator to maintain the stability of the output voltage when the power load changes rapidly. However, the timely diagnosis of excitation system faults is crucial to ensure the safe operation of the generator set. Existing fault diagnosis methods rely on comprehensive analysis of electrical signals at different positions of the excitation system. These methods have high requirements on the accuracy and diversity of signal information sources and are difficult to implement in practice. Since there is a direct electrical connection between the excitation voltage signal, excitation current signal, trigger angle signal and the excitation transformer, excitation rectifier cabinet, and generator rotor winding, the signal waveform contains rich state information. Therefore, it is of great significance to diagnose excitation system faults based on the excitation voltage signal.

[0003] Information fingerprint technology has been effectively applied in the field of audio and video recognition. By analyzing the unique characteristics of the signal and constructing a hash function, the signal can be classified and quickly found. At present, some studies have tried to apply information fingerprint technology to the field of fault identification of electrical systems, such as the identification of the fault location of the DC power grid. However, these methods are mainly based on the time-frequency information of the original signal, have poor robustness to signal time shift, and are highly dependent on the start time of signal acquisition. It is difficult to make consistent fingerprint mapping for signals without obvious trigger signs. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a method for identifying faults in the excitation rectifier circuit of a generator based on information fingerprints. The method uses the signal autocorrelation function within a period of time instead of the original signal to extract the time-frequency features and perform fingerprint mapping, thereby improving the robustness of the fingerprint code to the signal time shift, making the identification scheme of matching the fingerprint code with the fault fingerprint library suitable for periodic stable analog signals, thereby effectively realizing the rapid and accurate identification of different fault types in the three-phase bridge rectifier circuit of the excitation system.

[0005] The present invention adopts the following technical solution: A fault identification method for a generator excitation rectifier circuit based on information fingerprint, wherein the generator excitation rectifier circuit includes a group of three-phase bridge rectifier cabinets connected in parallel, and each three-phase bridge rectifier cabinet includes a group of bridge-connected thyristors; the specific steps of the fault identification method are as follows: Step 1: setting the generator excitation rectifier circuit fault identification to include level I rectifier cabinet fault identification and level II thyristor fault identification, wherein the level I rectifier cabinet fault identification is used to determine the damaged rectifier cabinet, and the level II thyristor fault identification is used to further determine the damaged thyristor in the rectifier cabinet; Step 2: Simulate the failure of each level I rectifier cabinet and t The time domain information of the directly collected quantity within 1 second is mapped to form the corresponding rough fingerprint of the I-level rectifier cabinet fault, 1< t 1<5, establish a level I fault identification library, in which the directly collected quantities include the trigger angle and excitation voltage; Step 3: Simulate the fault of each level II thyristor and t The time domain information of the directly collected quantity within 2 seconds is used to calculate the time domain information of the non-collected quantity, 10< t 2<30, the calculated non-collected time domain information is mapped to form the corresponding Class II rectifier fault detailed fingerprint, and a Class II fault identification library is established, where the non-collected quantities include excitation power, AC side excitation current and average energy; Step 4: Transplant the Level I fault identification library and Level II fault identification library to the on-site measurement and control platform of the generator excitation system. When the generator excitation rectifier circuit fails, collect t The time domain information directly collected within 1 second is mapped to generate fingerprint code and compared with the rough fingerprint of the level I fault identification library. GBER And calculate the rough bit error rate GBER The rough fingerprint with the highest similarity is determined, and the level I rectifier cabinet fault corresponding to the rough fingerprint with the highest similarity is sent up as the identification result for short-time relay protection of the generator excitation system; Step 5: When the generator excitation rectifier circuit fails, collect t Directly collect the time-frequency information within 2 seconds, calculate and generate the non-collected time domain information, generate the fingerprint code through the mapping algorithm and compare it with the refined fingerprint of the Level II fault identification library, and define the refined bit error rate ABER And calculate the bit error rate ABER The fine fingerprint with the highest similarity is determined, and the level II thyristor fault corresponding to the fine fingerprint with the highest similarity is sent up as the fault identification result for operation and maintenance of the excitation rectifier circuit and replacement of thyristor devices after the generator is shut down; Step 6: After the Level I fault identification library and the Level II fault identification library have run the set cycle in the generator excitation system measurement and control platform, the corresponding coarse fingerprint and fine fingerprint in the fault identification library are dynamically updated according to the identification library judgment logic; for the generator excitation rectifier circuit fault that cannot be identified by the Level I fault identification library and the Level II fault identification library, the corresponding Level III supplementary fault identification library and Level IV supplementary fault identification library are established respectively.

[0006] Further: In step 2, the steps of forming the rough fingerprint are as follows: Step 2.1) Based on the time domain information of the directly collected quantity, calculate its autocorrelation function and extract the previous k cycles, the expression is as follows: ; In the formula, is the time domain signal of the directly acquired quantity, is the length of the time domain signal sequence, T is the signal acquisition step length; Step 2.2) Perform wavelet transform reconstruction on the extracted autocorrelation function and divide the wavelet transform reconstruction result into signal frames, calculate the signal frame energy Zero crossing number , the expression is: ; ; In the formula, It is Layer The energy of a signal frame, It is Layer The number of zero crossings in a signal frame, Representative Layer reconstruction signal; Represents the number of sampling points per frame; represents a symbolic function; Step 2.3) and After mapping, the corresponding fingerprint information is as follows: ; ; In the formula, Indicates Tier The signal frame energy of the frame, is the corresponding fingerprint bit value, For the Tier The number of zero crossings in the frame, is the corresponding fingerprint bit value.

[0007] Further: In step 3, the steps of forming the detailed fingerprint are as follows: Step 3.1) Based on the time domain information of the directly collected quantities, calculate the non-collected quantities of excitation power, AC side excitation current and average energy, which are defined as follows: Step 3.1.1) Excitation power The formula is as follows: ; In the formula, is the DC side excitation voltage, is the excitation current; Step 3.1.2) Assumption M If the excitation currents in the rectifier cabinets are equal, the average excitation current on the AC side is The formula is as follows: ; ; In the formula, is the trigger angle, is the rectifier energy balance coefficient, is the current sharing coefficient, is the thyristor switching loss, is the heat loss of the thyristor, is the cooling system loss of the rectifier system, is the control system loss; if the rectifier cabinet does not have the intelligent current sharing function, the control system loss is ; Step 3.1.3) Average Energy The formula is as follows: ; ; In the formula, N is the length of the time domain signal sequence, T is the signal acquisition step length, ceiling represents the ceiling function; Step 3.2) Based on the calculated time domain information of the non-collected quantity, calculate its autocorrelation function and extract the previous k cycles, the expression is as follows: ; In the formula, is the time domain signal of the non-collected quantity, is the length of the time domain signal sequence, T is the signal acquisition step length; Step 3.3) Perform wavelet transform reconstruction on the extracted autocorrelation function and divide the wavelet transform reconstruction result into signal frames, calculate the signal frame energy Zero crossing number , the expression is: ; ; In the formula, It is Layer The energy of a signal frame, It is Layer The number of zero crossings in a signal frame, Representative Layer reconstruction signal; Represents the number of sampling points per frame; represents a symbolic function; Step 3.4) and After mapping, the corresponding fingerprint information is as follows: ; ; In the formula, Indicates Tier The signal frame energy of the frame, is the corresponding fingerprint bit value; For the Tier The number of zero crossings in the frame, is the corresponding fingerprint bit value.

[0008] Further: the process of step 4 is as follows: Defining the coarse bit error rate GBER To measure the similarity between rough fingerprints, and introduce energy correction coefficient , the expression is: ; ; In the formula, For the The number of different bit values ​​in the rough fingerprint layer, is the total number of fingerprints in each layer, is the total number of rough fingerprint layers, is the weight of each layer of fingerprint; , are respectively the average energy of the time domain waveform of the directly acquired quantity measured in actual operation and the average energy of the time domain waveform of the directly acquired quantity stored in the level I fault identification library, and the expression is: ; ; In the formula, , They are respectively the time domain signal of the directly acquired quantity measured in actual operation and the time domain signal of the directly acquired quantity stored in the level I fault identification library. and are the signal sequence lengths of the two respectively; The rough fingerprint generated by the actual operation fault is compared with the rough fingerprint of each fault in the level I fault identification library in turn to obtain the minimum rough bit error rate GBER min , minimum coarse bit error rate GBER min The corresponding rough fingerprint is the rough fingerprint with the highest similarity to the actual rectifier cabinet fault in the level I fault identification library.

[0009] Further: the process of step 5 is as follows: Defining Bit Error Rate ABER To measure the similarity between detailed fingerprints, and introduce energy correction coefficient , the expression is: ; ; In the formula, For the The number of different bit values ​​in the layer-by-layer fingerprint is is the total number of fingerprints in each layer, is the total number of layers of refined fingerprints, is the weight of each layer of fingerprint; , are the average energy of the non-collected time domain waveform measured in actual operation and the average energy of the non-collected time domain waveform stored in the Level II fault identification library, respectively. The expression is: ; ; In the formula, , They are the time domain signals of non-collected quantities measured in actual operation and the time domain signals of non-collected quantities stored in the Level II fault identification library. and are the signal sequence lengths of the two respectively; The refined fingerprint generated by the actual operation fault is compared with the refined fingerprint of each fault in the Level II fault identification library in turn to obtain the minimum fine bit error rate ABER min , minimum bit error rate ABER min The corresponding refined fingerprint is the refined fingerprint with the highest similarity to the actual thyristor fault in the Level II fault identification library.

[0010] Further: In step 6, the fault identification library is updated and supplemented according to the coarse / fine bit error rate values ​​of the actual fault matching, and the specific process is as follows: Step 6.1) When the matched coarse / fine bit error rate is less than the set threshold When the fingerprint code in the level I / level II fault identification library is dynamically updated, the steps are as follows: Step 6.1.1) Compare the coarse / fine fingerprints in the level I / level II fault identification library with the coarse / fine fingerprints generated by the actual operation fault, and identify the positions with different bit values ​​in the corresponding fingerprints; Step 6.1.2) For different positions, according to the set probability Replace the coarse / fine elements in the level I / level II fault identification library with the elements of the corresponding coarse / fine fingerprints in the actual fault fingerprint; Step 6.2) When the matched coarse / fine bit error rate is greater than or equal to the threshold ,The steps for establishing the Level III / Level IV supplementary fault identification library are as follows: Step 6.2.1) re-collect the time domain information of the directly collected quantity during the fault process twice, and according to the above step 4 / step 5), respectively calculate the coarse / fine bit error rate of the two collected signals, and obtain the corresponding fault identification result; Step 6.2.2) If the two groups of coarse / fine bit error rates calculated in step 6.2.1) are both less than the set threshold , and the fault identification results obtained are consistent, the result will be sent as the I-level rectifier cabinet / II-level thyristor fault identification result; otherwise, all the coarse / fine fingerprints of this fault and the time domain waveforms of the corresponding directly collected quantities are added to the III-level / IV-level supplementary fault identification library for subsequent identification and research, and a special fault warning is output.

[0011] The present invention provides a method for identifying faults in a generator excitation rectifier circuit based on information fingerprint, which has the following beneficial effects: (1) The fingerprint data generated by the method of the present invention is small in volume, and can be stored and quickly matched in large quantities for fault conditions under different parameters, and has a strong ability to distinguish different asymmetric fault states under different trigger angles; (2) The method of the present invention extracts the time-frequency information of the autocorrelation function of the original signal based on wavelet transform reconstruction, so that the excitation electric fingerprint is highly robust to noise influence and signal time shift, and the bit error rate of the signal fingerprint and the original fingerprint is reduced; (3) The multi-fault library parallel and update scheme in the method of the present invention can enable the simulated fault fingerprint library to be updated in real time according to the actual operation conditions, and output warnings for questionable data in a timely manner, so that the identification method has strong reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is the principle diagram of the excitation system of the present invention; Figure 2It is a schematic diagram of the information fingerprint structure of the present invention; Figure 3 The trigger angle of the present invention , the sampling signal waveform of the normal working part of the excitation system, 3a is the excitation voltage signal waveform, 3b is the excitation current signal waveform, 3c is the excitation active power signal waveform; Figure 4 The trigger angle of the present invention , only the signal waveforms are sampled before and after the thyristor VT1 is short-circuited, 4a is the excitation voltage signal waveform, 4b is the excitation current signal waveform, and 4c is the excitation active power signal waveform; Figure 5 The trigger angle of the present invention , only the excitation voltage time-frequency diagram before and after the thyristor VT1 is short-circuited, 5a is the excitation voltage signal waveform, 5b is the excitation current signal waveform, 5c is the excitation active power signal waveform; Figure 6 In the present invention, the relationship between the bit error rate of the fingerprint generated by the original signal and the autocorrelation function of the original signal and the original fingerprint after the sampled signal is shifted in time; Figure 7 In the present invention, the error rate of the fingerprint between the original signal and the original signal after different degrees of white noise are introduced; Figure 8 In the present invention, only when the VT1 thyristor is short-circuited, different trigger angles Voltage fingerprint and trigger angle The bit error rate between voltage fingerprints; Fig. 9 In the present invention, the bit error rates between fingerprints under 12 different fault types are selected. DETAILED DESCRIPTION

[0013] In order to describe the present invention more specifically, the technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0014] Embodiment: A method for identifying faults in a generator excitation rectifier circuit based on information fingerprints. The generator excitation rectifier circuit is composed of at least three three-phase bridge rectifier cabinets connected in parallel, and each rectifier cabinet is connected in a bridge manner by six thyristor devices. The specific process of the fault identification method is as follows: Step 1: Set the generator excitation rectifier circuit fault identification to include level I rectifier cabinet fault identification and level II thyristor fault identification, wherein level I rectifier cabinet fault identification is used to determine the damaged rectifier cabinet, and level II thyristor fault identification is used to further determine the damaged thyristor in the rectifier cabinet.

[0015] Step 2: simulate the faults of each level I rectifier cabinet, and map the time domain information of the directly collected quantities within a few seconds after each fault to form a rough fingerprint, thereby establishing a level I fault identification library, in which the directly collected quantities include the trigger angle and the excitation voltage.

[0016] Step 3, simulate each level II thyristor fault, calculate the time domain information of the non-collected quantity from the time domain information of the directly collected quantity within tens of seconds after each fault, and map them to form a detailed fingerprint, thereby establishing a level II fault identification library, in which the non-collected quantity includes excitation power, AC side excitation current and average energy.

[0017] Step 4: Transplant the level I and level II fault identification libraries to the on-site measurement and control platform of the generator excitation system. When the generator excitation rectifier circuit fails, collect the time domain information of the directly collected quantity within a few seconds, generate the fingerprint code through the mapping algorithm and compare it with the rough fingerprint of the level I fault identification library, and define and calculate the rough bit error rate. GBER The rough fingerprint with the highest similarity is determined, and the level I rectifier cabinet fault corresponding to the rough fingerprint is sent as the identification result for short-time relay protection of the generator excitation system.

[0018] Step 5: When the generator excitation rectifier circuit fails, the time-frequency information of the directly collected quantity within tens of seconds is collected, and the time domain information of the non-collected quantity is calculated and generated. The fingerprint code is generated by the mapping algorithm and compared with the refined fingerprint of the Level II fault identification library. The refined bit error rate is defined and calculated. ABER The refined fingerprint with the highest similarity is determined, and the level II thyristor fault corresponding to the refined fingerprint is sent as the fault identification result for operation and maintenance of the excitation rectifier circuit and replacement of thyristor devices after the generator is shut down.

[0019] Step 6: After the level I and level II fault identification libraries have run for a certain period in the generator excitation system measurement and control platform, the coarse fingerprint and fine fingerprint of the fault identification library are dynamically updated according to the identification library judgment logic; at the same time, for the generator excitation rectifier circuit fault that cannot be identified by the level I and level II fault identification libraries, level III and level IV supplementary fault identification libraries are established respectively to output special fault warnings.

[0020] The schematic diagram of the excitation system is as follows: Figure 1 The generator terminal voltage is input into the three-phase bridge rectifier circuit through the excitation transformer and the synchronous transformer, and then rectified by the thyristors VT1-VT6 to obtain a stable DC current, which finally passes through the generator rotor winding to realize the excitation operation of the generator set. The automatic voltage regulator (AVR) is based on the trigger angle. Provide corresponding 6-pulse signals to control the enablement of thyristors VT1-VT6. Figure 3 for When the thyristor is working normally, the excitation voltage, excitation current, and excitation active power curves are shown. When any thyristor has a short circuit or open circuit fault, the excitation voltage signal will change, and after reaching stability, it will present a curve with different characteristics with a period of 0.02s. The failure of a thyristor will destroy the symmetry of the three-phase AC source, and the proportion of high-frequency components will increase (see Figure 4 , Figure 5 ), so the characteristics of different frequency bands of the signal can be obtained through wavelet decomposition and reconstruction, and the fingerprint of the voltage signal can be constructed from this.

[0021] In step 2, the steps of forming the rough fingerprint are as follows: Step 2.1) Based on the time domain information of the directly collected quantity, calculate its autocorrelation function and extract the previous k cycles, the expression is as follows: ; In the formula, is the time domain signal of the directly acquired quantity, is the length of the time domain signal sequence, T is the signal acquisition step length; Step 2.2) Perform wavelet transform reconstruction on the intercepted autocorrelation function and divide the wavelet transform reconstruction result into signal frames, calculate the signal frame energy Zero crossing number , the expression is: ; ; In the formula, It is Layer The energy of a signal frame, It is Layer The number of zero crossings in a signal frame, Representative Layer reconstruction signal; Represents a symbolic function.

[0022] Step 2.3) and After mapping, the corresponding fingerprint information is as follows: ; ; In the formula, Indicates Tier The signal frame energy of the frame, is the corresponding fingerprint bit value; For the Tier The number of zero crossings in the frame, is the corresponding fingerprint bit value.

[0023] In step 3, the steps for forming the detailed fingerprint are as follows: Step 3.1) Based on the time domain information of the directly collected quantities, calculate the non-collected quantities of excitation power, AC side excitation current and average energy, which are defined as follows: Step 3.1.1) Excitation power The formula is as follows: ; In the formula, is the DC side excitation voltage, is the excitation current; Step 3.1.2) Assumption M If the excitation currents in the rectifier cabinets are equal, the average excitation current on the AC side is The formula is as follows: ; ; In the formula, is the trigger angle, is the rectifier energy balance coefficient, is the current sharing coefficient, is the thyristor switching loss, is the heat loss of the thyristor, is the cooling system loss of the rectifier system, is the control system loss; if the rectifier cabinet does not have the intelligent current sharing function, the control system loss is ; Step 3.1.3) Average Energy The formula is as follows: ; ; In the formula, N is the length of the time domain signal sequence, T is the signal acquisition step length, ceiling Represents the ceiling function.

[0024] Step 3.2) Based on the calculated time domain information of the non-collected quantity, calculate its autocorrelation function and extract the previous k cycles, the expression is as follows: ; In the formula, is the time domain signal of the non-collected quantity, is the length of the time domain signal sequence, T is the signal acquisition step.

[0025] Step 3.3) Perform wavelet transform reconstruction on the intercepted autocorrelation function and divide the wavelet transform reconstruction result into signal frames, calculate the signal frame energy Zero crossing number , the expression is: ; ; In the formula, It is Layer The energy of a signal frame, It is Layer The number of zero crossings in a signal frame, Representative Layer reconstruction signal; Represents the number of sampling points per frame; Represents a symbolic function.

[0026] Step 3.4) and After mapping, the corresponding fingerprint information is as follows: ; ; In the formula, Indicates Tier The signal frame energy of the frame, is the corresponding fingerprint bit value; For the Tier The number of zero crossings in the frame, is the corresponding fingerprint bit value.

[0027] In step 4, define the coarse bit error rate GBER To measure the similarity between rough fingerprints, and introduce energy correction coefficient , the expression is: ; ; In the formula, For the The number of different bit values ​​in the rough fingerprint layer, is the total number of fingerprints in each layer, is the total number of rough fingerprint layers, is the weight of each layer of fingerprint; , are respectively the average energy of the time domain waveform of the directly acquired quantity measured in actual operation and stored in the Level I fault identification library, and the expression is: ; ; In the formula, and They are respectively the time domain signals of the directly acquired quantities measured in actual operation and stored in the Level I fault identification library, and are the signal sequence lengths of the two respectively.

[0028] The rough fingerprint generated by the actual operation fault is compared with the rough fingerprint of each fault in the level I fault identification library in turn, and the minimum rough bit error rate can be obtained. GBER min , which corresponds to the rough fingerprint with the highest similarity to the actual rectifier cabinet fault in the level I fault identification library.

[0029] In step 5, define the fine bit error rate ABER To measure the similarity between detailed fingerprints, and introduce energy correction coefficient , the expression is: ; ; In the formula, For the The number of different bit values ​​in the layer-by-layer fingerprint is is the total number of fingerprints in each layer, is the total number of layers of refined fingerprints, is the weight of each layer of fingerprint; , are respectively the average energy of the non-collected time domain waveform measured in actual operation and stored in the Level II fault identification library, and the expression is: ; ; In the formula, and They are respectively the time domain signals measured in actual operation and the non-collected signals stored in the Level II fault identification library. and are the signal sequence lengths of the two respectively.

[0030] The minimum bit error rate can be obtained by comparing the refined fingerprint generated by the actual operation fault with the refined fingerprint of each fault in the Level II fault identification library. ABER min , which corresponds to the refined fingerprint with the highest similarity to the actual thyristor fault in the Level II fault identification library.

[0031] In step 2-5: for different trigger angles Under different error conditions, the sampling signals of the circuit are simulated, and the fault state, directly collected time domain signal, coarse fingerprint, and signal average energy are combined and stored to establish a level I fault identification library; the fault state, calculated non-collected time domain signal, fine fingerprint, and signal average energy are combined and stored to establish a level II fault identification library. After the fault occurs, the fault sampling signal waveforms with a time length of 1s and 10s are collected, and the coarse and fine fingerprint information are respectively produced, which are matched with the fingerprints in the level I and level II fault identification libraries and the coarse bit error rate is calculated. GBER , bit error rate ABER The fault type corresponding to the minimum coarse bit error rate is used as the fault identification result of the level I rectifier cabinet, which is used for the short-time relay protection of the generator excitation system; the fault type corresponding to the minimum fine bit error rate is used as the fault identification result of the level II thyristor, which is used for the operation and maintenance of the excitation rectifier circuit and the replacement of thyristor devices after the generator is shut down.

[0032] In step 6, the fault identification library will be updated and supplemented according to the coarse (fine) bit error rate values ​​matched by the actual fault. The specific process is as follows: 6.1) When the matching coarse (fine) bit error rate is less than the set threshold When the fingerprint code in the level I (level II) fault identification library is dynamically updated, the steps are as follows: Step 6.1.1) Compare the original fingerprint in the Level I (Level II) fault identification library with the rough (fine) fingerprint generated by the actual operation fault, and identify the positions with different bit values ​​in the corresponding fingerprints; Step 6.1.2) For each different position, according to the set probability Replace the elements of the original fingerprint in the Level I (Level II) fault identification library with the corresponding elements in the actual fault fingerprint.

[0033] Step 6.2) When the matched coarse (fine) bit error rate is greater than or equal to the threshold The steps for establishing the Level III (Level IV) supplementary fault identification library are as follows: Step 6.2.1) Re-collect the time domain information of the directly collected quantity during the fault process twice, calculate the coarse (fine) bit error rate of the two collected signals respectively according to step 4 (step 5), and obtain the corresponding fault identification result; Step 6.2.2) If the two groups of coarse (fine) bit error rates calculated in step 6.2.1) are both less than the set threshold , and the fault identification results obtained are consistent, the result will be sent as the fault identification result of the I-level rectifier cabinet (II-level thyristor); otherwise, all the coarse (fine) fingerprints of this fault and the time domain waveforms of the corresponding directly collected quantities are added to the III-level (IV-level) supplementary fault identification library for subsequent identification and research, and special fault warnings are output.

[0034] The process of updating and supplementing the fault identification library is shown in Figure 4 .

[0035] In order to accurately identify various fault types, the fingerprint mapping method should have good robustness and anti-collision capabilities. That is, for the same type of fault, the error rate between fingerprints should be small when affected by noise or time shift; while the error rate between fingerprints of different fault types should be large. In this regard, this embodiment built a generator excitation system simulation model in MATLAB / Simulink software and simulated and verified the proposed method. Figure 1 The schematic diagram of the excitation system is shown, and Table 1 shows the simulation parameters of the generator excitation system. Figure 2 The information fingerprint structure is shown, where layers 1-5 are composed of a6-a 10 and a 10 Signal frame energy generation, 6-8 layers: signal frame zero crossing number generation of d7-d9.

[0036] Table 1: Simulation parameters of generator excitation system

[0037] In order to verify the strong robustness of the method of generating fingerprints by autocorrelation function to the influence of time shift of the original signal, and considering that the coarse fingerprint and the fine fingerprint have almost the same mapping formation and fault matching process, this embodiment takes the formation and fault matching of the fine fingerprint as an example for verification and explanation.

[0038] Figure 6 The relationship between the bit error rate of the fingerprint generated by the original signal and the autocorrelation function of the original signal and the original fingerprint after the sampled signal (taking the excitation active power as an example) is shifted in time is shown. Figure 6 It can be seen that after using the autocorrelation function, the influence of the temporal shift is greatly reduced and is controlled below 5%. Therefore, the fingerprint mapping method of this embodiment uses the time-frequency domain information of the signal autocorrelation function instead of the original signal, which greatly improves the influence of the temporal shift of the periodic voltage signal on the fingerprint.

[0039] Figure 7 The bit error rate of the fingerprint between the original signal and the original signal after different degrees of white noise are introduced is shown. Figure 7 It can be seen that when the signal-to-noise ratio is greater than 10dB, the bit error rate is always less than 5%, and the bit error rate between fingerprints for voltage fingerprints under different fault types is 20%-80%. Therefore, the fingerprint mapping method of this embodiment can maintain good stability when the signal-to-noise ratio is above 10dB, and is highly robust to noise influence and signal time shift.

[0040] Figure 8 The following figure shows the different trigger angles under the fault type of only VT1 thyristor short circuit. Fingerprint information and trigger angle The bit error rate between fingerprint information. Figure 8 It can be seen that when the bit error rate is less than 5%, Controlled . Fig. 9 The bit error rates between fingerprints under 12 different excitation system fault types are shown, all of which are above 20%. The corresponding relationship between fault numbers and fault types is shown in Table 2.

[0041] Table 2: Corresponding numbers of fault states of rectifier circuits in the excitation system

[0042] This embodiment uses a 6-bit ternary array to represent the circuit fault state, where 1 represents an open circuit, -1 represents a short circuit, and 0 represents normal operation. For example, when both VT1 and VT2 are open circuits, the fault state is represented as (1, 1, 0, 0, 0, 0).

[0043] Therefore, this embodiment effectively distinguishes the waveforms of sampling signals at different trigger angles under the same fault condition, and the fingerprint mapping method of this embodiment has good anti-collision performance and can effectively distinguish different fault types.

[0044] comprehensive Figure 6-Figure 9 The method of this embodiment can effectively distinguish the waveform of the sampled signal under different trigger angles and fault conditions, and has strong robustness to the interference noise and time shift effects with a signal-to-noise ratio of more than 10dB relative to the original signal. In this embodiment, due to the symmetry between the phases of the input three-phase AC power, the three-phase symmetrical faults (such as only VT1 short circuit and only VT3 short circuit) have consistent steady-state voltage waveforms, and these two types of faults are classified as the same fault type.

[0045] The above description of the embodiments is to facilitate the understanding and application of the present invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments, and improvements and modifications made by those skilled in the art based on the disclosure of the present invention should be within the scope of protection of the present invention.

Claims

1. A method for identifying faults in a generator excitation rectifier circuit based on information fingerprint, characterized in that: The generator excitation rectifier circuit includes a group of parallel three-phase bridge rectifier cabinets, and each three-phase bridge rectifier cabinet includes a group of bridge-connected thyristors; the specific steps of the fault identification method are as follows: Step 1, setting the generator excitation rectifier circuit fault identification to include level I rectifier cabinet fault identification and level II thyristor fault identification; Step 2: simulate the faults of each level I rectifier cabinet, form a corresponding rough fingerprint of the level I rectifier cabinet fault according to the information collected after the fault of each level I rectifier cabinet, and establish a level I fault identification library; Step 3: Simulate the faults of each level II thyristor, form a corresponding level II rectifier fault fingerprint according to the information collected after the fault of each level II thyristor, and establish a level II fault identification library; Step 4: Generate the corresponding fingerprint code based on the information collected by the actual generator excitation rectifier fault system, compare it with the coarse fingerprint and fine fingerprint in the corresponding fault identification library, and define the coarse bit error rate respectively. GBER, Bit Error Rate ABER , the rough bit error rate calculated by GBER, Bit Error Rate ABER Determine the rough fingerprint and the fine fingerprint with the highest similarity, and send the fault corresponding to the rough fingerprint and the fine fingerprint with the highest similarity as the identification result; Step 5: Dynamically update the corresponding coarse fingerprint and fine fingerprint in the fault identification library; For the generator excitation rectifier circuit faults that cannot be identified by the Level I fault identification library and the Level II fault identification library, corresponding Level III supplementary fault identification library and Level IV supplementary fault identification library are established respectively.

2. The method for identifying faults in a generator excitation rectifier circuit based on information fingerprint according to claim 1 is characterized in that: In step 2, each level I rectifier cabinet fails. t The time domain information of the directly collected quantities within 1 second is mapped to form the corresponding rough fingerprint of the I-level rectifier cabinet fault, and the I-level fault identification library is established; the directly collected quantities include the trigger angle and the excitation voltage; wherein, the steps of forming the rough fingerprint are as follows: Step 2.1) Based on the time domain information of the directly collected quantity, calculate its autocorrelation function and extract the previous k cycles; Step 2.2) Perform wavelet transform reconstruction on the extracted autocorrelation function and divide the wavelet transform reconstruction result into signal frames, calculate the signal frame energy Zero crossing number ; Step 2.3) and Mapping is performed to obtain the corresponding fingerprint information.

3. The method for identifying faults in a generator excitation rectifier circuit based on information fingerprint according to claim 1 is characterized in that: In step 3, according to the failure of each level II thyristor t The time domain information of the directly collected quantity within 2 seconds is used to calculate the time domain information of the non-collected quantity. The calculated time domain information of the non-collected quantity is mapped to form the corresponding Class II rectifier cabinet fault detailed fingerprint, and a Class II fault identification library is established; the non-collected quantity includes excitation power, AC side excitation current and average energy; wherein, the steps for forming the detailed fingerprint are as follows: Step 3.1) Based on the time domain information of the directly collected quantities, calculate the non-collected quantities of excitation power, AC side excitation current and average energy; Step 3.2) Based on the calculated time domain information of the non-collected quantity, calculate its autocorrelation function and extract the previous k cycles; Step 3.3) Perform wavelet transform reconstruction on the extracted autocorrelation function and divide the wavelet transform reconstruction result into signal frames, calculate the signal frame energy Zero crossing number ; Step 3.4) and Mapping is performed to obtain the corresponding fingerprint information.

4. The method for identifying faults in a generator excitation rectifier circuit based on information fingerprint according to claim 1 is characterized in that: The process of step 4 is as follows: 4.1) Definition of coarse bit error rate GBER To measure the similarity between rough fingerprints, and introduce energy correction coefficient ; The rough fingerprint generated by the actual operation fault is compared with the rough fingerprint of each fault in the level I fault identification library in turn to obtain the minimum rough bit error rate GBER min , minimum coarse bit error rate GBER min The corresponding rough fingerprint is the rough fingerprint with the highest similarity to the actual rectifier cabinet fault in the level I fault identification library; 4.2) Definition of Bit Error Rate ABER To measure the similarity between detailed fingerprints, and introduce energy correction coefficient ; The refined fingerprint generated by the actual operation fault is compared with the refined fingerprint of each fault in the Level II fault identification library in turn to obtain the minimum fine bit error rate ABER min , minimum bit error rate ABER min The corresponding refined fingerprint is the refined fingerprint with the highest similarity to the actual thyristor fault in the Level II fault identification library.

5. The method for identifying faults in a generator excitation rectifier circuit based on information fingerprint according to claim 1 is characterized in that: In step 5, the fault identification library is updated and supplemented according to the coarse / fine bit error rate values ​​of the actual fault matching. The specific process is as follows: Step 5.1) When the matched coarse / fine bit error rate is less than the set threshold When the fingerprint code in the level I / level II fault identification library is dynamically updated, the steps are as follows: Step 5.1.1) Compare the coarse / fine fingerprints in the level I / level II fault identification library with the coarse / fine fingerprints generated by the actual operation fault, and identify the positions with different bit values ​​in the corresponding fingerprints; Step 5.1.2) For different positions, according to the set probability Replace the coarse / fine elements in the level I / level II fault identification library with the elements of the corresponding coarse / fine fingerprints in the actual fault fingerprint; Step 5.2) When the matched coarse / fine bit error rate is greater than or equal to the threshold ,The steps for establishing the Level III / Level IV supplementary fault identification library are as follows: Step 5.2.1) Re-collect the time domain information of the directly collected quantity during the fault process twice, calculate the coarse / fine bit error rate of the two collected signals respectively according to the step 4, and obtain the corresponding fault identification result; Step 5.2.2) If the two groups of coarse / fine bit error rates calculated in step 5.2.1) are both less than the set threshold , and the fault identification results obtained are consistent, the result will be sent as the I-level rectifier cabinet / II-level thyristor fault identification result; otherwise, all the coarse / fine fingerprints of this fault and the time domain waveforms of the corresponding directly collected quantities are added to the III-level / IV-level supplementary fault identification library for subsequent identification and research, and a special fault warning is output.

Images