Fault detection method for voltage-sharing capacitor in converter valve

Abstract

The invention discloses a fault detection method for a voltage-sharing capacitor in a converter valve, and relates to the technical field of converter valve fault detection. Acquiring a frequency response function of a voltage-sharing capacitor in the converter valve in a complex frequency domain form, selecting conjugate complex pairs in linear distribution in a preset frequency band as initial poles, and defining an auxiliary function; determining an order value of the frequency response function according to the frequency response function, the auxiliary function and the frequency response curve; solving a coefficient initial value of the frequency response function according to the initial pole, and determining a zero point of the auxiliary function according to the coefficient initial value and the initial pole; obtaining a target coefficient of the frequency response function by taking the zero point of the auxiliary function equal to the pole of the frequency response function as a constraint condition; determining a target zero point and a target pole of the frequency response function according to the target coefficient and the order value of the frequency response function; and determining a fault detection result of the voltage-sharing capacitor according to the target zero point and the target pole. According to the method, fault detection of the voltage-sharing capacitor in the converter valve can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of converter valve fault detection, and particularly relates to a method for detecting faults of grading capacitors in a converter valve. Background Art

[0002] A converter valve is a core device in a DC transmission project. It converts three-phase alternating current into direct current to achieve the transmission of DC electrical energy. In a converter valve, thyristors are the core components. However, under conditions such as steep wave impact, the voltage distribution inside the valve will change significantly. The uneven voltage distribution inside the valve will cause damage to thyristors. In order to improve the voltage distribution inside the valve and make it more uniform, it is necessary to install grading capacitors for thyristors, which can evenly distribute the voltage between two single valves inside a double valve or the voltage inside the valve, prevent uneven voltage sharing during the turn-on and turn-off processes of thyristor elements, avoid overvoltage on the elements and damage, and achieve voltage equalization and protection of the converter valve.

[0003] However, faults of the grading capacitors in the converter valve will bring many hazards to the converter valve. For example, it may cause uneven voltage in part, resulting in excessive voltage on the thyristors and causing thyristor breakdown. Moreover, when a grading capacitor fails, the transient voltage may rise rapidly to form a voltage impact, threatening the insulation of the components inside the converter valve and triggering faults such as insulation breakdown, leading to short circuit or open circuit of the converter valve. In addition, faults of the grading capacitors may cause problems such as reduced efficiency and increased losses.

[0004] Therefore, there is an urgent need for a method that can detect faults of grading capacitors in a converter valve. Summary of the Invention

[0005] Based on this, in view of the above technical problems, it is necessary to provide a method for detecting faults of grading capacitors in a converter valve, which can realize the fault detection of grading capacitors in the converter valve.

[0006] The present invention adopts the following technical solutions:

[0007] The present invention provides a method for detecting faults of grading capacitors in a converter valve, including:

[0008] Obtain the frequency response function of the grading capacitor in the converter valve in the complex frequency domain form, select a pair of conjugate complex numbers with a linear distribution as the initial poles within a preset frequency band, and define an auxiliary function according to the initial poles; the preset frequency band is the frequency band that needs to be vector-fitted;

[0009] Determine the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve;

[0010] Solve the initial values of the coefficients of the frequency response function according to the initial poles, and determine the zeros of the auxiliary function based on the initial values of the coefficients and the initial poles;

[0011] Taking the zeros of the auxiliary function being equal to the poles of the frequency response function as the constraint condition, use the zeros of the auxiliary function as the poles of the frequency response function to iteratively update the coefficients of the frequency response function to obtain the target coefficients of the frequency response function;

[0012] Determine the target zeros and target poles of the frequency response function according to the target coefficients and the order value of the frequency response function;

[0013] Determine the fault detection result of the equalizing capacitor according to the target zeros and target poles.

[0014] Optionally, the frequency response function f(s) in the complex frequency domain form is:

[0015]

[0016] where c n is the zero of f(s), a n is the pole of f(s), d and h are real numbers, s is the representation in the complex frequency domain, and N is the order of f(s).

[0017] Optionally, the auxiliary function σ(s) is:

[0018]

[0019] where is the pole of σ(s) and the initial pole of f(s), is the zero of σ(s).

[0020] Optionally, determine the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve, including:

[0021] Multiply the frequency response function by the auxiliary function to obtain a relational expression: The relational expression is

[0022] Obtain the simplified form of the relational expression. The simplified form of the relational expression is (σf) fit (s) = σ fit (s)f(s); where (σf) fit and σ fit (s) are written in the zero-pole form as: z n represents the zero of (σf) fit , represents the zero of σ fit (s).

[0023] Convert the abbreviated form of the relational expression into the linear form Ax = b; A is the coefficient matrix corresponding to each sampling frequency point, and b is f(s) corresponding to each sampling frequency point.

[0024] Expand both sides of the equation in linear form to obtain another representation of the frequency response function:

[0025]

[0026] where z n and respectively represent the zeros of (σf) fit and σ fit (s);

[0027] Determine the order value of the frequency response function according to another representation of the frequency response function and the frequency response curve.

[0028] Optionally, determine the order value of the frequency response function according to another representation of the frequency response function and the frequency response curve, including:

[0029] Obtain the fitting curves of another representation of the frequency response function at different orders;

[0030] Determine the order corresponding to the minimum deviation between the fitting curve and the frequency response curve as the order value of the frequency response function.

[0031] Optionally, solve the initial coefficient value of the frequency response function according to the initial poles, including:

[0032] Substitute the initial poles into Ax = b to solve for the initial coefficient value x.

[0033] Optionally, determine the fault detection result of the equalizing capacitor according to the target zeros and target poles, including:

[0034] Input the target zeros and target poles into the pre-trained support vector machine model to obtain the fault detection result of the equalizing capacitor.

[0035] Optionally, the construction process of the support vector machine model includes:

[0036] Obtain the frequency response curves of the converter valve equalizing capacitor under different fault conditions;

[0037] Determine the target zeros and target poles of the frequency response function corresponding to different fault conditions according to the frequency response curves of the converter valve equalizing capacitor under different fault conditions;

[0038] Take the target zeros and target poles of the frequency response functions corresponding to different fault conditions as the data sample set;

[0039] Train according to the data sample set to obtain a support vector machine model.

[0040] Optionally, the kernel function of the support vector machine is a radial basis function, and the range of the penalty coefficient of the support vector machine is [0.01, 1].

[0041] The present invention provides a fault detection device for the grading capacitors in a converter valve, including:

[0042] An acquisition module, configured to acquire the frequency response function of the grading capacitors in the converter valve in the complex frequency domain form, select conjugate complex pairs with linear distribution as the initial poles within a preset frequency band, and define an auxiliary function according to the initial poles; the preset frequency band is the frequency band that needs to be vector-fitted;

[0043] A first determination module, configured to determine the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve;

[0044] A second determination module, configured to solve the initial values of the coefficients of the frequency response function according to the initial poles, and determine the zeros of the auxiliary function according to the initial values of the coefficients and the initial poles;

[0045] A third determination module, configured to take the zeros of the auxiliary function equal to the poles of the frequency response function as a constraint condition, take the zeros of the auxiliary function as the poles of the frequency response function, and iteratively update the coefficients of the frequency response function to obtain the target coefficients of the frequency response function;

[0046] A fourth determination module, configured to determine the target zeros and target poles of the frequency response function according to the target coefficients and the order value of the frequency response function;

[0047] A fifth determination module, configured to determine the fault detection result of the grading capacitors according to the target zeros and target poles.

[0048] The present invention provides a computer-readable storage medium, where the storage medium stores a computer program, and when the computer program is executed by a processor, it implements the above-mentioned fault detection method for the grading capacitors in a converter valve.

[0049] The present invention provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the above-mentioned fault detection method for the grading capacitors in a converter valve.

[0050] At least one of the above technical solutions adopted by the present invention can achieve the following beneficial effects:

[0051] Based on the frequency response function of the voltage-sharing capacitors in the converter valve, through vector fitting and rational approximation, the poles of the frequency response function are iteratively adjusted to calculate the target zeros and poles of the frequency response function of the voltage-sharing capacitors in the converter valve. The fault detection of the voltage-sharing capacitors in the converter valve is carried out based on the zeros and poles, ensuring more accurate fault detection of the voltage-sharing capacitors in the converter valve. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The drawings described herein are used to provide a further understanding of the present invention and form a part of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0053] Figure 1 It is a schematic flowchart of a method for fault detection of voltage-sharing capacitors in a converter valve provided by the present invention;

[0054] Figure 2 It is a schematic flowchart of another method for fault detection of voltage-sharing capacitors in a converter valve provided by the present invention;

[0055] Figure 3 It is a schematic diagram of a device for fault detection of voltage-sharing capacitors in a converter valve provided by the present invention;

[0056] Figure 4 It is a schematic diagram of a computer device for implementing a method for fault detection of voltage-sharing capacitors in a converter valve provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the specific embodiments of the present invention and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0058] The following will describe in detail the technical solutions provided by each embodiment of the present invention with reference to the drawings.

[0059] Figure 1 It is a schematic flowchart of a method for fault detection of voltage-sharing capacitors in a converter valve in the present invention, specifically including the following steps:

[0060] S101, obtain the frequency response function of the voltage-sharing capacitors in the converter valve in the form of the complex frequency domain, select a linearly distributed conjugate complex pair as the initial poles within a preset frequency band, and define an auxiliary function according to the initial poles; the preset frequency band is the frequency band that needs to be vector-fitted.

[0061] Among them, the preset frequency band can be a frequency band of 1 kHz - 1 MHz.

[0062] Optionally, obtaining the frequency response function of the grading capacitor in the commutation valve in the complex frequency domain includes: obtaining the frequency response curve of the grading capacitor in the commutation valve and constructing the frequency response function of the grading capacitor in the complex frequency domain.

[0063] Applying a sinusoidal low-voltage sweep signal U at one end of the grading capacitor in the commutation valve in , and measuring the corresponding response voltage U at the other end o . The sweep frequency f ranges from 1 kHz to 500 kHz, with a step size of 1 kHz, and there are a total of 500 frequency points; according to the measured response voltage value and input voltage value at each frequency point, calculate the voltage ratio amplitude within the entire sweep frequency range, that is, the frequency response function of the grading capacitor in the commutation valve. Its calculation formula is as follows:

[0064] H(f i ) = |U o (f i ) / U in (f i )| (1)

[0065] Among them, the frequency response curve is the curve of the voltage ratio amplitude changing with frequency.

[0066] Writing the frequency response curve in the complex frequency domain form containing N orders, that is, the frequency response function f(s) in the complex frequency domain form containing N orders is:

[0067]

[0068] Among them, c n is the zero point of f(s), a n is the pole of f(s), d and h are real numbers, s is the representation of the complex frequency domain, and N is the order of f(s). c n , a n , d and h are unknown coefficients, c n and a n are real numbers or conjugate complex pairs.

[0069] Select conjugate complex pairs with a linear distribution as the initial poles within the frequency band [f min , f max that needs to be vector-fitted, and define the auxiliary function σ(s) as:

[0070]

[0071] Among them, is the pole of σ(s) and is the initial pole of f(s), is the zero point of σ(s).

[0072] S102. Determine the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve.

[0073] Optionally, determining the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve includes: multiplying the frequency response function by the auxiliary function to obtain a relational expression; obtaining a simplified form of the relational expression, and converting the simplified form of the relational expression into a linear form; expanding both sides of the linear form equation to obtain another representation form of the frequency response function; determining the order value of the frequency response function according to the another representation form of the frequency response function and the frequency response curve.

[0074] Specifically, the multiplication of the function f(s) and σ(s) can be expressed as follows:

[0075]

[0076] The rational approximation of the function σ(s) has the same poles as the rational approximation of σ(s)f(s). Multiplying formula (2) by formula (3) can obtain a relational expression, and the relational expression is:

[0077]

[0079] The simplified form of the relational expression is:

[0080] (σf) fit (s) = σ fit (s)f(s)(6)

[0081] Where (σf) fit and σ fit (s) are written in the form of zeros and poles, denoted as: Where z n represents the zero of (σf) fit , represents the zero of σ fit (s).

[0082] Considering each frequency point, convert the formula (6) corresponding to all frequency points into the linear form Ax = b; A is the coefficient matrix corresponding to each sampling frequency point, b is f(s) corresponding to each sampling frequency point, and x is

[0083] Express the left and right of Ax = b in the following form:

[0084]

[0085] Another representation of the frequency response function can be obtained as follows:

[0086]

[0087] At this time, the zeros of σ(s) fit (s) are the poles of f(s).

[0088] Optionally, according to another representation of the frequency response function and the frequency response curve, determine the order value of the frequency response function, including: obtaining the fitting curves of another representation of the frequency response function at different orders; determining the order corresponding to the minimum deviation between the fitting curve and the frequency response curve as the order value of the frequency response function.

[0089] Specifically, the order N can be set to vary from 1 to 100. As the order changes, observe the fitting situation. When the deviation between the frequency response curve obtained from formula (8) and the frequency response curve of the original measured equalizing capacitor is the smallest, select the order at this time as N.

[0090] S103, according to the initial poles, solve for the initial values of the coefficients of the frequency response function, and determine the zeros of the auxiliary function based on the initial values of the coefficients and the initial poles.

[0091] Optionally, according to the initial poles, solve for the initial values of the coefficients of the frequency response function, including: substituting the initial poles into Ax = b and solving to obtain the initial values of the coefficients x.

[0092] At this time, σ(s)f(s) and σ(s) can be obtained through formula (2). After obtaining the function of σ(s), find the zeros of σ(s); specifically, according to Ax = b, x, that is, the correlation coefficient of σ(s), can be obtained. Further, by setting its rational function expression to 0, its zeros can be obtained.

[0093] S104, with the condition that the zeros of the auxiliary function are equal to the poles of the auxiliary function, take the zeros of the auxiliary function as the poles of the frequency response function, and iteratively update the coefficients of the frequency response function to obtain the target coefficients of the frequency response function.

[0094] Take the zeros of σ(s) as the new poles of f(s), continue to solve Ax = b, and iterate until finally the zeros of σ(s) are equal to its poles (the true zeros of f(s)), as shown in formula (7). At this time, σ fit (s) = 1. At this time, the rational approximation formulas (3) and (4) of the frequency response function converge. According to c n , a n , d, and h, these coefficients can characterize the frequency response function f(s), and then the zero-pole values can be obtained according to f(s).

[0095] The c obtained when formulas (3) and (4) converge n , a n , d, and h are determined as the target coefficients of the frequency response function.

[0096] S105. Determine the target zeros and target poles of the frequency response function according to the target coefficients and order values of the frequency response function.

[0097] Based on the target coefficients and order values of the above frequency response function, determine the final frequency response function. Determine the zeros and poles according to this frequency response function, and respectively determine the zeros and poles as the target zeros and target poles.

[0098] S106. Determine the fault detection result of the voltage-sharing capacitor according to the target zeros and target poles.

[0099] In one embodiment, determining the fault detection result of the voltage-sharing capacitor according to the target zeros and target poles includes: inputting the target zeros and target poles into a pre-trained support vector machine model to obtain the fault detection result of the voltage-sharing capacitor.

[0100] Among them, the construction process of the support vector machine model includes: obtaining the frequency response curves of the converter valve voltage-sharing capacitor under different fault conditions; determining the target zeros and target poles of the corresponding frequency response functions under different fault conditions according to the frequency response curves of the converter valve voltage-sharing capacitor under different fault conditions; using the target zeros and target poles of the corresponding frequency response functions under different fault conditions as the data sample set; training according to the data sample set to obtain the support vector machine model.

[0101] It should be noted that the method for calculating the target zeros and target poles of the frequency response function corresponding to different fault conditions is the same as that for calculating the target zeros and target poles of the frequency response function in the above embodiment, and will not be elaborated in this embodiment.

[0102] Select the radial basis kernel function as the kernel function of the support vector machine (SVM). The range of the penalty coefficient c is [0.01, 1]. Establish an SVM classifier containing the kernel function. Use all the solved target zeros and poles as the data sample set, divide it into a training set and a test set according to 7:3. Input the training set into the SVM classifier for training, use the test set for testing, obtain the trained SVM model, and use the target zeros and poles obtained by vector matching of the frequency response function measured by the voltage-sharing capacitor in the converter valve as the input features of the SVM model to realize the detection of the fault of the voltage-sharing capacitor in the converter valve.

[0103] The present invention also provides a method for fault detection of the grading capacitors in a converter valve. Based on the measured frequency response function of the grading capacitors in the converter valve, through rational approximation and vector fitting, the zeros and poles of the frequency response function of the grading capacitors in the converter valve under different conditions are calculated, and these are used as the input features of the support vector machine model for fault detection of the grading capacitors in the converter valve; as Figure 2 shown, this embodiment includes the following steps:

[0104] S201, test and collect the frequency response function values of the grading capacitors of the converter valve under different fault conditions.

[0105] S202, write the frequency response function of the grading capacitors of the converter valve in the complex frequency domain form f(s) containing N orders.

[0106] S203, use conjugate complex pairs as the initial poles and define the function σ(s).

[0107] S204, deduce the relational expression between σ(s) and f(s) and determine the order of f(s).

[0108] S205, solve the coefficients of f(s), calculate the zeros of σ(s), which are the poles of f(s), consider all frequency points, and calculate all the zeros and poles of f(s) through iterative loops.

[0109] S206, calculate the zeros and poles under all fault conditions, use them as features, and construct a data sample set.

[0110] S207, construct an SVM model, use the data sample set as the model input features, train and test the model, and obtain the constructed SVM model.

[0111] S208, realize the fault detection of the grading capacitors in the converter valve through the SVM model.

[0112] This method extracts the true features of the grading capacitors in the converter valve, and can avoid problems such as insufficient and incomplete extraction of features caused by using data index methods and other methods when extracting the deviation of the frequency response function of the grading capacitors in the converter valve under different conditions, making the detection accuracy of the support vector machine model based on zeros and poles as feature inputs higher, and providing a more accurate method for fault detection of the grading capacitors in the converter valve.

[0113] When applying the method for fault detection of the grading capacitors in the converter valve provided by the present invention, it is not necessary to execute according to Figure 1 the order of the steps shown. The specific execution order of each step can be determined as needed, and the present invention does not limit this.

[0114] The above is a fault detection method for grading capacitors in a commutation valve provided by one or more embodiments of the present invention. Based on the same concept, the present invention also provides a corresponding fault detection device for grading capacitors in a commutation valve, as Figure 3 shown.

[0115] Figure 3 FIG. 6 is a schematic diagram of a fault detection device for grading capacitors in a commutation valve provided by the present invention. The device 300 includes:

[0116] An acquisition module 301, configured to acquire the frequency response function of the grading capacitor in the commutation valve in the complex frequency domain, select a conjugate complex pair with a linear distribution as the initial poles within a preset frequency band, and define an auxiliary function according to the initial poles; the preset frequency band is the frequency band that needs to perform vector fitting;

[0117] A first determination module 302, configured to determine the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve;

[0118] A second determination module 303, configured to solve the initial coefficient value of the frequency response function according to the initial poles, and determine the zero points of the auxiliary function according to the coefficient initial value and the initial poles;

[0119] A third determination module 304, configured to use the zero points of the auxiliary function being equal to the poles of the frequency response function as a constraint condition, use the zero points of the auxiliary function as the poles of the frequency response function, and iteratively update the coefficients of the frequency response function to obtain the target coefficients of the frequency response function;

[0120] A fourth determination module 305, configured to determine the target zero points and target poles of the frequency response function according to the target coefficients and the order value of the frequency response function;

[0121] A fifth determination module 306, configured to determine the fault detection result of the grading capacitor according to the target zero points and target poles.

[0122] For the specific limitations on the fault detection device for grading capacitors in a commutation valve, reference can be made to the limitations on the fault detection method for grading capacitors in a commutation valve in the above text, which will not be elaborated here. Each module in the above-mentioned fault detection device for grading capacitors in a commutation valve can be implemented in whole or in part by software, hardware, and their combination. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0123] The present invention also provides a computer-readable storage medium, which stores a computer program, and the computer program can be used to execute the above Figure 1A fault detection method for voltage-sharing capacitors in a commutation valve is provided.

[0124] The present invention also provides Figure 4 a schematic structural diagram of the computer device shown in, as Figure 4 shown, at the hardware level, the computer device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory. Of course, it may also include other hardware required for other services. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to implement the above Figure 1 fault detection method for voltage-sharing capacitors in a commutation valve provided.

[0125] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memories. The non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical memory, etc. The volatile memory can include random access memory (RAM) or an external cache. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0126] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope recorded by the present invention.

Claims

1. A fault detection method for voltage-sharing capacitors in a converter valve, characterized in that, Including: Obtain the frequency response function of the voltage-sharing capacitor in the commutation valve in the complex frequency domain form, select a pair of conjugate complex numbers with a linear distribution as the initial poles within a preset frequency band, and define an auxiliary function according to the initial poles; the preset frequency band is the frequency band that needs to be vector-fitted; Determine the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve; Solve the initial values of the coefficients of the frequency response function according to the initial poles, and determine the zeros of the auxiliary function according to the initial values of the coefficients and the initial poles; Taking the zeros of the auxiliary function being equal to the poles of the frequency response function as a constraint condition, use the zeros of the auxiliary function as the poles of the frequency response function, and iteratively update the coefficients of the frequency response function to obtain the target coefficients of the frequency response function; Determine the target zeros and target poles of the frequency response function according to the target coefficients of the frequency response function and the order value; Determine the fault detection result of the voltage-sharing capacitor according to the target zeros and the target poles; 2. The method according to claim 1, wherein The frequency response function f(s) in the complex frequency domain form is: where c n is the zero of f(s), a n is the pole of f(s), d and h are coefficients, s is the representation in the complex frequency domain, and N is the order of f(s).

3. The method according to claim 2, wherein The auxiliary function σ(s) is: Among them, is a pole of σ(s) and an initial pole of f(s), is a zero of σ(s).

4. The method according to claim 3, wherein The determining the order value of the frequency response function according to the frequency response function, the auxiliary function, and the frequency response curve includes: Multiply the frequency response function by the auxiliary function to obtain a relational expression: The relational expression is Obtain the shorthand form of the relational expression, and the shorthand form of the relational expression is (σf) fit (s) = σ fit (s)f(s); where, (σf) fit and σ fit (s) is written in the form of zeros and poles as: z n represents the zero of (σf) fit and represents the zero of σ fit (s); Convert the abbreviated form of the relational expression into the linear form Ax = b; A is the coefficient matrix corresponding to each sampling frequency point, and b is f(s) corresponding to each sampling frequency point. Expand both sides of the equation in linear form to obtain another representation of the frequency response function: Among them, z n and respectively represent the zeros of (σf) fit and σ fit (s); Determine the order value of the frequency response function according to another representation of the frequency response function and the frequency response curve.

5. The method according to claim 4, characterized in that, The determining the order value of the frequency response function according to another representation of the frequency response function and the frequency response curve includes: Obtain the fitting curves of another representation of the frequency response function at different orders; Determine the order value corresponding to the minimum deviation between the fitting curve and the frequency response curve as the order value of the frequency response function.

6. The method according to claim 4, characterized in that, The solving the initial values of the coefficients of the frequency response function according to the initial poles includes: Substitute the initial poles into Ax = b and solve to obtain the initial values of the coefficients x.

7. The method according to claim 1, wherein The determining the fault detection result of the voltage-sharing capacitor according to the target zeros and the target poles includes: Input the target zeros and the target poles into a pre-trained support vector machine model to obtain the fault detection result of the voltage-sharing capacitor.

8. The method according to claim 7, characterized in that, The construction process of the support vector machine model includes: Obtain the frequency response curves of the voltage-sharing capacitors of the commutation valve under different fault conditions; Determine the target zeros and target poles of the corresponding frequency response functions under different fault conditions according to the frequency response curves of the voltage-sharing capacitors of the commutation valve under different fault conditions; Use the target zeros and target poles of the corresponding frequency response functions under different fault conditions as a data sample set; Train according to the data sample set to obtain the support vector machine model.

9. The method according to claim 8, wherein The kernel function of the support vector machine is a radial basis function, and the range of the penalty coefficient of the support vector machine is [0.01, 1].

Images