Distributed power grid disturbance location method based on improved high-frequency energy method and matrix method

Abstract

The present invention discloses a method for locating disturbances in a distributed power grid based on an improved high-frequency energy method and a matrix method. The method comprises the following steps: decomposing and reconstructing disturbance data to obtain high-frequency power and high-frequency energy; determining the relative direction between the disturbance source and the monitoring point by using an improved high-frequency energy method directional criterion; forming a directional discrimination information matrix and a system structure information matrix, establishing a particle swarm optimization model, and limiting the particle swarm search motion space and particle motion span; constructing an evaluation function that is tolerant to directional discrimination information and is suitable for power grids containing distributed energy; and utilizing a particle swarm algorithm for iterative search to determine the optimal particle based on the evaluation function to accurately locate the disturbance source. By combining the improved high-frequency energy method directional criterion with an optimization matrix, the present invention improves the accuracy of locating power quality disturbance sources in power grids containing distributed energy.

Description

Technical Field

[0001] The present invention relates to the field of power quality positioning of a distributed energy grid, and in particular to a distributed power grid disturbance positioning method based on an improved high-frequency energy method and a matrix method. Background Art

[0002] With the integration of numerous renewable energy power generation projects into the grid as distributed energy sources, the structure, power flow, and operation of the power grid have been impacted. This has not only introduced new power quality disturbances but also increased the difficulty of disturbance analysis. Accurately locating the source of power quality disturbances in distribution networks is crucial for quickly resolving power quality issues, reducing economic losses, and determining responsibility for incidents.

[0003] Existing disturbance source localization algorithms rely too much on the accuracy of direction discrimination information, and the direction discrimination and positioning accuracy are low in power grids containing distributed energy. Summary of the Invention

[0004] The purpose of the present invention is to provide a distributed power grid disturbance location method based on an improved high-frequency energy method and a matrix method, thereby more accurately locating power quality disturbances in a power grid containing distributed energy.

[0005] The technical solution to achieve the purpose of the present invention is: a distributed power grid disturbance location method based on an improved high-frequency energy method and a matrix method, comprising the following steps:

[0006] Step 1: perform wavelet packet discretization decomposition and reconstruction on the original disturbance signal data, and use the reconstruction information to obtain the high-frequency power and high-frequency energy in the full signal to obtain the reconstructed high-frequency signal;

[0007] Step 2: Determine the time when the disturbance occurs based on the reconstructed high-frequency signal, and determine the relative direction between the disturbance source and the monitoring point by using the new direction criterion of the improved high-frequency energy method;

[0008] Step 3: Form the direction discrimination information matrix and the system structure information matrix, establish the particle swarm optimization model, and limit the particle swarm search motion space and particle motion span;

[0009] Step 4: Construct an evaluation function that is tolerant to directional discrimination information and applicable to distributed energy grids. The smaller the value of the evaluation function, the better the corresponding result.

[0010] Step 5: Use the particle swarm algorithm to iteratively search in space, determine the optimal particle position based on the evaluation function, and locate the disturbance source in combination with the system structure information matrix.

[0011] Compared with the existing technology, the present invention has the following significant advantages: (1) It proposes to use the sign of the sudden change of high-frequency energy at the disturbance moment as the disturbance direction criterion, thereby improving the accuracy of the direction discrimination of the disturbance source in the power grid containing distributed energy; (2) It designs an evaluation function including the credibility information of the direction discrimination and the DG compensation, which can not only reduce the negative impact of the distributed energy grid connection on the disturbance location, but also has the fault tolerance of the direction discrimination information. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 This is a flow chart of the distributed power grid disturbance location method based on the improved high-frequency energy method and matrix method of the present invention.

[0013] Figure 2 The following is a typical distribution network structure diagram containing two distributed energy sources.

[0014] Figure 3 Schematic diagram of high-frequency energy mutation at a certain monitoring point at the moment of disturbance.

[0015] Figure 4 Schematic diagram of high-frequency energy fluctuations at a certain monitoring point at the moment of disturbance.

[0016] Figure 5 Schematic diagram of the change in the optimal evaluation function value during the algorithm's iterative search process. DETAILED DESCRIPTION

[0017] like Figure 1 As shown, the distributed power grid disturbance location method based on the improved high-frequency energy method and the matrix method includes the following steps:

[0018] Step 1: perform wavelet packet discretization decomposition and reconstruction on the original disturbance signal data, and use the reconstruction information to obtain the high-frequency power and high-frequency energy in the full signal to obtain the reconstructed high-frequency signal;

[0019] Step 2: Determine the time when the disturbance occurs based on the reconstructed high-frequency signal, and determine the relative direction between the disturbance source and the monitoring point by using the new direction criterion of the improved high-frequency energy method;

[0020] Step 3: Form the direction discrimination information matrix and the system structure information matrix, establish the particle swarm optimization model, and limit the particle swarm search motion space and particle motion span;

[0021] Step 4: Construct an evaluation function that is tolerant to directional discrimination information and applicable to distributed energy grids. The smaller the value of the evaluation function, the better the corresponding result.

[0022] Step 5: Use the particle swarm algorithm to iteratively search in space, determine the optimal particle position based on the evaluation function, and locate the disturbance source in combination with the system structure information matrix.

[0023] Furthermore, in step 1, the specific method for obtaining the high-frequency power and high-frequency energy of the signal is:

[0024] The original disturbance voltage and current signals are decomposed into three layers of wavelet packets and their high-frequency components are reconstructed. The high-frequency power is obtained by calculating the voltage high-frequency component and the current high-frequency component:

[0025] p h =u h i h (1)

[0026] Where p h is the high frequency power, u h is the high-frequency component of voltage, i h is the high frequency component of the current.

[0027] Integrating the high-frequency power gives the high-frequency energy:

[0028] W h =∫p h dt (2)

[0029] Where W h It is high frequency energy.

[0030] Furthermore, in step 2, the specific method for determining the direction of the disturbance source using the improved high-frequency energy method is:

[0031] The theoretical premise of the high-frequency energy method is to split the system experiencing power quality disturbances into a superposition of an ideal three-phase system and a high-frequency disturbance source. When a disturbance occurs, all high-frequency components within the system are emitted by the disturbance source and propagate to both sides of the line. Unlike ideal steady-state three-phase systems, high-frequency components are always present in actual power grid systems. When distributed energy resources are connected to the grid, the operational characteristics of these resources dictate that these high-frequency components within the system will persist and fluctuate continuously. In this case, using the sign of the final value of the high-frequency energy as the criterion for direction determination may lead to misjudgment due to the influence of the high-frequency components generated by the distributed energy resources within the system.

[0032] The improved high-frequency energy method observes the high-frequency energy in a short period of time before and after the disturbance occurs. The sign of the sudden change in the high-frequency energy value before and after the disturbance is used to determine the direction of the disturbance source. If the sign of the sudden change in the high-frequency energy value at the moment of the disturbance is positive, it means that the disturbance source is upstream of the monitoring point; if the sign of the sudden change is negative, it means that the disturbance source is downstream of the monitoring point.

[0033] Furthermore, in step 3, a direction discrimination information matrix and a system structure information matrix are formed, a particle swarm optimization model is established, and the particle swarm search motion space and particle motion span are restricted, as follows:

[0034] (3.1) Determine the directional discrimination information matrix D in the system based on the directional discrimination information returned by each monitoring point in the system when the disturbance occurs. m×1 , where m is the total number of monitors in the system:

[0035] D m×1 =[d1…d m ] T (3)

[0036] Where, d k The value of depends on the judgment result of the kth monitoring point. When the disturbance source is located upstream of the monitoring point, the judgment result is regular. k =1, otherwise d k =-1.

[0037] (3.2) The structural information matrix C is formed based on the topological structure information of the distribution network and the location information of the monitoring instruments in the system. l×m , where l is the number of line segments in the system:

[0038]

[0039] Where c ij Indicates the positional relationship between the line segment and the monitoring instrument in the distribution network. If the i-th line segment is located in the downstream direction c of the j-th monitoring instrument, ij =-1, otherwise c ij =1.

[0040] This matrix representation method is only applicable to unidirectional radial networks. In order to ensure power supply reliability and quality, current systems mostly operate in a bidirectional power supply or ring network state. In this case, analogous to the line relay protection system, it is only necessary to install monitors at both ends of the line to split the network into two radial networks during analysis.

[0041] (3.3) Establish a particle swarm optimization model. The optimal position of the particle swarm is the optimal solution to the positioning problem. The particles move in the specified solution space at a specified speed. The entire particle swarm will be continuously updated as the algorithm operation progresses. Because the required solution is the direction discrimination matrix corresponding to the disturbance at a certain point in the system, let the algorithm particle be the direction discrimination information matrix D m×1 Matrix particles X with the same structure i , where i represents the particle number:

[0042] X i =[x1…x m ] T (5)

[0043] (3.4) To shorten the particle search time and increase the convergence rate, the matrix particles are restricted in their possible positions so that the positions they can reach within the motion space correspond to the direction discriminant matrix corresponding to the disturbance source located on a certain line segment. The matrix particle's movement span is limited to 1, meaning that after each iteration, the matrix moves to the direction discriminant matrix corresponding to the previous line segment (or the next one), ensuring global traversal and avoiding local convergence.

[0044] Furthermore, in step 4, the specific method of constructing the evaluation function is:

[0045] Because the direction discrimination information may contain errors, the credibility of the direction discrimination information needs to be considered. An evaluation function with fault tolerance for direction discrimination information and applicable to distributed energy grids is constructed from multiple aspects:

[0046] (1) The probability of misjudgment of the monitor in actual system operation is very small, so the matrix particles and the direction discrimination information matrix D m×1 The greater the difference, the greater the evaluation value and the worse the evaluation result.

[0047] (2) The farther the distance between the monitoring point and the disturbance source, the smaller the high-frequency disturbance energy value transmitted to the location, and the smaller the high-frequency energy mutation, the more susceptible it is to interference such as noise and misjudgment. In this case, the credibility of the direction discrimination information is low. Taking the maximum value of all high-frequency energy mutations monitored in the system as the benchmark, the ratio of other monitoring values to the benchmark is used to represent the distance coefficient between the monitoring point and the disturbance source. The farther the distance, the smaller the coefficient, indicating that the credibility of the direction discrimination information of the monitoring point is lower. The calculation result is recorded as α. When the matrix particle X i and the disturbance direction discriminant matrix D m×1 When the difference position corresponding distance coefficient α is small (α≤0.3), the evaluation value is compensated.

[0048] (3) Disturbance direction discrimination is based on the sign of the sudden change in high-frequency energy within a short period of time after the disturbance occurs. However, if a large fluctuation in high-frequency energy occurs within a short period of time after the disturbance occurs, it means that the positive and negative polarity of the disturbance power fluctuates seriously during the integration process. In this case, the credibility of the direction discrimination information is low. Let the change amplitude be the ratio of the abnormal change amplitude of high-frequency energy within a short period of time after the disturbance occurs to the peak value ΔW. When the matrix particle X i and the disturbance direction discriminant matrix D m×1 When the difference bit corresponding change ratio ΔW exceeds 50%, the evaluation value is compensated.

[0049] (4) On the grid-connected lines with distributed energy, high-frequency energy analysis is easily affected by distributed energy. When the matrix particle X i and the disturbance direction discriminant matrix D m×1 When the line corresponding to the difference position is a grid-connected line containing distributed energy, the evaluation value is compensated.

[0050] (5) In the actual system, virtual monitoring instruments are set up in some line sections to reduce costs. The status information of the virtual monitoring instruments is derived from the status of other monitoring instruments on the same bus. Because there are estimation errors in the inference process, when the matrix particle X i and the disturbance direction discriminant matrix D m×1 When the line corresponding to the difference bit is a virtual monitor line, the evaluation value is compensated.

[0051] To sum up, the evaluation function is constructed as follows:

[0052]

[0053] Wherein, ω1, ω2, ω3, ω4, and ω5 are compensation coefficients with a value range of (0 to 5), and the specific value depends on the importance of the influencing factor. Function f1(x, α) returns -1 when x is non-zero and the α value meets the judgment condition (α ≤ 0.3), returns 1 when x is non-zero and the α value meets the judgment condition (α ≥ 0.7), and otherwise returns 0. Function f2(x, ΔW) returns 1 when x is non-zero and the ΔW value meets the judgment condition (ΔW ≥ 0.5), and otherwise returns 0. Function f3(x, k) returns 1 when x is non-zero and k corresponds to a grid-connected line containing distributed energy resources, and otherwise returns 0. Function f4(x, k) returns 1 when x is non-zero and k corresponds to a virtual monitoring instrument line, and otherwise returns 0.

[0054] Furthermore, in step 5, the specific method of iterative search using the particle swarm algorithm to accurately locate the disturbance source is:

[0055] The particle swarm iterative search rule is:

[0056]

[0057]

[0058] Where, represents the velocity of the matrix particle at the k+1th iteration; and are the position and velocity of the matrix particle at the kth iteration respectively; ω represents the inertia factor, c1 and c2 represent the acceleration factor, r1 k 、 and Assign a random number between 0 and 1; represents the optimal value of the individual particle at the kth iteration; Represents the optimal value of the particle swarm at the kth iteration; is the threshold function; D L(i+1)It represents the direction discrimination matrix corresponding to the disturbance source located at the i+1th line segment, D L(i-1) It represents the direction discrimination matrix corresponding to the disturbance source located in the i-1th line segment.

[0059] The threshold function is used to limit the direction of movement of matrix particles (upward or downward) according to the particle velocity output threshold. At the same time, in order to avoid the saturation problem of the threshold function, the relationship between the threshold function and the particle velocity is set as follows:

[0060]

[0061] During the iteration process, the individual optimal value of the particle and the global optimal value are continuously updated. When the number of iterations reaches the set maximum value, the matrix particle corresponding to the global optimal value is output. The line segment corresponding to the matrix particle is the disturbance source positioning result.

[0062] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0063] Example

[0064] In order to verify the effectiveness of the solution of the present invention, Figure 2 The following simulation experiment is conducted on a typical structure distribution network with two distributed energy sources. When the disturbance occurs in the eighth line segment L8 in the distribution network, the parameters in the algorithm are set as follows:

[0065] The particle swarm size of the algorithm is 3, the maximum number of iterations is 20, and the inertia factor of the particle swarm is set to ω = 0.7, the acceleration factors c1 = 0.5, c2 = 0.5, and the evaluation function compensation coefficients ω1 = 1, ω2 = 1.5, ω3 = 0.5, ω4 = 0.15, ω5 = 0.1.

[0066] Follow steps 1 and 2 to obtain the high-frequency energy when the disturbance occurs at monitor 2, such as Figure 3 As shown in the figure, because line L2 is farther from the disturbance source, the high-frequency disturbance energy is smaller. Although reverse transmission of high-frequency energy occurs when the disturbance occurs, the final value remains positive. The original high-frequency energy method misjudges the direction of the disturbance. The high-frequency energy mutation sign criterion proposed in this method can accurately determine the direction of the disturbance source.

[0067] According to step 3, the disturbance monitoring direction discrimination information matrix is obtained by combining formula (3) with the measured data as follows:

[0068] D m×1 =[1-1111-1-1-111111] T(10)

[0069] From formula (10), we can see that the direction judgment of the monitor at 6 points is misjudged. Figure 4 It shows that the location is far from the disturbance source and there is a severe high-frequency energy fluctuation in the presence of distributed energy grid connection, which ultimately leads to misjudgment. Figure 2 The distribution network structure information and monitoring instrument location information shown in the figure are used to form a structural information matrix according to formula (4) as follows, and the particle swarm search motion space is restricted accordingly:

[0070]

[0071] According to step 5, perform particle swarm iterative search and evaluate the results during the search process using the evaluation function in step 4. The matrix particle with the minimum evaluation value is finally obtained as follows:

[0072] X i =[1-11111-1-111111] T (12)

[0073] It can be seen that it corresponds to the 8th line segment L8, and the positioning result of this method is accurate. Figure 5 The change process of the optimal evaluation function value during the iterative search process is shown.

[0074] The above-described embodiments are merely typical implementations of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A distributed power grid disturbance location method based on improved high-frequency energy method and matrix method, characterized in that: The following steps are involved: Step 1: perform wavelet packet discretization decomposition and reconstruction on the original disturbance signal data, and use the reconstruction information to obtain the high-frequency power and high-frequency energy in the full signal to obtain the reconstructed high-frequency signal; Step 2: Determine the time when the disturbance occurs based on the reconstructed high-frequency signal, and determine the relative direction between the disturbance source and the monitoring point by improving the new direction criterion of the high-frequency energy method, specifically: Observe the high-frequency energy situation within the time range of ∆t before and after the disturbance occurs, and use the sign of the mutation of the high-frequency energy value before and after the disturbance to determine the direction of the disturbance source: If the sign of the high-frequency energy mutation is positive at the moment of disturbance, it means that the disturbance source is located upstream of the monitoring point; If the sign of the high-frequency energy mutation is negative at the moment of disturbance, it means that the disturbance source is located downstream of the monitoring point; Step 3: Form the direction discrimination information matrix and the system structure information matrix, establish the particle swarm optimization model, and limit the particle swarm search motion space and particle motion span; Step 4: Construct an evaluation function that is tolerant to directional discrimination information and applicable to distributed energy grids. The smaller the value of the evaluation function, the better the corresponding result. Step 5: Use the particle swarm algorithm to iteratively search in space, determine the optimal particle position based on the evaluation function, and locate the disturbance source in combination with the system structure information matrix.

2. The distributed power grid disturbance location method based on the improved high-frequency energy method and matrix method according to claim 1 is characterized in that: In step 1, the high-frequency power and high-frequency energy in the full signal are obtained using the reconstructed information, specifically: The original disturbance voltage and current signals are decomposed into three layers of wavelet packets and the high-frequency components are reconstructed. The high-frequency power is obtained by calculating the voltage high-frequency component and the current high-frequency component: (1) Where, is the high frequency power, is the high-frequency component of voltage, is the high-frequency component of the current; Integrating the high-frequency power gives the high-frequency energy: (2) Where, It is high frequency energy.

3. The distributed power grid disturbance location method based on the improved high-frequency energy method and matrix method according to claim 1 is characterized in that: In step 3, the direction discrimination information matrix and the system structure information matrix are formed, and a particle swarm optimization model is established to limit the particle swarm search motion space and particle motion span, as follows: (3.1) Determine the direction discrimination information matrix within the system ,in is the total number of monitors in the system: (3) Where, The value depends on the The judgment result of the monitoring point is normal when the disturbance source is located upstream of the monitoring point. ,on the contrary ; (3.2) Form a structural information matrix based on the topological structure information of the distribution network and the location information of the monitoring instruments in the system ,in is the number of line segments in the system: (4) Where, Indicates the position relationship between the line segment and the monitor in the distribution network. The line section is located at Downstream direction of each monitor ,on the contrary ; (3.3) Establish a particle swarm optimization model, and the algorithm particles are taken as the direction discrimination information matrix Matrix particles with the same structure ,in Indicates the particle number: (5) (3.4) Restrict the movable positions of the matrix particles so that the position states that the matrix particles can reach in the motion space are all the direction discrimination matrices corresponding to when the disturbance source is located in a certain line segment; limit the movement span of the matrix particles to 1, that is, after each iteration, move to the direction discrimination matrix corresponding to the previous or next line segment to ensure global traversal.

4. The distributed power grid disturbance location method based on the improved high-frequency energy method and matrix method according to claim 1 is characterized in that: In step 4, an evaluation function with directional discrimination information fault tolerance and applicable to distributed energy grids is constructed, specifically: (4.1) Matrix particles and direction discrimination information matrix The greater the difference between them, the greater the evaluation value and the worse the evaluation result; (4.2) Taking the maximum value of all high-frequency energy mutations monitored in the system as the benchmark, the ratio of other monitoring values to the benchmark is used to represent the distance coefficient between the monitoring point and the disturbance source. The farther the distance, the smaller the coefficient, indicating that the credibility of the direction discrimination information of the monitoring point is lower. The calculation result is recorded as , when the matrix particles and the perturbation direction discriminant matrix Difference bit corresponding distance coefficient Compensate for the evaluation value when (4.3) Let the change amplitude be the ratio of the abnormal change amplitude of high-frequency energy to the peak value within a short period of time after the disturbance occurs. , when the matrix particles and the perturbation direction discriminant matrix Difference bit corresponding change ratio When it exceeds 50%, the evaluation value will be compensated; (4.4) When the matrix particles and the perturbation direction discriminant matrix When the line corresponding to the difference position is a grid-connected line containing distributed energy, the evaluation value is compensated; (4.5) When the matrix particles and the perturbation direction discriminant matrix When the line corresponding to the difference bit is a virtual monitor line, the evaluation value is compensated; To sum up, the evaluation function is constructed as follows: (6) Where, , , , , Is the compensation coefficient, the value range is 0~5, the specific value depends on the importance of the setting of this influencing factor; function when Non-zero and The value meets the judgment condition When the return value is -1, Non-zero and The value meets the judgment condition Returns a value of 1 when , otherwise it returns a value of 0; function when Non-zero and The value meets the judgment condition Returns a value of 1 when , otherwise it returns a value of 0; function when Non-zero and The return value is 1 when the corresponding line is connected to the grid with distributed energy, otherwise it returns 0; the function when Non-zero and If the corresponding line is a virtual monitor, the return value is 1; otherwise, the return value is 0.

5. The distributed power grid disturbance location method based on the improved high-frequency energy method and matrix method according to claim 1 is characterized in that: In step 5, the particle swarm algorithm is used to iteratively search in space, the optimal particle position is determined based on the evaluation function, and the disturbance source is located in combination with the system structure information matrix. The particle swarm iterative search rule is: (7) (8) Where, Representatives in the The velocity of the matrix particles at the iteration; and In the The position and velocity of the matrix particles at the iteration; represents the inertia factor, 、 represents the acceleration factor, 、 and Assign random numbers between 0 and 1 respectively; Representative The optimal value of the individual particle at the iteration; Representative The optimal value of the particle swarm at the iteration; is the threshold function; Indicates that the disturbance source is located at The direction discrimination matrix corresponding to the line segment is, Indicates that the disturbance source is located at The direction discrimination matrix corresponding to the line segment; The relationship between the threshold function and the particle speed is set as: (9) During the iteration process, the individual optimal value of the particle and the global optimal value are continuously updated. When the number of iterations reaches the set maximum value, the matrix particle corresponding to the global optimal value is output. The line segment corresponding to the matrix particle is the disturbance source positioning result.

Images