Power distribution network single-phase earth fault line selection method based on calculation decomposition quality factor improved VMD

Abstract

The invention discloses a power distribution network single-phase earth fault line selection method based on VMD improved by calculating a decomposition quality factor. The method comprises the following steps: step 1, collecting a transient zero-sequence voltage of a power distribution network model bus; 2, collecting transient zero-sequence current signals of the power distribution network model 0.1 second before and after the fault occurs; 3, calculating a decomposition quality factor, and determining an optimal decomposition mode number K; step 4, adaptively decomposing transient zero-sequence current signals of each outgoing line of the power distribution network model by using the improved VMD algorithm; 5, calculating an energy entropy and a polarity measurement factor of each outgoing line of the power distribution network model; and 6, comparing the energy entropy and the polarity measurement factor of each outgoing line of the power distribution network model, and selecting a fault line through double criteria. According to the method, the accuracy and efficiency of fault line selection of the power distribution network are improved, and effective identification of fault and non-fault lines under the conditions of different fault distances and different grounding resistances and in different neutral point grounding modes is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and more particularly to a single-phase grounding fault line selection method for a distribution network based on improving VMD by calculating and decomposing a quality factor. Background Art

[0002] Most of my country's 3-66kV distribution networks are small current grounding systems, which are usually divided into two types: neutral point ungrounded and neutral point grounded through arc suppression coils. About 80% of faults are single-phase grounding faults. With the rapid development of the economy, the demand for electricity for production and construction has continued to increase, the transformation of urban and rural distribution networks has accelerated, the scale of distribution systems has expanded rapidly, and the number and distance of cables and overhead lines have increased significantly compared with the past. When a single-phase grounding fault occurs, the voltage of the faulty phase will decrease, and the voltage of the other two phases will increase, causing the insulation of the distribution equipment to be under high voltage for a long time, which is easy to cause insulation aging or even breakdown, and may develop into a more serious phase-to-phase short-circuit fault, resulting in a three-phase short-circuit shock; the increase in ground potential also increases the risk of electric shock, especially in an environment with low ground resistance, which is easy to cause safety accidents. Although a single-phase grounding fault is not as serious as a phase-to-phase short circuit, long-term operation will still affect the safety and stability of the distribution network, so it should be detected and handled as soon as possible.

[0003] For systems with neutral points grounded through arc suppression coils, due to the compensation effect of inductive current, the zero-sequence current value is further reduced during single-phase grounding faults, making it difficult to collect, which places higher requirements on the detection of faulty equipment. Traditional methods for selecting lines for small current grounding faults are difficult to achieve high accuracy in complex distribution networks and are prone to misjudgment or missed judgments, especially in weakly grounded systems. These methods have low sensitivity and reliability, so new algorithms and technologies are needed to improve the accuracy of fault identification. Small current grounding fault line selection technology has become an important research topic and hotspot at home and abroad. Summary of the invention

[0004] The purpose of this method is to propose a single-phase grounding fault line selection method for distribution network based on calculating the decomposed quality factor to improve VMD, so as to solve the current technical problems of low efficiency and low accuracy of fault line selection.

[0005] To achieve the above object, the present invention comprises the following steps:

[0006] Step 1: collect the transient zero-sequence voltage of the busbar of the distribution network model, and determine whether a single-phase grounding fault has occurred by judging the transient zero-sequence voltage of the system. If it is greater than 20% of the rated voltage, start the fault line selection process.

[0007] Step 2: Collect the transient zero-sequence current signal of the distribution network model 0.1 seconds before and after the fault occurs.

[0008] Step 3: Calculate the decomposition quality factor and determine the optimal decomposition mode number K.

[0009] Step 4: Use the improved VMD algorithm to adaptively decompose the transient zero-sequence current signal of each outgoing line of the distribution network model to obtain K IMF components.

[0010] Step 5: Calculate the energy entropy and polarity measurement factor of each outgoing line of the distribution network model.

[0011] Step 6: Compare the energy entropy and polarity measurement factor of each outgoing line of the distribution network model. Select a line as the reference line. When a line satisfies the maximum energy entropy value and the polarity of the polarity measurement factor of the line is different from that of other lines, the line is determined to be a faulty line; otherwise, it is determined to be a distribution network bus fault.

[0012] The present invention is also characterized in that, in step 3, the process of calculating the decomposition quality factor is as follows:

[0013] a. The difference coefficient C is used to characterize the difference between the IMF components, and its definition is as follows:

[0014]

[0015] Where H1, H2, ..., H K Respectively represent the information entropy of the corresponding decomposed K IMF values, is the mean information entropy of the decomposed K IMF components. From the above formula, we can see that the closer the information entropy of each IMF component is to the mean information entropy, the smaller the C value is, the smaller the difference between the IMF components is, and the decomposition effect is not obvious. To ensure a good decomposition effect, the C value should be as large as possible.

[0016] b. The error coefficient Δ is used to characterize the error between the original signal and the reconstructed signal. Its definition is as follows:

[0017] Δ=sum{abs[f-(u1+u2+···+u K )]}

[0018] Where f represents the original signal, u1,u2,...,u K Represents the IMF components after decomposition. The smaller the value of the error coefficient Δ is, the closer the original signal is to the reconstructed signal, and the better the decomposition effect is.

[0019] c. The VMD decomposition quality factor can be defined by the difference coefficient and error coefficient as follows:

[0020]

[0021] In order to achieve the best effect of VMD decomposition, the decomposition quality factor must be maximized, that is, the characteristics of the decomposed component signals are quite different while the reconstructed signal is close to the original signal. To reduce the amount of calculation, the maximum K value of the program is set to 10. By calculating the size of the decomposition quality factor Q under each K value, the largest Q value is selected, and its corresponding K value is the optimal number of decomposition modes.

[0022] In step 4, the VMD algorithm flow is as follows;

[0023] a. Construction of variational problem

[0024] The original signal f(t) is decomposed into K intrinsic modal components, and the sum of the broadband of the K IMF components is minimum and the sum of the K IMF components is equal to f(t). The variational model with constraints is shown as follows:

[0025]

[0026] In the formula, u k (t)={u1,u2,···,u k} represents the set of all modal components, ω k ={ω1,ω2,···,ω k} represents the center frequency of each IMF component, δ(t) represents the unit pulse function, and “*” represents the convolution calculation.

[0027] b. Solution of variational problems

[0028] By introducing the Lagrangian quadratic penalty factor α and the Lagrangian multiplication operator λ, the variational problem is converted into an unconstrained variational form, and the obtained augmented Lagrangian function is shown as follows:

[0029]

[0030] The minimization problem is transformed into the saddle point problem of the augmented Lagrangian function. The alternating direction multiplier method is used to calculate the parameters Perform multiple iterations to solve each modal component Update it with:

[0031]

[0032] Where ω k =ω k+1 , Fourier transform of the above equation can be obtained Mathematical expression in the frequency domain:

[0033]

[0034] The center frequency of each IMF is obtained in the same way. and the Lagrange multiplication operator The updated expression in the frequency domain is as follows:

[0035]

[0036] Set parameters The initial value of n is 0, and the parameter u is updated iteratively. k ,ω k ,λ, until the convergence condition shown in the following formula is met, the iteration is stopped.

[0037]

[0038] In step 5, the selected fault feature energy entropy and polarity measurement factor are as follows:

[0039] a. Energy entropy is an extension of entropy in the energy domain. Entropy describes the degree of confusion of a system. Information entropy is a measure of information uncertainty. Its definition was proposed by Shannon as follows:

[0040] H(x)=E[I(x i )]=-∑p i logp i

[0041] The decomposed K IMF components are represented by u1(t),u2(t),···,u k (t) represents that the energy definition of each IMF is as follows:

[0042]

[0043] Since the IMF components are orthogonal, the energy of the signal is the sum of the energies of all IMF components. Based on the definition of information entropy, the energy entropy formula is as follows:

[0044]

[0045] b. The polarity measurement factor α is measured by selecting a reference route, and the high-frequency components of other routes are respectively inner-producted with the selected reference route. Its definition is as follows:

[0046]

[0047] i x (t) is the high frequency component of line x, and i0(t) is the high frequency component of the selected reference line.

[0048] Compared with the prior art, the present invention has the following beneficial effects: effectively improving the efficiency and accuracy of single-phase grounding fault line selection. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 is a flow chart of the fault line selection method of the present invention; DETAILED DESCRIPTION

[0050] The technical solution of the present invention is described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0051] Step 1: collect the transient zero-sequence voltage of the busbar of the distribution network model, and determine whether a single-phase grounding fault has occurred by judging the transient zero-sequence voltage of the system. If it is greater than 20% of the rated voltage, start the fault line selection process.

[0052] Step 2: Collect the transient zero-sequence current signal of the distribution network model 0.1 seconds before and after the fault occurs.

[0053] Step 3: Calculate the decomposition quality factor and determine the optimal decomposition mode number K.

[0054] In step 3, the process of calculating the decomposition quality factor is as follows:

[0055] a. The difference coefficient C is used to characterize the difference between the IMF components, and its definition is as follows:

[0056]

[0057] Where H1, H2, ..., H K Respectively represent the information entropy of the corresponding decomposed K IMF values, is the mean information entropy of the decomposed K IMF components.

[0058] b. The error coefficient Δ is used to characterize the error between the original signal and the reconstructed signal. Its definition is as follows:

[0059] Δ=sum{abs[f-(u1+u2+···+u K )]}

[0060] Where f represents the original signal, u1,u2,...,u K Represents the decomposed IMF components.

[0061] c. The VMD decomposition quality factor can be defined by the difference coefficient and error coefficient as follows:

[0062]

[0063] Step 4: Use the improved VMD algorithm to adaptively decompose the transient zero-sequence current signal of each outgoing line of the distribution network model to obtain K IMF components.

[0064] In step 4, the VMD algorithm flow is as follows;

[0065] a. Construction of variational problem

[0066] The original signal f(t) is decomposed into K intrinsic modal components, and the sum of the broadband of the K IMF components is minimum and the sum of the K IMF components is equal to f(t). The variational model with constraints is shown as follows:

[0067]

[0068] In the formula, u k (t)={u1,u2,···,u k} represents the set of all modal components, ω k ={ω1,ω2,···,ω k} represents the center frequency of each IMF component, δ(t) represents the unit pulse function, and “*” represents the convolution calculation.

[0069] b. Solution of variational problems

[0070] By introducing the Lagrangian quadratic penalty factor α and the Lagrangian multiplication operator λ, the variational problem is converted into an unconstrained variational form, and the obtained augmented Lagrangian function is shown as follows:

[0071]

[0072] The minimization problem is transformed into the saddle point problem of the augmented Lagrangian function. The alternating direction multiplier method is used to calculate the parameters Perform multiple iterations to solve each modal component Update it with:

[0073]

[0074] Where ω k =ω k+1 , Fourier transform of the above equation can be obtained Mathematical expression in the frequency domain:

[0075]

[0076] The center frequency of each IMF is obtained in the same way. and the Lagrange multiplication operator The updated expression in the frequency domain is as follows:

[0077]

[0078] Set parameters The initial value of n is 0, and the parameter u is updated iteratively. k ,ω k ,λ, until the convergence condition shown in the following formula is met, the iteration is stopped.

[0079]

[0080] Step 5: Calculate the energy entropy and polarity measurement factor of each outgoing line of the distribution network model.

[0081] In step 5, the selected fault feature energy entropy and polarity measurement factor are as follows:

[0082] a. Energy entropy is an extension of entropy in the energy domain. Entropy describes the degree of confusion of a system. Information entropy is a measure of information uncertainty. Its definition was proposed by Shannon as follows:

[0083] H(x)=E[I(x i )]=-∑p i logp i

[0084] The decomposed K IMF components are represented by u1(t),u2(t),···,u k (t) represents that the energy definition of each IMF is as follows:

[0085]

[0086] Since the IMF components are orthogonal, the energy of the signal is the sum of the energies of all IMF components. Based on the definition of information entropy, the energy entropy formula is as follows:

[0087]

[0088] b. The polarity measurement factor α is measured by selecting a reference route, and the high-frequency components of other routes are respectively inner-producted with the selected reference route. Its definition is as follows:

[0089]

[0090] i x (t) is the high frequency component of line x, and i0(t) is the high frequency component of the selected reference line.

[0091] Step 6: Compare the energy entropy and polarity measurement factor of each outgoing line of the distribution network model. Select line 1 as the reference line. When a line satisfies the maximum energy entropy value and the polarity of the polarity measurement factor of the line is different from that of other lines, the line is determined to be a faulty line; otherwise, it is determined to be a distribution network bus fault.

Claims

1. A method for selecting a single-phase grounding fault line in a distribution network based on calculating and decomposing quality factors to improve VMD, characterized in that: The steps include: Step 1: collect the transient zero-sequence voltage of the busbar of the distribution network model, and determine whether a single-phase grounding fault has occurred by judging the transient zero-sequence voltage of the system. If it is greater than 20% of the rated voltage, start the fault line selection process. Step 2: Collect the transient zero-sequence current signal of the distribution network model 0.1 seconds before and after the fault occurs. Step 3: Calculate the decomposition quality factor and determine the optimal decomposition mode number K. Step 4: Use the improved VMD algorithm to adaptively decompose the transient zero-sequence current signal of each outgoing line of the distribution network model to obtain K IMF components. Step 5: Calculate the energy entropy and polarity measurement factor of each outgoing line of the distribution network model. Step 6: Compare the energy entropy and polarity measurement factor of each outgoing line of the distribution network model. Select line 1 as the reference line. When a line satisfies the maximum energy entropy value and the polarity of the polarity measurement factor of the line is different from that of other lines, the line is determined to be a fault line. Otherwise, it is determined to be a distribution network bus fault.

2. According to the method for selecting a single-phase grounding fault line in a distribution network based on improving VMD by calculating decomposition quality factor according to claim 1, in step 3, the process of calculating the decomposition quality factor is as follows: a. The difference coefficient C is used to characterize the difference between the IMF components, and its definition is as follows: Where H1, H2, ..., H K Respectively represent the information entropy of the corresponding decomposed K IMF values, is the mean information entropy of the decomposed K IMF components. b. The error coefficient Δ is used to characterize the error between the original signal and the reconstructed signal. Its definition is as follows: Δ=sum{abs[f-(u1+u2+···+u K )]} Where f represents the original signal, u1,u2,...,u K Represents the decomposed IMF components. c. The VMD decomposition quality factor can be defined by the difference coefficient and error coefficient as follows: In order to optimize the effect of VMD decomposition, the decomposition quality factor must be maximized, that is, the characteristics of the decomposed component signals are quite different while the reconstructed signal is close to the original signal.

3. According to the method for selecting a single-phase grounding fault line in a distribution network based on improving VMD by calculating and decomposing the quality factor according to claim 1, in step 4, the VMD algorithm flow is as follows; a. Construction of variational problem The original signal f(t) is decomposed into K intrinsic modal components, and the sum of the broadband of the K IMF components is minimum and the sum of the K IMF components is equal to f(t). The variational model with constraints is shown as follows: In the formula, u k (t)={u1,u2,···,u k } represents the set of all modal components, ω k ={ω1,ω2,···,ω k } represents the center frequency of each IMF component, δ(t) represents the unit pulse function, and "*" represents the convolution calculation. b. Solution of variational problems By introducing the Lagrangian quadratic penalty factor α and the Lagrangian multiplication operator λ, the variational problem is converted into an unconstrained variational form, and the obtained augmented Lagrangian function is shown as follows: The minimization problem is transformed into the saddle point problem of the augmented Lagrangian function. The alternating direction multiplier method is used to calculate the parameters Perform multiple iterations to solve each modal component Update it with the following formula: Where ω k =ω k+1 , Fourier transform of the above equation can be obtained Mathematical expression in the frequency domain: The center frequency of each IMF is obtained in the same way. and the Lagrange multiplication operator The updated expression in the frequency domain is as follows: Set parameters The initial value of n is 0, and the parameter u is updated iteratively. k ,ω k ,λ, until the convergence condition shown in the following formula is met, the iteration is stopped.

4. According to the method for selecting a single-phase grounding fault line in a distribution network based on calculating and decomposing the quality factor to improve VMD according to claim 1, in step 5, the selected fault feature energy entropy and polarity measurement factor are as follows: a. Energy entropy is an extension of entropy in the energy domain. Entropy describes the degree of confusion of a system. Information entropy is a measure of information uncertainty. Its definition was proposed by Shannon as follows: H(x)=E[I(x i )]=-∑p i logp i The decomposed K IMF components are represented by u1(t),u2(t),···,u k (t) represents that the energy definition of each IMF is as follows: Since the IMF components are orthogonal, the energy of the signal is the sum of the energies of all IMF components. Based on the definition of information entropy, the energy entropy formula is as follows: b. The polarity measurement factor α is measured by selecting a reference route, and the high-frequency components of other routes are respectively inner-producted with the selected reference route. Its definition is as follows: i x (t) is the high frequency component of line x, and i0(t) is the high frequency component of the selected reference line.

Images