Open-circuit fault diagnosis method for three-phase six-switch tube Vienna rectifier of direct-current charging pile

Abstract

The invention relates to a direct current charging pile three-phase six-switch tube Vienna rectifier open-circuit fault diagnosis method, and belongs to the field of direct current charging pile fault diagnosis. The method comprises the following steps: acquiring alternating-current side three-phase current and direct-current side voltage data of the Vienna rectifier in real time; frequency domain information of alternating-current side three-phase current and direct-current side voltage is extracted from the collected data through fast Fourier transform, and then normalized open-circuit fault characteristic quantity is obtained; and inputting the normalized open-circuit fault characteristic quantity into the trained open-circuit fault diagnosis model of the Vienna rectifier of the direct-current charging pile to obtain an open-circuit fault diagnosis result of the Vienna rectifier. The Vienna rectifier single-tube and double-tube open-circuit fault diagnosis method can realize Vienna rectifier single-tube and double-tube open-circuit fault diagnosis only by acquiring alternating current side current and direct current side voltage data, and has an important value for reducing the fault maintenance time and the operation and maintenance cost of a direct current charging pile.

Description

Technical Field

[0001] The invention belongs to the field of fault diagnosis of a DC charging pile and relates to a method for diagnosing an open circuit fault of a three-phase six-switch Vienna rectifier of a DC charging pile. Background Art

[0002] At present, the electric vehicle market is expanding rapidly, and the construction of charging facilities is also accelerating. In 2023, the number of public charging piles in China has reached 2.626 million, of which 1.14 million are DC charging piles, an increase of 114% year-on-year. DC charging piles work in complex environments such as high temperature and high humidity at different altitudes, and operate under high voltage, high current, complex electrical noise and other working conditions for a long time, causing the mechanical strength and electrical performance of various components in the charging pile to decrease rapidly, and serious sudden shutdown failures occur. Among them, the switch tube (generally referring to IGBT and MOSFET) is a key component for power conversion in the power converter of the DC charging pile, which is prone to short circuit faults and open circuit faults. Statistics show that nearly 34% of the failure rate in the power converter is caused by the failure of the switch tube. Short circuit faults will generate easily detectable instantaneous overcurrents, which are usually converted into open circuit faults by fuses. Open circuit faults are not easy to be found when they occur, especially when single or two switch tubes open circuit faults occur, no instantaneous overcurrent or undervoltage will be generated, which has a strong hidden characteristic, but long-term open circuit fault operation will cause other fault-free switch tubes to fail.

[0003] At present, the mainstream DC charging pile power converter mainly includes the front-stage AC / DC rectifier and the rear-stage DC / DC converter. Among them, the Vienna rectifier is often used in the front-stage AC / DC converter because of its advantages such as low cost, high conversion efficiency, and high power factor. According to the circuit topology, the Vienna rectifier is divided into two types: three-phase three-switch tube and three-phase six-switch tube. For the open-circuit fault diagnosis of the Vienna rectifier tube, the current diagnostic methods are mostly for the three-phase three-switch tube Vienna rectifier with a single switch tube structure, and it is impossible to realize the fault diagnosis of the simultaneous open circuit of two switch tubes in the three-phase six-switch tube Vienna rectifier. In addition, for the extraction of fault feature quantities, the current diagnostic methods are mostly for the time domain fault data extraction of the Vienna rectifier voltage and current. However, in the three-phase six-switch tube Vienna rectifier, when the two switch tubes fail at the same time, the time domain waveform changes are complex, and it is difficult to extract effective open circuit fault feature quantities. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a method for diagnosing an open-circuit fault of a three-phase six-switch Vienna rectifier of a DC charging pile. Based on the three-phase current on the AC side and the DC side voltage monitoring data of the Vienna rectifier, and using an RBF-based Vienna rectifier fault diagnosis model, the open-circuit fault diagnosis of the Vienna rectifier of the DC charging pile is realized, thereby reducing the fault inspection time and operation and maintenance costs of the DC charging pile.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A method for diagnosing an open circuit fault of a Vienna rectifier of a DC charging pile, comprising:

[0007] S1, real-time collection of AC side three-phase current and DC side voltage data of Vienna rectifier;

[0008] S2. Using Fast Fourier Transform (FFT) to extract the frequency domain information of the three-phase current on the AC side and the voltage on the DC side from the collected data, and then obtain the normalized open circuit fault feature quantity;

[0009] S3, training a DC charging pile Vienna rectifier open circuit fault diagnosis model based on radial basis function (RBF) neural network;

[0010] S4. Input the normalized open-circuit fault feature into the trained DC charging pile Vienna rectifier open-circuit fault diagnosis model to obtain the open-circuit fault diagnosis result of the Vienna rectifier.

[0011] Furthermore, the open circuit fault types of the Vienna rectifier include single-tube open circuit fault, double-tube same-phase open circuit fault, double-tube non-same-phase group open circuit fault and double-tube non-same-phase different-group open circuit fault.

[0012] Furthermore, in step S1, a current sensor and a voltage sensor are used to collect T (T>1) cycles of Vienna rectifier AC side three-phase current and DC side voltage monitoring data in real time at a sampling frequency f. The AC side three-phase current data of T cycles is I M (0) I M (1) I M (2),…,I M (N-1), M = a, b, c, M represents the phase of the current; T cycles of DC side voltage data V(0), V(1), V(2), ..., V(N-1),

[0013] Furthermore, in step S2, FFT is used to extract frequency domain information of the monitoring data and calculate open circuit fault feature quantities, wherein the open circuit fault feature quantities include: fault feature quantities based on the three-phase current on the AC side and fault feature quantities based on the voltage on the DC side.

[0014] Calculate the fault characteristic quantity based on the three-phase current on the AC side: Use FFT to extract frequency domain information from the three-phase current data on the AC side:

[0015]

[0016] In the formula, I M [ω] represents the amplitude of the current frequency domain information at ω, ω∈[0,N-1], I M (n) represents the phase current on the AC side, and N represents the data volume;

[0017] Using formula (1), when ω = 0, the fault characteristic quantity - the DC component of the AC current is obtained:

[0018]

[0019] Where, X Z Represents the DC component of each phase current, X1, X2, and X3 represent the DC components of phase current a, b, and c respectively.

[0020] Calculate the fault characteristic quantity based on the DC side voltage: Use FFT to extract frequency domain information from the DC side voltage data:

[0021]

[0022] In the formula, there is a one-to-one correspondence between V(n) and the frequency domain component V[ω], ω∈[0,N-1]; V[ω] represents the amplitude at the frequency domain information ω; N represents the amount of data; V(n) represents the DC side voltage;

[0023] The fault characteristic quantities based on the DC side voltage data include: mean value, standard deviation, skewness, kurtosis, peak-to-peak value, root mean square, crest factor, waveform factor, and energy. These fault characteristic quantities are obtained using the following formulas (4) to (12).

[0024] 1) Average value:

[0025]

[0026] 2) Standard Deviation:

[0027]

[0028] 3) Skewness:

[0029]

[0030] 4) Kurtosis:

[0031]

[0032] 5) Peak-to-peak value:

[0033]

[0034] 6) Root Mean Square:

[0035]

[0036] 7) Crest factor:

[0037]

[0038] 8) Form factor:

[0039]

[0040] 9) Energy:

[0041]

[0042] According to the above open circuit fault characteristic quantity, the normalized open circuit fault characteristic quantity is obtained as follows:

[0043]

[0044] Where, X k represents the characteristic quantity of the kth open circuit fault, X k * is the normalized open circuit fault characteristic quantity, They represent the upper and lower thresholds of the open circuit fault characteristic quantity, respectively. The upper and lower thresholds are obtained based on historical monitoring data. Represents the kth open circuit fault characteristic quantity of the sth segment of historical monitoring data.

[0045] Furthermore, the normalized open circuit fault feature quantities in S2 are combined into

[0046] Furthermore, in step S3, the training process of the DC charging pile Vienna rectifier open circuit fault diagnosis model includes:

[0047] Label the normal state of the Vienna rectifier and the fault state corresponding to each open circuit fault type;

[0048] The historical monitoring data of the three-phase current on the AC side and the voltage on the DC side of the Vienna rectifier in the normal state and the open circuit fault state for s time periods are obtained, and the normalized open circuit fault characteristic quantity combination X in each time period is calculated according to equations (1) to (13):(1)* , X (2)* , …, X (s)* , and get the training samples:

[0049]

[0050] in, represents the kth normalized open circuit fault characteristic value of the sth segment of historical monitoring data, F s For X (s)* The corresponding label.

[0051] The training samples are divided into a training set and a test set, and the DC charging pile Vienna rectifier open circuit fault diagnosis model is trained, wherein the normalized open circuit fault feature quantity combination X (s)* As model input, the corresponding label is used as model output;

[0052] After the training is completed, determine whether the accuracy of the model meets the requirements. If it does, output the model; otherwise, retrain.

[0053] Further, in step S4, the open circuit fault diagnosis is performed by using the trained DC charging pile Vienna rectifier open circuit fault diagnosis model, including: firstly, the normalized open circuit fault feature quantity calculated according to the collected data is combined into Then X * The input is into the Vienna rectifier open circuit fault diagnosis model of the DC charging pile, and the corresponding open circuit fault type is determined according to the label output by the model.

[0054] The beneficial effects of the present invention are as follows: the present invention uses fast Fourier transform to extract the frequency domain information of the AC side current and DC side voltage of the three-phase six-switch Vienna rectifier, and then obtains the fault characteristic quantity based on the three-phase current on the AC side and the fault characteristic quantity based on the DC side voltage, and applies the DC charging pile Vienna rectifier open circuit fault diagnosis model based on RBF neural network to obtain the open circuit fault diagnosis result. The present invention only needs to collect the AC side current and DC side voltage data of the DC charging pile Vienna rectifier in real time, and can realize the diagnosis of single and double tube open circuit faults of Vienna rectifier, which has important engineering value for reducing the fault inspection time and operation and maintenance costs of DC charging piles.

[0055] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0057] Figure 1 A circuit diagram of a Vienna rectifier of a DC charging pile diagnosed by an embodiment of the present invention;

[0058] Figure 2 A schematic diagram of a flow chart of an open circuit fault diagnosis of a Vienna rectifier according to an embodiment of the present invention;

[0059] Figure 3 The current waveform diagram of open circuit fault under different fault types;

[0060] Figure 4 The voltage waveforms of open circuit faults under different fault types. DETAILED DESCRIPTION

[0061] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0062] The present invention provides a method for diagnosing an open circuit fault of a three-phase six-switch Vienna rectifier of a DC charging pile. The method is based on the monitoring data of the three-phase current on the AC side and the voltage on the DC side of the Vienna rectifier, and uses an RBF neural network to establish a fault diagnosis model to realize the open circuit fault diagnosis of the Vienna rectifier of the DC charging pile.

[0063] An embodiment of the present invention is directed to Figure 1 The DC charging pile three-phase six-switch tube Vienna rectifier shown in the figure is used for fault diagnosis. The single-tube and double-tube open circuit fault types of the three-phase six-switch tube DC charging pile Vienna rectifier are 21 types, as shown in Table 1. The main parameters of the three-phase six-switch tube DC charging pile Vienna rectifier are shown in Table 2. This embodiment combines Figure 2 The Vienna rectifier open circuit fault diagnosis flow chart shown in the figure illustrates the implementation steps of the method of the present invention through three parts: data acquisition part, fault feature extraction and open circuit fault diagnosis.

[0064] Table 1 Vienna rectifier fault types

[0065]

[0066] Table 2 Vienna rectifier circuit parameter design

[0067]

[0068] Step 1: Data collection: Use current sensors and voltage sensors to collect T (T>1) cycles of Vienna rectifier AC side three-phase current and DC side voltage data in real time.

[0069] Taking the three-phase six-switch Vienna rectifier of the DC charging pile as an example, when the load voltage is 600V, the current sensor and voltage sensor are used to collect the AC side current data and DC side voltage data of the Vienna rectifier in real time, such as Figure 3 and Figure 4 As shown, the sampling frequency is 10kHz. Combined with Table 1, it can be seen that the current and voltage waveforms of different open circuit fault types T1, T3, T7, and T10 are different.

[0070] Step 2: Fault feature extraction part. Use FFT to extract the frequency domain information of the collected data to calculate the open circuit fault feature, and finally normalize the fault feature according to equations (1) to (13).

[0071] For example, for Figure 3 The data of 5 cycles in the Sa2 open circuit fault are obtained according to formula (1) to formula (13), and the fault characteristic quantity is:

[0072] X=[0.01 0.43 87.85 9199 50.84 0.43 118.78 41.07 4581 1.45 8.577.13],

[0073] The fault characteristic quantity after normalization is:

[0074] X * =[0.35 0.41 0.60 0.06 0.12 0.41 0.25 0.66 0.07 0.07 0.48 0.38]

[0075] Among them, the upper limit thresholds of fault feature quantities are: 0.0278, 1.04, 139.8, 24000, 200, 1.04, 151, 59, 48496, 19.54, 17.99, and 18.54; the lower limit thresholds are: 0.0006, 0.004, 10, 8300, 30, 0.004, 108, 6.63, 1457, 0.003, 0.005, and 0.004.

[0076] Step 3: Open circuit fault diagnosis part. The fault characteristic quantity X after normalization in step 2 is * The output result is obtained by inputting it into the RBF-based DC charging pile Vienna rectifier open circuit fault diagnosis model, and then querying the fault number in Table 1 to obtain the open circuit fault diagnosis result.

[0077] For example, replace X in step 2 * The input is input into the RBF-based DC charging pile Vienna rectifier open circuit fault diagnosis model, and the fault label is T3. The query table 1 shows that the Sb2 switch tube has an open circuit fault, and the fault diagnosis result is correct.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A method for diagnosing open circuit faults of a three-phase six-switch Vienna rectifier of a DC charging pile, characterized in that: The method includes: Real-time collection of three-phase current on the AC side and voltage on the DC side of the three-phase six-switch Vienna rectifier; Fast Fourier transform is used to extract the frequency domain information of the three-phase current on the AC side and the voltage on the DC side from the collected monitoring data, and then the normalized fault feature quantity is obtained; The normalized open-circuit fault feature quantity is input into the trained DC charging pile Vienna rectifier open-circuit fault diagnosis model to obtain the open-circuit fault diagnosis result of the Vienna rectifier; the fault types of the Vienna rectifier include single-tube open-circuit fault, double-tube same-phase open-circuit fault, double-tube non-same-group open-circuit fault and double-tube non-same-phase different-group open-circuit fault.

2. The method according to claim 1, characterized in that: The frequency domain information extracted from the three-phase current data on the AC side using fast Fourier transform is: In the formula, I M [ω] represents the amplitude at frequency domain information ω, I M (n) represents the phase current on the AC side, N represents the data volume, and M represents the phase of the current; According to the above formula, when ω=0, the fault characteristic quantity based on the three-phase current on the AC side, that is, the DC component of the AC current, is obtained: Where, X Z is the DC component of each phase current, and X1, X2, and X3 are the DC components of the three-phase currents a, b, and c respectively.

3. The method according to claim 1, characterized in that: The frequency domain information extracted from the DC side voltage data using fast Fourier transform is: Where V[ω] represents the amplitude at frequency domain information ω; N represents the amount of data; V(n) represents the DC side voltage; According to the above formula, the fault characteristic quantities based on the DC side voltage data include: 1) Average value: 2) Standard Deviation: 3) Skewness: 4) Kurtosis: 5) Peak-to-peak value: X8 = V max -V min , V max = max{V[1], V[2], V[3], …, V[ω], …, V[N - 1]}, V min = min{V[1], V[2], V[3], …, V[ω], …, V[N - 1]}; 6) Root Mean Square: 7) Crest factor: 8) Form factor: 9) Energy:

4. The method according to claim 1, characterized in that: The obtaining of the normalized fault characteristic quantity comprises: calculating the normalized open circuit fault characteristic quantity according to the open circuit fault characteristic quantity, wherein the open circuit fault characteristic quantity comprises a fault characteristic quantity based on the three-phase current data on the AC side and a fault characteristic quantity based on the voltage data on the DC side; The normalized open circuit fault characteristic is calculated by the following formula: Where, X k represents the characteristic quantity of the kth open circuit fault, where X1~X3 are the fault characteristic quantities based on the three-phase current data measured by AC, X4~X 12 is the fault characteristic quantity based on DC side voltage data, X k * is the normalized open circuit fault characteristic, X kmax , X kmin They represent the upper and lower thresholds of the open circuit fault characteristic quantity, Represents the kth open circuit fault characteristic quantity of the sth segment of historical monitoring data.

5. The method according to claim 1, characterized in that: The training process of the DC charging pile Vienna rectifier open circuit fault diagnosis model includes: Label the normal state of the Vienna rectifier and the fault state corresponding to each open circuit fault type; Obtain the historical monitoring data of s periods of the three-phase current on the AC side and the voltage on the DC side of the Vienna rectifier under normal and open-circuit fault conditions, calculate the normalized open-circuit fault feature value in each period, and obtain the training samples: In the formula, represents the kth normalized open circuit fault characteristic value of the sth segment of historical monitoring data, k = 1, 2, ..., 12, represents the combination of k normalized open-circuit fault feature quantities of the sth segment of historical monitoring data, F s Representation and The corresponding label; The training samples are divided into a training set and a test set, and a DC charging pile Vienna rectifier open circuit fault diagnosis model based on a radial basis function neural network is trained, wherein the input of the DC charging pile Vienna rectifier open circuit fault diagnosis model is a normalized open circuit fault feature quantity, and the output is a label; After the training is completed, it is determined whether the accuracy of the DC charging pile Vienna rectifier open circuit fault diagnosis model meets the requirements. If it meets the requirements, the model is output, otherwise the training is performed again.

6. The method according to claim 5, characterized in that: The open circuit fault diagnosis is performed by using the trained DC charging pile Vienna rectifier open circuit fault diagnosis model, which includes: firstly, the normalized open circuit fault feature quantity calculated according to the collected data is combined into Then X * The input is into the Vienna rectifier open circuit fault diagnosis model of the DC charging pile, and the corresponding open circuit fault type is determined according to the label output by the model.

Images