Fault judgment circuit based on current signal

Abstract

The invention relates to the field of power distribution of a power system, in particular to a fault judgment circuit based on a current signal, which comprises a topological network power transmission network, and a beat frequency device, a mutual inductor, a signal acquisition unit and an execution unit are arranged on a node; the signal acquisition unit is used for receiving a signal transmitted by the beat frequency device of the upstream node through the mutual inductor, and is also used for receiving a signal output by the mutual inductor of the corresponding node; the signal processing unit is used for carrying out noise reduction processing on the signal received by the signal acquisition unit; the data self-learning processing system is connected with the signal processing unit to calculate data denoised by the signal processing unit, and is connected with the beat frequency device of each node to control parameters of output signals of the beat frequency devices; the execution units are arranged on the corresponding nodes and connected with the data self-learning processing system, the data self-learning processing system controls the execution units to enable the corresponding nodes to be disconnected and connected according to data of the signal processing unit, and self-repairing can be achieved when a circuit breaks down.

Description

Technical Field

[0001] The present invention relates to the field of power distribution of electric power systems, and in particular to a fault judgment circuit based on current signals. Background Art

[0002] The distribution network refers to a power grid that receives electric energy from the transmission network or regional power plants, and distributes it locally or step by step according to the voltage to various users through distribution facilities. It is composed of overhead lines, cables, towers, distribution transformers, disconnectors, reactive power compensators and some ancillary facilities, and plays an important role in distributing electric energy in the power grid. With the increasing demand for electricity in various industries, especially the rapid increase in industrial and civil loads in areas with rapid economic development, the number of distribution systems and their distribution equipment has increased significantly, the network structure of the distribution system has become more and more complex, and the power supply operation mode has changed more and more frequently. The 110kv high-voltage distribution system is a relatively critical link in the entire distribution network, and there are increasingly higher requirements for the stability of the distribution network operation, the effectiveness of regulation, and the reliability of fault section judgment. Therefore, accurate judgment of the fault section of the more complex distribution network and rapid handling of faults have become an important part of the distribution network work.

[0003] Topology is currently used in regional distribution networks, and it is an important means of distribution network analysis. At present, distribution automation can only correctly handle centralized equipment, some of which cannot properly cooperate with boundary switches, etc. In the event of a fault, its self-recovery ability cannot meet actual needs. Summary of the invention

[0004] The purpose of the present invention is to provide a fault judgment circuit based on current signal to solve the above problem and quickly restore power supply when a fault occurs in the line.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A fault judgment circuit based on current signal, comprising:

[0007] The topological network transmission network has frequency beaters, mutual inductors, signal acquisition units, and execution units at its nodes;

[0008] The signal acquisition unit is connected to the mutual inductor of the corresponding node, receives the signal transmitted by the beat frequency device of the upstream node through the mutual inductor, and is also used to receive the signal output by the mutual inductor of the corresponding node;

[0009] A signal processing unit, connected to the signal acquisition unit, and configured to perform noise reduction processing on the signal received by the signal acquisition unit;

[0010] A data self-learning processing system, connected to the signal processing unit to calculate the data after noise reduction of the signal processing unit, and connected to the beater of each node to control the parameters of the beater output signal;

[0011] An execution unit is arranged on a corresponding node and connected to the data self-learning processing system. The data self-learning processing system controls the execution unit to disconnect or connect the corresponding nodes according to the data of the signal processing unit;

[0012] The topology network transmission network is a ring network topology circuit or a hybrid topology circuit.

[0013] Preferably, the output signal of the beat frequency generator includes time, phase, amplitude and frequency.

[0014] Preferably, the signal processing unit is a wavelet transform noise reduction processor.

[0015] Preferably, the nodes of the topological network transmission network are numbered, and the beat frequency devices, mutual inductors, signal acquisition units, and execution units corresponding to the nodes are also numbered corresponding to the nodes;

[0016] Processing the time, phase, amplitude and frequency output signals of the beat frequency generator after noise reduction by the signal processing unit and the voltage and current signals of each node by the data self-learning processing system;

[0017] After being trained by the data self-learning processing system, when a fault occurs, the corresponding numbers of the nodes, frequency beaters, mutual inductors, signal acquisition units, and execution units are used to control the corresponding frequency beaters to transform the detection signal, and the timing, phase, amplitude and frequency, voltage, and current signals on the downstream nodes are combined to control the execution unit that can restore the corresponding fault position to operate, thereby realizing the path of the control circuit.

[0018] Preferably, the data self-learning processing system is configured as follows:

[0019] Create a Sequential model setting and set the number of input channels. The number of input channels is the same as the number of parameters to be detected. The number of input channels is 8, which are time, phase, amplitude, frequency, voltage, current, and number parameters.

[0020] Add a 1D convolutional layer;

[0021] Add a pooling layer with a pooling window size of 2;

[0022] Second convolution and pooling, add one-dimensional convolution layer again;

[0023] Add the maximum pooling layer again with a pooling window size of 2;

[0024] The flattening layer flattens the output of the convolutional layer into a one-dimensional vector;

[0025] Fully connected layer, set the corresponding number of neurons;

[0026] Output layer, number of neurons: set according to the number of fault categories, activation function is softmax.

[0027] Preferably, in the added one-dimensional convolution layer, the convolution layer uses 32 convolution kernels of size 3 and adopts a ReLU activation function.

[0028] Preferably, in the one-dimensional convolution layer added again, the convolution layer uses 64 convolution kernels of size 3, and the activation function is ReLU.

[0029] Preferably, the number of neurons in the fully connected layer is 128, and the activation function is ReLU.

[0030] The present invention has the following technical effects:

[0031] The present invention can collect voltage and current information in the circuit by arranging a beater, a mutual inductor, a signal acquisition unit, an execution unit, and a signal acquisition unit on each node. It can also detect the relevant information of the information output by the beater of the upstream node and transmit it to the downstream node through the components on each node. After the signal is subjected to noise reduction processing by the signal processing unit, it is self-learned by the data self-learning processing system, and various input signals are used to determine whether a fault occurs between the nodes. Then, the corresponding execution unit is controlled to open and close the circuit to achieve self-healing of faults in the power supply process of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0033] Figure 1 This is a flow chart of circuit failure recovery of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] A fault judgment circuit based on current signal, comprising:

[0037] The topological network transmission network has frequency beaters, mutual inductors, signal acquisition units, and execution units at its nodes;

[0038] A signal acquisition unit is connected to the mutual inductor of the corresponding node, receives the signal transmitted by the beat frequency device of the upstream node through the mutual inductor, and is also used to receive the signal output by the mutual inductor of the corresponding node;

[0039] A signal processing unit, connected to the signal acquisition unit, and configured to perform noise reduction processing on the signal received by the signal acquisition unit;

[0040] The data self-learning processing system is connected to the signal processing unit to calculate the data after noise reduction of the signal processing unit, and is connected to the beater of each node to control the parameters of the beater output signal;

[0041] The execution unit is arranged on the corresponding node and connected to the data self-learning processing system. The data self-learning processing system controls the execution unit to disconnect or connect the corresponding nodes according to the data of the signal processing unit.

[0042] The topology network transmission network is a ring network topology circuit or a hybrid topology circuit.

[0043] The present invention can collect voltage and current information in the circuit by arranging a beater, a mutual inductor, a signal acquisition unit, an execution unit, and a signal acquisition unit on each node. It can also detect the relevant information of the information output by the beater of the upstream node and transmit it to the downstream node through the components on each node. After the signal is subjected to noise reduction processing by the signal processing unit, it is self-learned by the data self-learning processing system, and various input signals are used to determine whether a fault occurs between the nodes. Then, the corresponding execution unit is controlled to open and close the circuit to achieve self-healing of faults in the power supply process of the power grid.

[0044] According to the further optimization scheme, the output signal of the beat frequency generator includes time, phase, amplitude and frequency, the mutual inductor includes a PT mutual inductor and a CT mutual inductor, the PT mutual inductor is used to detect the voltage of the circuit on the node, and the CT mutual inductor is used to detect the current of the circuit on the node.

[0045] To further optimize the solution, the signal processing unit is a wavelet transform noise reduction processor.

[0046] To further optimize the scheme, the nodes of the topological network transmission network are numbered, and the beat frequency generators, mutual inductors, signal acquisition units, and execution units corresponding to the nodes are also numbered corresponding to the nodes;

[0047] The data self-learning processing system processes the time, phase, amplitude and frequency output signals of the beat frequency generator after noise reduction by the signal processing unit, as well as the voltage and current signals of each node;

[0048] After being trained by the data self-learning processing system, when a fault occurs, the corresponding numbers of the nodes, frequency beaters, mutual inductors, signal acquisition units, and execution units are used to control the corresponding frequency beaters to transform the detection signal, and the timing, phase, amplitude and frequency, voltage, and current signals on the downstream nodes are combined to control the execution unit operation that can restore the corresponding fault location, thereby realizing the path of the control circuit.

[0049] To further optimize the solution, the data self-learning processing system is set up as follows:

[0050] Create a Sequential model setting and set the number of input channels. The number of input channels should be the same as the number of parameters to be detected.

[0051] Add a 1D convolutional layer;

[0052] Add a pooling layer with a pooling window size of 2;

[0053] Second convolution and pooling, add one-dimensional convolution layer again;

[0054] Add the maximum pooling layer again with a pooling window size of 2;

[0055] The flattening layer flattens the output of the convolutional layer into a one-dimensional vector;

[0056] For the fully connected layer, set the corresponding number of neurons; the number of neurons is twice the number of neurons in the output layer. For the output layer, the number of neurons is set according to the number of fault categories, and the activation function is softmax.

[0057] To further optimize the solution, a one-dimensional convolution layer is added. The convolution layer uses 32 convolution kernels of size 3 and adopts the ReLU activation function.

[0058] To further optimize the solution, we added a one-dimensional convolution layer again. The convolution layer used 64 convolution kernels of size 3 and the activation function was ReLU.

[0059] To further optimize the scheme, the number of neurons in the fully connected layer is 128 and the activation function is ReLU.

[0060] Train the model:

[0061] Number of training cycles: 100;

[0062] Batch size: 64

[0063] Validation set ratio: 0.1;

[0064] After neural network training, the data self-learning processing system is connected to the corresponding node of the power grid. When a fault occurs in the circuit, the data self-learning processing system can determine the fault point through the input parameters, control the frequency beater at the fault point to change the output parameters, and further verify the function of the fault. When the fault is determined, the execution unit with the corresponding number is controlled to change the circuit breakers and paths between the nodes to achieve self-repair of the power grid fault.

[0065] In the description of the present invention, it should be understood that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside" and "outside" etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0066] The embodiments described above are only descriptions of the preferred modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A fault judgment circuit based on current signal, characterized in that: include: The topological network transmission network has frequency beaters, mutual inductors, signal acquisition units, and execution units at its nodes; The signal acquisition unit is connected to the mutual inductor of the corresponding node, receives the signal transmitted by the beat frequency device of the upstream node through the mutual inductor, and is also used to receive the signal output by the mutual inductor of the corresponding node; A signal processing unit, connected to the signal acquisition unit, and configured to perform noise reduction processing on the signal received by the signal acquisition unit; A data self-learning processing system, connected to the signal processing unit to calculate the data after noise reduction of the signal processing unit, and connected to the beater of each node to control the parameters of the beater output signal; An execution unit is arranged on a corresponding node and connected to the data self-learning processing system. The data self-learning processing system controls the execution unit to disconnect or connect the corresponding nodes according to the data of the signal processing unit; The topology network transmission network is a ring network topology circuit or a hybrid topology circuit.

2. The current signal-based fault judgment circuit according to claim 1, characterized in that: The output signal of the beat frequency generator includes time, phase, amplitude and frequency.

3. The current signal-based fault judgment circuit according to claim 2, characterized in that: The signal processing unit is a wavelet transform noise reduction processor.

4. The current signal-based fault judgment circuit according to claim 3, characterized in that: The nodes of the topological network transmission network are numbered, and the beat frequency devices, mutual inductors, signal acquisition units, and execution units corresponding to the nodes are also numbered corresponding to the nodes; Processing the time, phase, amplitude and frequency output signals of the beat frequency generator after noise reduction by the signal processing unit and the voltage and current signals of each node by the data self-learning processing system; After being trained by the data self-learning processing system, when a fault occurs, the corresponding numbers of the nodes, frequency beaters, mutual inductors, signal acquisition units, and execution units are used to control the corresponding frequency beaters to transform the detection signal, and the timing, phase, amplitude and frequency, voltage, and current signals on the downstream nodes are combined to control the execution unit that can restore the corresponding fault position to operate, thereby realizing the path of the control circuit.

5. The current signal-based fault judgment circuit according to claim 4, characterized in that: The data self-learning processing system is configured as follows: Create a Sequential model setting and set the number of input channels. The number of input channels should be the same as the number of parameters to be detected. Add a 1D convolutional layer; Add a pooling layer with a pooling window size of 2; Second convolution and pooling, add one-dimensional convolution layer again; Add the maximum pooling layer again with a pooling window size of 2; Flattening layer, which flattens the output of the convolutional layer into a one-dimensional vector; Fully connected layer, set the corresponding number of neurons; Output layer, number of neurons: set according to the number of fault categories, activation function is softmax.

6. The current signal-based fault judgment circuit according to claim 5, characterized in that: In the added one-dimensional convolution layer, the convolution layer uses 32 convolution kernels of size 3 and adopts the ReLU activation function.

7. The current signal-based fault judgment circuit according to claim 5, characterized in that: In the one-dimensional convolution layer added again, the convolution layer uses 64 convolution kernels of size 3, and the activation function is ReLU.

8. The current signal-based fault judgment circuit according to claim 5, characterized in that: The number of neurons in the fully connected layer is 128, and the activation function is ReLU.

Images