Voltage sag stabilizing system and method

Abstract

The invention discloses a voltage sag voltage stabilization method. The method comprises the following steps: S1, obtaining operation data of each electrical device; s2, inputting the operation data into a recognition model, and obtaining the operation state of each electrical device; and S3, when the operation state is a voltage abnormal state, cutting off a current power supply of the electrical equipment, connecting a bypass power supply, and supplying power to the electrical equipment through the bypass power supply. Detection is carried out through the method, multiple voltage detection devices do not need to be purchased, only a detector of the electrical device needs to be used, operation data are obtained through the Internet of Things, the detection cost is greatly reduced, and meanwhile the detection speed is increased. Generally, the influence of the instantaneous voltage change on the operation data of the electrical equipment can be almost ignored. Compared with the prior art that voltage detection is carried out through voltage detection equipment, the accuracy is higher, and the risk of misjudgment is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of circuit intelligent control, and in particular to a voltage sag stabilization system and method. Background Art

[0002] As the scale of distribution network gradually expands and the number of cable lines increases, it will be affected by various faults in actual operation. If the voltage is insufficient, it will affect the stability of the power system and the safe operation of electrical equipment.

[0003] In the prior art, a voltage detection device is usually installed on each electrical device to determine whether the voltage requirement is met by detecting whether the voltage on the electrical device reaches the rated voltage. When the voltage is insufficient, the power supply needs to be stabilized to ensure that the electrical device has sufficient voltage to work.

[0004] However, the existing voltage detection method requires the purchase of additional detection equipment, which requires a lot of cost resources. In addition, the detection will be affected by instantaneous voltage changes, which may lead to the risk of misjudgment. Summary of the invention

[0005] In view of the above-mentioned defects, the object of the present invention is to provide a voltage sag stabilization system and method to improve the accuracy of low voltage detection and quickly achieve voltage stabilization of the equipment.

[0006] To achieve this purpose, the present invention adopts the following technical solution: a voltage sag stabilization method, comprising the following steps:

[0007] Step S1: Obtaining the operating data of each electrical device;

[0008] Step S2: inputting the operation data into the recognition model to obtain the operation status of each electrical device;

[0009] Step S3: When the operating state is an abnormal voltage state, the current power supply of the electrical equipment is cut off, the bypass power supply is connected, and the electrical equipment is powered by the bypass power supply.

[0010] Preferably, before executing step S2, the following steps need to be executed:

[0011] Step A: Build a virtual twin model;

[0012] Step B: Initialize the X value to 1;

[0013] Step C: Obtain the X value, and use the X value as the startup quantity of the electrical equipment in the virtual twin model;

[0014] Step D: installing the electrical equipment into the virtual twin model in sequence, setting the operating voltage to the standard voltage, and obtaining the first operating state vector of the electrical equipment;

[0015] Step E: the value of X is increased by 1, and it is determined whether the value of X is greater than the value of N, where the value of N is the total number of electrical devices in a certain circuit. If not, steps C to E are executed again. If greater, the first operating state vector set F1 = {f1, f2, ..., fN} of the electrical devices is output;

[0016] Step F: Calculate the similarity between two vectors in the first state set, and construct a connection matrix based on the similarity;

[0017] Step G: Calculate the attention coefficient between each adjacent first running state vector through the learning function and the connection matrix;

[0018] The adjustment coefficient of the weight matrix in the recognition model is obtained through the attention coefficient.

[0019] Preferably, the recognition model in step S2 performs the following steps:

[0020] Step S21: marking the number of the acquired operating states with abnormal data as a first number, and marking the operating states with abnormal data as a second vector;

[0021] Step S22: based on the first quantity, obtaining a corresponding first running state vector from the first running state vector collection;

[0022] Step S23: Mapping the plurality of second vectors and the first operating state vector to the same space, and then using the judgment matrix to measure whether an abnormal voltage state occurs.

[0023] Preferably, the formula for measuring whether the voltage abnormality state occurs through the judgment matrix is ​​as follows:

[0024] P=max(W d *ω i-1 [U1][U2],W d [U1][U2],W d *ω i+1 [U1][U2]);

[0025] When the P value is greater than the probability threshold, it is judged that the voltage is abnormal. d is the weight matrix, ω i-1 is the adjustment coefficient calculated by the first running state vector and the previous adjacent first running state vector, ω i,i+1 is the adjustment coefficient calculated by the first running state vector and the next adjacent first running state vector, U1 is the mapping of the first running state vector in space, and U2 is the mapping of the second vector in space.

[0026] Specifically, step G is as follows:

[0027] The formula for obtaining the attention coefficient is as follows:

[0028]

[0029] Where att is a function composed of LeakyReLU and Linear, W is the first weight coefficient,

[0030] h i =W h f i ,h j =W h f j , W h is the second weight coefficient, where N i is the lowest i elements in the adjacency matrix;

[0031] The formula for obtaining the adjustment coefficient is as follows:

[0032]

[0033] sim(i,j) is f i With f j The similarity of

[0034] Preferably, the method further comprises the following steps:

[0035] The voltage data after the bypass power supply supplies power to the electrical equipment must meet the following conditions: the output three-phase voltage cannot be less than 100% of the rated voltage, and the impact on the electrical equipment end or the power supply end must not be greater than 105% of the rated voltage.

[0036] A voltage sag stabilization system, using the voltage sag stabilization method, comprises an operation data acquisition module, an analysis module and a voltage stabilization module;

[0037] The operation data acquisition module is used to acquire the operation data of each electrical device;

[0038] The analysis module is used to input the operation data into the recognition model to obtain the operation status of each electrical device;

[0039] The voltage stabilizing module is used to cut off the current power supply of the electrical equipment when the operating state is an abnormal voltage state, connect the bypass power supply, and supply power to the electrical equipment through the bypass power supply.

[0040] Preferably, it further comprises an adjustment coefficient acquisition module, wherein the adjustment coefficient acquisition module comprises a construction submodule, an initialization submodule, a setting submodule, a virtual vector acquisition submodule, a loop module, a relationship acquisition submodule and a coefficient acquisition submodule;

[0041] The construction submodule is used to build a virtual twin model;

[0042] The initialization submodule is used to initialize the X value to 1;

[0043] The setting submodule is used to obtain the X value, which is used as the startup quantity of the electrical equipment in the virtual twin model;

[0044] The virtual vector acquisition submodule is used to sequentially install the electrical equipment into the virtual twin model, set the operating voltage to the standard voltage, and obtain the first operating state vector of the electrical equipment;

[0045] The loop module is used to increase the value of X by 1 and determine whether the value of X is greater than the value of N, where the value of N is the total number of electrical devices in a certain circuit. If not, the setting submodule and the virtual vector acquisition submodule are repeatedly called. If greater, the first operation state vector set of the electrical device is output;

[0046] The relationship acquisition submodule is used to calculate the similarity between two nodes in the first state set, and construct a connection matrix based on the similarity;

[0047] The coefficient acquisition submodule is used to calculate the attention coefficient between each adjacent vector through the learning function and the connection matrix;

[0048] The adjustment coefficient of the weight matrix in the recognition model is obtained through the attention coefficient.

[0049] Preferably, the analysis module includes a marking submodule, a determination submodule and a result judgment submodule;

[0050] The marking submodule is used to mark the number of the acquired data abnormal operation status as a first number, and mark the data abnormal operation status as a second vector;

[0051] The determining submodule is used to obtain a corresponding first operating state vector in the first operating state vector set based on the first quantity;

[0052] The result judgment submodule is used to map the plurality of second vectors and the first operation state vector to the same space, and then measure whether an abnormal voltage state occurs through the judgment matrix.

[0053] One of the above technical solutions has the following advantages or beneficial effects: by using this method for detection, there is no need to purchase multiple voltage detection devices, only the detectors provided by the electrical equipment need to be used, and the operation data can be obtained through the Internet of Things, which greatly reduces the cost of detection and also improves the speed of detection. Usually, the operation data of electrical equipment is almost negligibly affected by the change of instantaneous voltage. Compared with the existing voltage detection by voltage detection equipment, the accuracy of voltage detection will be higher, reducing the risk of misjudgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 is a flow chart of an embodiment of the method of the present invention.

[0055] Figure 2 It is a schematic diagram of the structure of an embodiment of the system of the present invention. DETAILED DESCRIPTION

[0056] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0057] In the description of the embodiments of the present invention, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0058] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, the meaning of "plurality" is two or more. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0059] like Figures 1-2 As shown, a voltage sag stabilization method comprises the following steps:

[0060] Step S1: Obtaining the operating data of each electrical device;

[0061] Step S2: inputting the operation data into the recognition model to obtain the operation status of each electrical device;

[0062] Step S3: When the operating state is an abnormal voltage state, the current power supply of the electrical equipment is cut off, the bypass power supply is connected, and the electrical equipment is powered by the bypass power supply.

[0063] Since the cost of directly using voltage detection equipment to detect the circuit is high, and it is used to be affected by the change of instantaneous voltage, there is a risk of misjudgment. For this reason, in the present invention, the voltage state is judged by obtaining the operating data of each electrical device, because in the case of insufficient voltage, the operating data of each electrical device will be different from the operating data in normal state, and the difference can be used to determine whether the operating state is in a low voltage operating state. By using this method for detection, there is no need to purchase multiple voltage detection devices, only the detector provided by the electrical equipment needs to be used, and the operating data is obtained through the Internet of Things, which greatly reduces the cost of detection and also improves the speed of detection. Usually, the operating data of electrical equipment is almost negligibly affected by the change of instantaneous voltage. Compared with the existing voltage detection by voltage detection equipment, the accuracy will be higher, reducing the risk of misjudgment. When the operating state is an abnormal voltage state, the current power supply of the electrical equipment will be cut off, and the power supply of the electrical equipment will be switched to the bypass power supply, wherein the bypass power supply is an independent power supply, which is not connected to the power grid, so it can continuously provide stable electricity to the electrical equipment in a short time, so that the voltage stabilization effect can be achieved in a short time, thereby ensuring the normal operation of the electrical equipment.

[0064] Preferably, before executing step S2, the following steps need to be executed:

[0065] Step A: Build a virtual twin model;

[0066] Step B: Initialize the X value to 1;

[0067] Step C: Obtain the X value, and use the X value as the startup quantity of the electrical equipment in the virtual twin model;

[0068] Step D: installing the electrical equipment into the virtual twin model in sequence, setting the operating voltage to the standard voltage, and obtaining the first operating state vector of the electrical equipment;

[0069] Step E: the value of X is increased by 1, and it is determined whether the value of X is greater than the value of N, where the value of N is the total number of electrical devices in a certain circuit. If not, steps C to E are executed again. If greater, the first operating state vector set F1 = {f1, f2, ..., fN} of the electrical devices is output;

[0070] Step F: Calculate the similarity between two vectors in the first state set, and construct a connection matrix based on the similarity;

[0071] Step G: Calculate the attention coefficient between each adjacent first running state vector through the learning function and the connection matrix;

[0072] The adjustment coefficient of the weight matrix in the recognition model is obtained through the attention coefficient.

[0073] It is very difficult to obtain the characteristics of low voltage under different electrical quantities one by one. Therefore, in the present invention, a virtual twin model is constructed according to the circuit to be detected, and the characteristics at low voltage are obtained in the virtual twin model. However, in actual production, each electrical device in the circuit is not necessarily in working condition. Therefore, the first operating state vector of low voltage under different startup numbers is gradually obtained in the virtual twin model to meet the actual detection needs. For example, in the first operating state vector collection, f1 represents the first operating state vector when only one electrical device is started, f2 represents the first operating state vector when two electrical devices are started, and so on.

[0074] In the experiment, the first running state vector is obtained, and the features of some adjacent first running state vectors are similar (for example, the value of f1 may be 0.1x, and the value of f2 may be 0.102x). In actual operation, since there may be a voltage stabilizing device in some electrical equipment, it is difficult to accurately select the corresponding first running state vector for judgment in the subsequent judgment. For this reason, in the present invention, based on the similarity between the two vectors, an adjacency matrix is ​​constructed based on the similarity, wherein the method of constructing the adjacency matrix is ​​the same as the existing construction method, and a parameter is defined as the threshold value of whether the nodes between the adjacency matrix are connected. When the similarity of the two first running state vectors is greater than the parameter, it means that the two first running state vectors are adjacent in the adjacency matrix. On the contrary, when the similarity of the two first running state vectors is less than the parameter, it means that the two first running state vectors are not adjacent in the adjacency matrix. After obtaining the adjacency matrix, the attention coefficient between each adjacent first running state vector can be calculated by the learning function, and the weight in the subsequent recognition model can be adjusted by the first running state vector, thereby improving the accuracy of detection.

[0075] Preferably, the recognition model in step S2 performs the following steps:

[0076] Step S21: marking the number of the acquired operating states with abnormal data as a first number, and marking the operating states with abnormal data as a second vector;

[0077] Step S22: based on the first quantity, obtaining a corresponding first running state vector from the first running state vector collection;

[0078] Step S23: Mapping the plurality of second vectors and the first operating state vector to the same space, and then using the judgment matrix to measure whether an abnormal voltage state occurs.

[0079] In the recognition model, the number of operating states with data anomalies will be obtained, where the data anomaly is not necessarily caused by voltage, but may also be caused by internal components of the electrical equipment. Therefore, it is not possible to determine whether it is a low voltage state at this time. After obtaining the number of operating states with data anomalies, the corresponding first operating state vector is determined according to the first number. Then, the probability of whether it is a voltage anomaly is calculated in the judgment matrix through each second vector and the first operating state vector. When the probability is greater than the probability threshold, it can be explained that there is a low voltage situation at present, and voltage stabilization processing is required.

[0080] Preferably, the formula for measuring whether the voltage abnormality state occurs through the judgment matrix is ​​as follows:

[0081] P=max(W d *ω i-1 [U1][U2],W d [U1][U2],W d *ω i+1 [U1][U2]);

[0082] When the P value is greater than the probability threshold, it is judged that the voltage is abnormal. d is the weight matrix, ω i-1 is the adjustment coefficient calculated by the first running state vector and the previous adjacent first running state vector, ω i,i+1 is the adjustment coefficient calculated by the first running state vector and the next adjacent first running state vector, U1 is the mapping of the first running state vector in space, and U2 is the mapping of the second vector in space.

[0083] Specifically, step G is as follows:

[0084] The formula for obtaining the attention coefficient is as follows:

[0085]

[0086] Where att is a function composed of LeakyReLU and Linear, W is the first weight coefficient,

[0087] h i =W h f i ,h j =W h f j , W h is the second weight coefficient, where N iis the lowest i elements in the adjacency matrix;

[0088] The formula for obtaining the adjustment coefficient is as follows:

[0089]

[0090] sim(i,j) is f i With f j The similarity of

[0091] Preferably, the method further comprises the following steps:

[0092] The voltage data after the bypass power supply supplies power to the electrical equipment must meet the following conditions: the output three-phase voltage cannot be less than 100% of the rated voltage, and the impact on the electrical equipment end or the power supply end must not be greater than 105% of the rated voltage.

[0093] By limiting voltage surges to within 105% of the rated voltage, bypass supplies help protect electrical equipment from damage caused by excessive voltage. This helps extend the life of equipment and reduces equipment failures and repair costs caused by voltage problems.

[0094] A voltage sag stabilization system, using the voltage sag stabilization method, comprises an operation data acquisition module, an analysis module and a voltage stabilization module;

[0095] The operation data acquisition module is used to acquire the operation data of each electrical device;

[0096] The analysis module is used to input the operation data into the recognition model to obtain the operation status of each electrical device;

[0097] The voltage stabilizing module is used to cut off the current power supply of the electrical equipment when the operating state is an abnormal voltage state, connect the bypass power supply, and supply power to the electrical equipment through the bypass power supply.

[0098] Preferably, it further comprises an adjustment coefficient acquisition module, wherein the adjustment coefficient acquisition module comprises a construction submodule, an initialization submodule, a setting submodule, a virtual vector acquisition submodule, a loop module, a relationship acquisition submodule and a coefficient acquisition submodule;

[0099] The construction submodule is used to build a virtual twin model;

[0100] The initialization submodule is used to initialize the X value to 1;

[0101] The setting submodule is used to obtain the X value, which is used as the startup quantity of the electrical equipment in the virtual twin model;

[0102] The virtual vector acquisition submodule is used to sequentially install the electrical equipment into the virtual twin model, set the operating voltage to the standard voltage, and obtain the first operating state vector of the electrical equipment;

[0103] The loop module is used to increase the value of X by 1 and determine whether the value of X is greater than the value of N, where the value of N is the total number of electrical devices in a certain circuit. If not, the setting submodule and the virtual vector acquisition submodule are repeatedly called. If greater, the first operation state vector set of the electrical device is output;

[0104] The relationship acquisition submodule is used to calculate the similarity between two nodes in the first state set, and construct a connection matrix based on the similarity;

[0105] The coefficient acquisition submodule is used to calculate the attention coefficient between each adjacent vector through the learning function and the connection matrix;

[0106] The adjustment coefficient of the weight matrix in the recognition model is obtained through the attention coefficient.

[0107] Preferably, the analysis module includes a marking submodule, a determination submodule and a result judgment submodule;

[0108] The marking submodule is used to mark the number of the acquired data abnormal operation status as a first number, and mark the data abnormal operation status as a second vector;

[0109] The determining submodule is used to obtain a corresponding first operating state vector in the first operating state vector set based on the first quantity;

[0110] The result judgment submodule is used to map the plurality of second vectors and the first operation state vector to the same space, and then measure whether an abnormal voltage state occurs through the judgment matrix.

[0111] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0112] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A voltage sag stabilization method, characterized in that: The steps include: Step S1: Obtaining the operating data of each electrical device; Step S2: inputting the operation data into the recognition model to obtain the operation status of each electrical device; Step S3: When the operating state is an abnormal voltage state, the current power supply of the electrical equipment is cut off, the bypass power supply is connected, and the electrical equipment is powered by the bypass power supply.

2. A voltage sag stabilization method according to claim 1, characterized in that: Before executing step S2, the following steps need to be performed: Step A: Build a virtual twin model; Step B: Initialize the value of X to 1; Step C: Obtain the X value, and use the X value as the startup quantity of the electrical equipment in the virtual twin model; Step D: installing the electrical equipment into the virtual twin model in sequence, setting the operating voltage to the standard voltage, and obtaining the first operating state vector of the electrical equipment; Step E: the value of X is increased by 1, and it is determined whether the value of X is greater than the value of N, where the value of N is the total number of electrical devices in a certain circuit. If not, steps C to E are executed again. If greater, the first operating state vector set F1 = {f1, f2, ..., fN} of the electrical devices is output; Step F: Calculate the similarity between two vectors in the first state set, and construct a connection matrix based on the similarity; Step G: Calculate the attention coefficient between each adjacent first running state vector through the learning function and the connection matrix; The adjustment coefficient of the weight matrix in the recognition model is obtained through the attention coefficient.

3. A voltage sag stabilization method according to claim 2, characterized in that: In step S2, the recognition model performs the following steps: Step S21: marking the number of the acquired operating states with abnormal data as a first number, and marking the operating states with abnormal data as a second vector; Step S22: based on the first quantity, obtaining a corresponding first running state vector from the first running state vector collection; Step S23: Mapping the plurality of second vectors and the first operating state vector to the same space, and then using the judgment matrix to measure whether an abnormal voltage state occurs.

4. A voltage sag stabilization method according to claim 3, characterized in that: The formula for measuring whether the voltage abnormality occurs through the judgment matrix is ​​as follows: P=max(W d *oh i-1 [U1][U2],W d [U1][U2],W d *oh i+1 [U1][U2]); When the P value is greater than the probability threshold, it is judged that the voltage is abnormal. d is the weight matrix, ω i-1 is the adjustment coefficient calculated by the first running state vector and the previous adjacent first running state vector, ω i,i+1 is the adjustment coefficient calculated by the first running state vector and the next adjacent first running state vector, U1 is the mapping of the first running state vector in space, and U2 is the mapping of the second vector in space.

5. A voltage sag stabilization method according to claim 2, characterized in that: The details of step G are as follows: The formula for obtaining the attention coefficient is as follows: Where att is a function composed of LeakyReLU and Linear, W is the first weight coefficient, h i =W h f i ,h j =W h f j , W h is the second weight coefficient, where N i is the lowest i elements in the adjacency matrix; The formula for obtaining the adjustment coefficient is as follows: sim(i,j) is f i With f j The similarity of 6. A voltage sag stabilization method according to claim 1, characterized in that: The following steps are also included: The voltage data after the bypass power supply supplies power to the electrical equipment must meet the following conditions: the output three-phase voltage cannot be less than 100% of the rated voltage, and the impact on the electrical equipment end or the power supply end must not be greater than 105% of the rated voltage.

7. A voltage sag stabilization system, characterized in that: A voltage sag stabilization method according to any one of claims 1 to 6, comprising an operation data acquisition module, an analysis module and a voltage stabilization module; The operation data acquisition module is used to acquire the operation data of each electrical device; The analysis module is used to input the operation data into the recognition model to obtain the operation status of each electrical device; The voltage stabilizing module is used to cut off the current power supply of the electrical equipment when the operating state is an abnormal voltage state, connect the bypass power supply, and supply power to the electrical equipment through the bypass power supply.

8. The voltage sag stabilization system according to claim 7, characterized in that: It also includes an adjustment coefficient acquisition module, which includes a construction submodule, an initialization submodule, a setting submodule, a virtual vector acquisition submodule, a loop module, a relationship acquisition submodule and a coefficient acquisition submodule; The construction submodule is used to build a virtual twin model; The initialization submodule is used to initialize the X value to 1; The setting submodule is used to obtain the X value, which is used as the startup quantity of the electrical equipment in the virtual twin model; The virtual vector acquisition submodule is used to sequentially install the electrical equipment into the virtual twin model, set the operating voltage to the standard voltage, and obtain the first operating state vector of the electrical equipment; The loop module is used to increase the value of X by 1 and determine whether the value of X is greater than the value of N, where the value of N is the total number of electrical devices in a certain circuit. If not, the setting submodule and the virtual vector acquisition submodule are repeatedly called. If greater, the first operation state vector set of the electrical device is output; The relationship acquisition submodule is used to calculate the similarity between two nodes in the first state set, and construct a connection matrix based on the similarity; The coefficient acquisition submodule is used to calculate the attention coefficient between each adjacent vector through the learning function and the connection matrix; The adjustment coefficient of the weight matrix in the recognition model is obtained through the attention coefficient.

9. The voltage sag stabilization system according to claim 8, characterized in that: The analysis module includes a marking submodule, a determination submodule and a result judgment submodule; The marking submodule is used to mark the number of the acquired data abnormal operation status as a first number, and mark the data abnormal operation status as a second vector; The determining submodule is used to obtain a corresponding first operating state vector in the first operating state vector set based on the first quantity; The result judgment submodule is used to map the plurality of second vectors and the first operation state vector to the same space, and then measure whether an abnormal voltage state occurs through the judgment matrix.

Images