Distribution network differential protection method based on dynamic time bending distance and floating threshold

Abstract

The present invention discloses a distribution network differential protection method based on dynamic time warping distance and floating threshold, comprising the following steps: Step 1: current sampling devices at both ends of the protected line in the distribution line perform current signal sampling at a fixed frequency; Step 2: when the current fault detection device detects that there is a sudden change current fault in the distribution line, the current sampling devices at both ends of the protected line collect the current signals within A milliseconds before the fault moment and within B milliseconds after the fault to form current sampling sequences i1 and i2; Step 3: after the normalized current sequence #imgabs0# is inverted, the DTW distance and the differential protection floating threshold braking threshold value are calculated with the normalized current sequence #imgabs1#; Step 4: Calculate whether the DTW distance and the differential protection floating threshold braking threshold value meet the judgment criteria. The present invention utilizes the good tolerance of the dynamic time warping algorithm to the time domain, and adds a floating threshold to prevent the false operation of the protection, thereby improving the reliability of the protection.

Description

Technical Field

[0001] The invention relates to the technical field of distribution network protection, and in particular to a distribution network differential protection method based on dynamic time bending distance and floating threshold. Background Art

[0002] The distribution network is directly connected to the users, and is a key link in ensuring the quality of power supply, improving the efficiency of power system operation and the ability to accept renewable energy. Most traditional distribution networks are single-power radial structures, and the load type is usually traditional load, and the boundary between source-grid-load is very clear. With the development of distribution networks and changes in user needs, distributed power sources are gradually connected to distribution networks, and the boundary between source-grid-load becomes blurred. After the distributed power sources are connected, the fault characteristics of the distribution network are changed, causing the original protection devices to reduce sensitivity, refuse to operate or malfunction. Obviously, current protection and other protection methods based on single-ended electrical quantities are increasingly difficult to reflect complex distribution network faults.

[0003] Differential protection based on two-terminal electrical quantities is known as the most ideal protection method. It can be used as the main protection without dead zones and has become the first choice for protection in many occasions. Applying differential protection to power grids can quickly and accurately locate and remove faults, greatly reducing the impact of faults on power grids. However, differential protection has high requirements for synchronization, and many transmission channels cannot meet the requirements of differential protection. At present, optical fiber is the main transmission channel for differential protection and is widely used in differential protection of transmission lines. If optical fiber is used in distribution networks, it will bring problems such as high construction cost, difficulty in laying, and susceptibility to damage.

[0004] The arrival of 5G provides a transmission channel for distribution network differential protection with performance comparable to that of optical fiber. As a type of wireless communication, it has advantages that optical fiber does not have, such as convenient deployment, low cost, and easy upgrade and transformation. It is a transmission medium that is very suitable for distribution network differential protection. However, when 5G is used as a transmission medium for differential protection, there is a problem of delay jitter, which causes the transmitted current waveform to be distorted, which may affect the reliability of protection. Existing related research uses the dynamic time bending distance algorithm to overcome the delay jitter problem due to its good tolerance to the time axis, but the fixed threshold may cause the protection to refuse to operate or malfunction, and it is difficult to adapt to different distribution network structures and line parameters. Summary of the invention

[0005] The purpose of the present invention is to provide a distribution network differential protection method based on dynamic time bending distance and floating threshold. The present invention utilizes the good tolerance of the dynamic time warping algorithm to the time domain, overcomes the problem of delay jitter in the 5G transmission process, and adds a floating threshold to prevent false operation of the protection, thereby improving the reliability of the protection.

[0006] To achieve this purpose, the present invention is designed to provide a distribution network differential protection method based on dynamic time bending distance and floating threshold, which includes the following steps:

[0007] Step 1: The current sampling devices at both ends of the protected line in the distribution line sample current signals at a fixed frequency, and the sudden change current fault detection devices at both ends of the protected line respectively perform real-time sudden change current detection on the current signals collected by the corresponding current sampling devices;

[0008] Step 2: When the current fault detection device detects that there is a sudden current fault in the distribution line, the current sampling device at one end of the protected line collects the current signals within A milliseconds before the fault and within B milliseconds after the fault to form a current sampling sequence i 1 At this time, the current sampling device at the other end of the protected line collects the current signals within A milliseconds before the fault and within B milliseconds after the fault to form a current sampling sequence i 2 ;

[0009] Step 3: The differential protection device at one end of the protected line responds to the current sequence i 1 Perform normalization to obtain the normalized current sequence The differential protection device at the other end of the protected line responds to the current sequence i 2 Perform normalization to obtain the normalized current sequence The differential protection device at the other end of the protected line converts the normalized current sequence The normalized current sequence is sent to the differential protection device at one end of the protected line by wireless communication. The inverted and normalized current sequence DTW distance D(k) and differential protection floating threshold braking threshold D res (k) calculation;

[0010] Step 4: Calculate the DTW distance D(k) and the differential protection floating threshold braking threshold D res (k) Whether it satisfies the criterion D(k)>K×D res (k)+D0, where K is the braking coefficient of the protection criterion, and D0 is the delay jitter threshold value used to balance the lack of communication synchronization. When the criterion is met, the differential protection device performs protection action, and the fault condition is a fault on the protected line.

[0011] Beneficial effects of the present invention:

[0012] 1. The present invention applies the dynamic time warping distance algorithm to the distribution network differential protection, and uses the algorithm's good tolerance to the time domain to solve the problem of slight asynchrony and delay jitter in the 5G environment. In addition, the method has good recognition ability for various short-circuit types, and can determine the short-circuit type according to the number of phases exceeding the floating threshold.

[0013] 2. The present invention introduces a floating threshold. When a fault occurs within the zone, the DTW distance is large and the floating threshold is small, and the protection can be reliably operated. When a fault occurs outside the zone or it is normal, the DTW distance is small and the floating threshold is large, which has a good braking effect and the protection will not be malfunctioned.

[0014] 3. The present invention can be used in the differential protection of distribution networks in a 5G environment, which saves the cost of laying optical fibers and does not require the addition of external timing devices, greatly saving the cost of differential protection and having certain engineering promotion value. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of the process of the present invention;

[0016] Figure 2 This is a comparison chart of the original sampling point and the sampling point after 5G transmission;

[0017] Figure 3 The optimal path of DTW distance is illustrated by taking two sequences: sequence p =

[12479] and sequence q =

[35186] as examples;

[0018] Figure 4a The DTW distance and floating threshold rule when the fault occurs within the area;

[0019] Figure 4b The DTW distance and floating threshold rule for out-of-area faults;

[0020] Figure 5a To verify the model of the present invention, one: single power supply radial structure;

[0021] Figure 5b To verify the model 2 of the present invention: closed-loop operation structure;

[0022] Figure 6a The DTW change of phase A when a two-phase short circuit occurs in the single-power radial distribution network;

[0023] Figure 6b The DTW change of phase A when a two-phase short circuit occurs in the closed-loop distribution network;

[0024] Figure 7 The DTW change of phase A when a two-phase short circuit occurs outside the area in the closed-loop distribution network;

[0025] in, Figure 6a , Figure 6b , Figure 7 They are composed of three waveforms, from top to bottom: original current on both sides, normalized current on both sides, DTW distance and floating threshold. Figure 5a In the figure, l1~l10 are ten lines. A fault occurring at f1 in section MN is called an intra-section fault, and a fault occurring at f2 on the right side of N is an extra-section fault. DG1 and DG2 are two distributed power sources. DETAILED DESCRIPTION

[0026] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0027] like Figures 1 to 7 The distribution network differential protection method based on dynamic time bending distance and floating threshold shown in the figure comprises the following steps:

[0028] Step 1: The current sampling devices at both ends of the protected line in the distribution line sample the current signal at a fixed frequency of f = 1000 Hz. Theoretically, the sampling points after transmission are as follows: Figure 2 As shown, the sudden current fault detection devices at both ends of the protected line respectively perform real-time sudden current detection on the current signals collected by the corresponding current sampling devices;

[0029] Step 2: When the current fault detection device detects that there is a sudden current fault in the distribution line, the current sampling device at one end of the protected line collects the current signals within 5 milliseconds before the fault and within 15 milliseconds after the fault to form a current sampling sequence i 1 ,i 1 ={i 11 ,i 12 ,…i 1k ,…i 1n}, at this time, the current sampling device at the other end of the protected line collects the current signals within A milliseconds before the fault and within B milliseconds after the fault to form a current sampling sequence i 2 ,i 2 ={i 21 ,i 22 ,…i 2k ,…i 2n}, where i 1n Indicates the nth current sampling value collected at one end of the protection line, i 2n Indicates the nth current sampling value collected at the other end of the protection line;

[0030] Step 3: The differential protection device at one end of the protected line responds to the current sequence i 1 Perform normalization to obtain the normalized current sequence The differential protection device at the other end of the protected line responds to the current sequence i 2Perform normalization to obtain the normalized current sequence The differential protection device at the other end of the protected line converts the normalized current sequence The normalized current sequence is sent to the differential protection device at one end of the protected line through 5G wireless communication. The inverted and normalized current sequence DTW distance D(k) and differential protection floating threshold braking threshold D res (k) calculation;

[0031] Step 4: Calculate the DTW distance D(k) and the differential protection floating threshold braking threshold D res (k) Whether it satisfies the criterion D(k)>K×D res (k)+D0, where K is the braking coefficient of the protection criterion, which is generally 0.5-1, and D0 is the delay jitter threshold value used to balance the synchronization of communication. When the criterion is established, the differential protection device performs protection action. The fault condition is a fault on the protected line. In this embodiment, the value of K is 0.8, and D 0 Set to 5.

[0032] When the judgment criterion is not met, the differential protection device does not operate. At this time, the non-protected line is at fault or operating normally.

[0033] In step 1 of the above technical solution, when the sudden current fault detection device performs real-time sudden current detection on the collected current signal and detects that N consecutive sudden current values ​​are greater than a preset value, it is determined to be a sudden current fault.

[0034] In step 3 of the above technical solution, the current sequence i 1 The specific method for normalization is: 1 * =(i 1k -i 1min ) / (i 1max -i 1min ), where i 1k To protect the kth current sampling value at one end of the line, i 1max To protect the maximum value of the current sampling value at one end of the line, i 1min To protect the minimum value of the current sampling value at one end of the line, after normalization, i 1 * ∈[0,1].

[0035] In step 3, the current sequence i 2 The specific method for normalization is: 2 * =(i 2k-i 2min ) / (i 2max -i 2min ), where i 2k To protect the kth current sampling value at the other end of the line, i 2max To protect the maximum value of the current sampling value at the other end of the line, i 2min To protect the minimum value of the current sampling value at the other end of the line, after normalization, i 2 * ∈[0,1].

[0036] In step 3 of the above technical solution, the normalized current sequence The inverted and normalized current sequence DTW distance The differential protection floating threshold braking threshold value Figure 4a The changing rules of DTW distance and floating threshold when there is a fault in the zone are shown. It can be seen that the DTW distance is larger (left figure) and the floating threshold is smaller (right figure), and the protection can be operated reliably; Figure 4b The variation pattern of DTW distance and floating threshold in the case of out-of-zone faults is shown. It can be seen that the DTW distance is almost zero (left figure) and the floating threshold is very large (right figure), which has a braking effect.

[0037] The specific method of DTW distance calculation is:

[0038] First, the normalized current series is calculated. The inverted and normalized current sequence The Euclidean distance is then used to obtain the DTW distance matrix, and the sum of the smallest DTW distance paths in the DTW distance matrix is ​​obtained, which is the normalized current sequence. The inverted and normalized current sequence DTW distance.

[0039] DTW distance calculation method

[0040] First, calculate the Euclidean distance between the two sequences:

[0041]

[0042] Then calculate the DTW distance matrix D as follows

[0043]

[0044] According to the above method, an optimal matching path will be obtained, and the final DTW distance will be

[0045]

[0046] in, Represents the normalized current series There are m current data in total. Represents the normalized current series There are n current data in total. for and The DTW distance D(m,n) is the value obtained by adding the values ​​on the closest path selected by the D matrix from the above paths. res (k), we only need to replace Change to

[0047] In the above technical solution, starting from the upper left corner of the DTW distance matrix, the path includes downward, rightward and lower-right directions. The path with the shortest distance is selected from the above paths, and the values ​​on the path are added to obtain the DTW distance.

[0048] In step 1 of the above technical solution, the calculation formula of the sudden change current is ΔI k =|i k -i k-N |, where i k is the instantaneous sampling value of the current, i k-N It is the instantaneous sampling value of the current one sampling cycle ago.

[0049] In the above technical solution, if the sudden change of the current exceeds the set value I for 5 consecutive times, 0 Then, there is a sudden current fault in the distribution line.

[0050] The following is an explanation of the above steps combined with specific simulation situations:

[0051] FIG5 is a simulation model of the present invention. There are two simulation models, namely a single power supply radial structure and a closed-loop operation structure, to study the fault condition of the line between the M side and the N side.

[0052] Example 1: Use Figure 5a The simulation model has a single power supply radial structure and the simulation time is 0.8s. At 0.1s, an internal fault occurs at f1, and at 0.5s, an external fault occurs at f2. The current sampling values ​​of the M side and the N side within 0.8s are monitored and recorded as i 1 with i 2 When the sudden current detects a fault, the currents on both sides are normalized and recorded as and Thus, the DTW example and the main components of the floating threshold are calculated. When the criterion D(k)>0.8·D is met resThe protection will be activated when (k)+5, and will not be activated if it is not satisfied. Figure 6a This is the change law of the fault phase A when a fault occurs in the AB two-phase short-circuit zone of a single power supply system. The simulation results show that the DTW distance exceeds the floating threshold at 0.1154s, proving that a fault has occurred in phase A.

[0053] Example 2: Use Figure 5b The closed-loop operation structure in the simulation model has a simulation duration of 0.8s. At 0.1s, an internal fault occurs, i.e., a fault occurs at f1, and at 0.5s, an external fault occurs, i.e., a fault occurs at f2. The current sampling values ​​of the M side and the N side within 0.8s are monitored and recorded as i 1 with i 2 When the sudden current detects a fault, the currents on both sides are normalized and recorded as and Thus, the DTW example and the main components of the floating threshold are calculated. When the criterion D(k)>0.8·D is met res The protection will be activated when (k)+5, and will not be activated if it is not satisfied. Figure 6b This is the change law of the fault phase A when a fault occurs in the short-circuit zone of the two-phase AB in the closed-loop operation system. The simulation results show that the DTW distance exceeds the floating threshold at 0.1033s, proving that a fault has occurred in phase A. Figure 7 This is the change law of the fault phase A when the closed-loop operation system AB two-phase short circuit fault occurs outside the zone. The simulation results show that when the fault occurs outside the zone, the DTW distance is always lower than the floating threshold and the protection will not malfunction.

[0054] The above two embodiments can verify that the distribution network differential protection based on dynamic time bending distance and floating threshold can accurately distinguish between faults inside and outside the area, and can achieve reliable operation inside the area and no false operation outside the area, so as to achieve the purpose of safe and stable operation of the distribution network system.

[0055] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

Claims

1. A distribution network differential protection method based on dynamic time bending distance and floating threshold, characterized in that: It includes the following steps: Step 1: The current sampling devices at both ends of the protected line in the distribution line sample current signals at a fixed frequency, and the sudden change current fault detection devices at both ends of the protected line respectively perform real-time sudden change current detection on the current signals collected by the corresponding current sampling devices; Step 2: When the current fault detection device detects that there is a sudden current fault in the distribution line, the current sampling device at one end of the protected line collects the current signals within A milliseconds before the fault and within B milliseconds after the fault to form a current sampling sequence i1. At this time, the current sampling device at the other end of the protected line collects the current signals within A milliseconds before the fault and within B milliseconds after the fault to form a current sampling sequence i2. Step 3: The differential protection device at one end of the protected line normalizes the current sequence i1 to obtain the normalized current sequence The differential protection device at the other end of the protected line normalizes the current sequence i2 to obtain the normalized current sequence The differential protection device at the other end of the protected line converts the normalized current sequence The normalized current sequence is sent to the differential protection device at one end of the protected line by wireless communication. The inverted and normalized current sequence DTW distance D(k) and differential protection floating threshold braking threshold D res (k) calculation; Step 4: Calculate the DTW distance D(k) and the differential protection floating threshold braking threshold D res (k) Whether it satisfies the criterion D(k)>K×D res (k)+D0, where K is the braking coefficient of the protection criterion, and D0 is the delay jitter threshold value used to balance the synchronization caused by communication failure. When the criterion is met, the differential protection device performs protection action, and the fault condition is a fault on the protected line; In step 3, the normalized current sequence The inverted and normalized current sequence DTW distance The differential protection floating threshold braking threshold value The specific method of DTW distance calculation is: First, the normalized current series is calculated. The inverted and normalized current sequence The Euclidean distance is then used to obtain the DTW distance matrix, and the sum of the smallest DTW distance paths in the DTW distance matrix is ​​obtained, which is the normalized current sequence. The inverted and normalized current sequence DTW distance; The DTW distance calculation method is: First, calculate the Euclidean distance between the two sequences: Then calculate the DTW distance matrix D as follows According to the above method, an optimal matching path will be obtained, and the final DTW distance will be in, Represents the normalized current series There are m current data in total. Represents the normalized current series There are n current data in total. for and DTW distance, D(m,n) is the value of the D matrix selecting the path with the shortest Euclidean distance from the above paths and adding the values ​​on the shortest path. To calculate D res (k), we only need to replace Change to 2. The distribution network differential protection method based on dynamic time bending distance and floating threshold according to claim 1 is characterized in that: In step 1, when the sudden current fault detection device performs real-time sudden current detection on the collected current signal and detects that N consecutive sudden current values ​​are greater than a preset value, it is determined to be a sudden current fault.

3. The distribution network differential protection method based on dynamic time bending distance and floating threshold according to claim 1 is characterized in that: In step 3, the specific method for normalizing the current sequence i1 is: i1 * =(i 1k -i 1min ) / (i 1max -i 1min ), where i 1k To protect the kth current sampling value at one end of the line, i 1max To protect the maximum value of the current sampling value at one end of the line, i 1min To protect the minimum value of the current sampling value at one end of the line, after normalization, i1 * ∈[0,1].

4. The distribution network differential protection method based on dynamic time bending distance and floating threshold according to claim 1 is characterized in that: In step 3, the specific method for normalizing the current sequence i2 is: i2 * =(i 2k -i 2min ) / (i 2max -i 2min ), where i 2k To protect the kth current sampling value at the other end of the line, i 2max To protect the maximum value of the current sampling value at the other end of the line, i 2min In order to protect the minimum value of the current sampling value at the other end of the line, after normalization, i2 * ∈[0,1].

5. The distribution network differential protection method based on dynamic time bending distance and floating threshold according to claim 1 is characterized in that: Starting from the upper left corner of the DTW distance matrix, the paths include downward, right, and lower-right directions. From the above paths, the path with the shortest distance is selected, and the values ​​on the path are added to obtain the DTW distance.

6. The distribution network differential protection method based on dynamic time bending distance and floating threshold according to claim 1 is characterized in that: In step 1, the calculation formula of the sudden change current is ΔI k =|i k -i k-N |, where i k is the instantaneous sampling value of the current, i k-N It is the instantaneous sampling value of the current one sampling cycle ago.

7. The distribution network differential protection method based on dynamic time bending distance and floating threshold according to claim 1 is characterized in that: If the sudden change in current exceeds the set value I0 for five consecutive times, there is a sudden change current fault in the distribution line.

Images