Low-voltage transformer area line loss diagnosis method, system and device and storage medium

Abstract

The invention discloses a low-voltage transformer area line loss diagnosis method, system and equipment and a storage medium, and the method comprises the steps: collecting the voltage and current of nodes in each branch box and a metering box, and calculating the electric quantity, phase angle, fundamental current amplitude and odd harmonic current amplitude of a set number; calculating the weight of each node; according to the electric quantity and the corresponding normalized weight, calculating transformer area bus loss and all segment line loss, and calculating all comprehensive segment line loss rates by combining the relative change rates corresponding to all the line loss; and constructing and training a line loss time sequence prediction model, inputting all the comprehensive segment line loss rates at the moment into the line loss time sequence prediction model to predict all the comprehensive segment line loss rates at the next moment, and performing early warning when the predicted comprehensive segment line loss rates at the next moment exceed a set line loss rate threshold. According to the method, the bus loss and the segment line loss are accurately calculated, the comprehensive segment line loss rate at the next moment is predicted, early warning can be carried out in advance, and the field operation efficiency is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of line loss diagnosis, and specifically relates to a low-voltage area line loss diagnosis method, system, equipment and storage medium. Background Art

[0002] Substation line loss is an important operating technical indicator of power grid enterprises, which consists of technical line loss and management line loss. Due to the existence of technical line loss, substation line loss is generally within a certain range. If it exceeds the normal range, it is considered abnormal line loss and requires manual investigation and treatment.

[0003] CN119125667A provides a method, device, electronic device and medium for calculating line loss in a low-voltage substation. The method is applied to a three-phase IoT meter of a target low-voltage substation, including: reading the electrical quantity of the three-phase IoT meter and the electrical quantity of each single-phase meter, and obtaining the electrical characteristic quantity of the target low-voltage substation; calculating a first line loss result based on the electrical quantity of the three-phase IoT meter and the electrical quantity of each single-phase meter; inputting the electrical characteristic quantity into a pre-trained line loss calculation model to obtain a second line loss result; wherein the line loss calculation model is obtained by training a random forest model based on an improved bat algorithm; performing weighted calculation on the first line loss result and the second line loss result to obtain the line loss result of the target low-voltage substation. However, the invention is affected by the data collection completeness rate and clock synchronization rate, and the calculation accuracy of line loss in low-voltage substations is not high. At the same time, it does not take into account the complexity of the low-voltage substation power supply network and the large number of cable buried line scenarios, involving low-voltage lines, residential meter boxes, electricity meters, and transformers and other equipment. When front-line personnel face difficult substations such as high losses, negative losses, unreasonable deviations between the two rates, and suspected unmetered electricity use, it is difficult to troubleshoot problems. Moreover, with the continuous improvement of the penetration rate of new loads such as new energy vehicles, 5G base stations, and distributed photovoltaics, the complexity of power supply and use in substations has increased, which has aggravated the difficulty of line loss investigation and management in low-voltage substations. Summary of the invention

[0004] In order to solve the problems of difficulty and low accuracy in line loss measurement and calculation in the prior art, the present invention provides a low-voltage substation line loss diagnosis device, method and system, which synchronously measures the key electrical parameters of the substation nodes in real time through multiple node measurement units, comprehensively calculates and predicts line loss abnormal points, and improves the efficiency of substation line loss management.

[0005] The present invention adopts the following technical solution.

[0006] The first aspect of the present invention provides a low voltage area line loss diagnosis method, which is characterized by comprising the following contents: Survey the nodes contained in the branch boxes and meter boxes at each level from the transformer outlet to the meter box step by step; The voltage and current of the node are collected at the set time interval, and the electrical quantity, phase angle, fundamental current amplitude and the set number of odd harmonic current amplitudes of the node are calculated based on the voltage and current; The weight of each node is calculated according to the phase angle of the node, the fundamental current amplitude and the amplitude of a set number of odd harmonic currents, and the weight is normalized; Calculate the total line loss of the substation area and all segment line losses according to the power and the corresponding normalized weights, and calculate all comprehensive segment line loss rates according to the relative change rates of the total line loss of the substation area, all segment line losses and all line losses; Construct and train a line loss time series prediction model, input all the comprehensive segmented line loss rates at this moment into the line loss time series prediction model to predict all the comprehensive segmented line loss rates at the next moment, and issue an early warning when the predicted comprehensive segmented line loss rate at the next moment exceeds the set line loss rate threshold; add all the comprehensive segmented line loss rates at this moment to the training set to update the line loss time series prediction model.

[0007] Preferably, the nodes included in the branch boxes and meter boxes at all levels from the transformer outlet to the meter box are specifically: Each level of branch box contains several branch boxes, and the A-phase, B-phase, and C-phase cables in front of the outlet terminals of each branch box serve as a node respectively; The A-phase, B-phase, and C-phase cables in each meter box each serve as a node.

[0008] Preferably, the weight of each node is calculated according to the phase angle of the node, the fundamental current amplitude and the set number of odd harmonic current amplitudes, and the weight is normalized, specifically: Weight The calculation formula is:

[0009] in, , is the number of branch box levels; For the x indivual k The first branch box j The weight of the node, when k is equal to m +1, k The level branch box is a metering box; , j 1, 2, and 3 correspond to the nodes of phase A, phase B, and phase C respectively; is the fundamental current amplitude of the node; is the 2i+1 harmonic current amplitude of the node; N is the number of odd harmonic currents set; is the phase angle of the node; , These are all set sensitivity parameters; is the importance ratio of the set 2i+1 harmonic current; The normalized weight is obtained by dividing the weight of each node by the sum of the weights of all nodes.

[0010] Preferably, the calculation of the area total line loss and the segment line loss is performed by multiplying the power by the corresponding normalized weight, specifically: This moment x indivual k Sectional line loss of branch box Specifically:

[0011] in, For the x indivual k The first branch box j The power of each node; After normalization, x indivual k The first branch box j The weight of each node; For the x indivual k The first branch box connected to the indivual k +1 level branch box j The power of each node, is the normalized x indivual k The first branch box connected to the indivual k +1 level branch box j The weight of each node; For the x indivual k Connected to the branch box k +Number of level 1 branch boxes; The total line loss is the sum of all segment line losses.

[0012] Preferably, the total line loss and segment line loss of the substation area are calculated according to the power and the corresponding normalized weights, and the comprehensive segment line loss rate is calculated according to the relative change rate of the total line loss of the substation area and all segment line losses and all line losses, specifically: This moment x indivual k Comprehensive segment line loss rate of branch box The calculation formula is:

[0013] in, is the total bus loss at this moment; For this moment x indivual k The relative change rate of the segment line loss of the branch box, Equal to the difference between this moment and the previous moment x indivual k The difference of the segment line loss of the branch box at the first level is divided by the x indivual k Sectional line loss of the branch box; is the relative change rate of the total bus loss at this moment, It is equal to the difference between the bus loss at this moment and the previous moment divided by the bus loss at this moment.

[0014] Preferably, the line loss time series prediction model adopts an echo state network, which includes an input layer, a dynamic reservoir and an output layer, and the dynamic reservoir is composed of neurons connected by random sparse connections; The dynamic reservoir state update function of the echo state network is:

[0015] in, is the dynamic reservoir state vector at time step t; is the vector output by the input layer at time step t; is the dynamic reservoir state vector at time step t-1; is the weight matrix from the input layer to the dynamic reservoir; is the weight matrix inside the dynamic reservoir; is the bias term inside the dynamic reservoir; is a non-linear activation function.

[0016] Preferably, all the comprehensive segmented line loss rates at this moment are added to the training set to update the line loss time series prediction model, and all weight matrices of the echo state network are not updated and trained, and only the weight matrix of the output layer of the network is updated.

[0017] The second aspect of the present invention provides a low voltage area line loss diagnosis system using the method described in the first aspect of the present invention, comprising a plurality of node measurement units, a weight calculation unit, a comprehensive segmented line loss rate calculation unit and a prediction unit, characterized in that: The node measurement unit is installed on each node contained in the branch box and the meter box at each level from the outlet of the transformer in the substation area to the meter box; it is used to collect the voltage and current of the node at a set time interval, and calculate the electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes of the node according to the voltage and current; Weight calculation unit: used to calculate the weight of each node according to the phase angle of the node, the fundamental current amplitude and the set number of odd harmonic current amplitudes, and normalize the weight; Comprehensive section line loss rate calculation unit: used to calculate the total line loss of the substation area and all section line losses according to the power and the corresponding normalized weights, and calculate all comprehensive section line loss rates according to the total line loss of the substation area and all section line losses and the relative change rate corresponding to all line losses; Prediction unit: used to build and train the line loss time series prediction model. All the comprehensive segmented line loss rates at this moment are input into the line loss time series prediction model to predict all the comprehensive segmented line loss rates at the next moment. When the predicted comprehensive segmented line loss rate at the next moment exceeds the set line loss rate threshold, an early warning is issued; all the comprehensive segmented line loss rates at this moment are added to the training set to update the line loss time series prediction model.

[0018] Preferably, the node measurement unit is composed of a voltage measurement module, a current measurement module, a data storage module, a data communication module, a synchronous clock module, a power supply module and a microprocessor; The voltage measurement module and the current measurement module synchronously measure the voltage and current at the node under the control of the microprocessor; the synchronous clock module provides clock information for the microprocessor; the microprocessor calculates the electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes based on the measured voltage and current; the data storage module is used to store the measured voltage and current values, as well as the calculated electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes; the data communication module is used for the node measurement unit to communicate with the weight calculation unit and the comprehensive segmented line loss rate calculation unit.

[0019] Preferably, the voltage measurement module adopts a direct access mode, and the voltage measurement module is directly connected to the measured node through a wire, the connection between the wire and the voltage measurement module adopts a banana head or a BNC interface, and the connection between the wire and the measured node adopts an alligator clip or a neodymium iron boron magnet; The current measurement module adopts an open-type current clamp access mode, the open-type current clamp is clamped on the phase line, and the lead-out line of the open-type current clamp is connected to the current measurement module by using a BNC interface.

[0020] The third aspect of the present invention proposes a device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the steps of the low-voltage substation line loss diagnosis method described in the first aspect of the present invention.

[0021] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, uses the steps of the low-voltage substation line loss diagnosis method described in the first aspect of the present invention.

[0022] The beneficial effect of the present invention is that, compared with the prior art, the present invention calculates various electrical parameters by collecting the voltage and current of each node; calculates the weight of each node according to the electrical parameters, and comprehensively adjusts the calculated electrical quantity according to the operating status of each node in the low-voltage substation, so as to adapt to line loss detection in complex situations. The main line loss and segmented line loss are calculated according to the electrical quantity and weight, and only the node measurement and control unit needs to be installed without other debugging, which reduces the equipment installation and debugging time. The comprehensive segmented line loss rate is calculated by the main line loss and segmented line loss and their change rate, so that the measurement results of abnormal detection are more accurate. The comprehensive segmented line loss rate at the next moment is predicted by the time series prediction model, and early warning can be given in advance, which greatly improves the efficiency of on-site operations and saves labor costs. After the prediction, the comprehensive segmented line loss rate at this moment is added to the training set to update the model, complete online learning, and improve the accuracy of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a flow chart of the line loss diagnosis method for low voltage area; Figure 2 Schematic diagram of the installation of node measurement units in low-voltage distribution station areas; Figure 3 Block diagram of the node measurement unit; Figure 4 An example diagram of the installation of a node measurement unit in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0025] like Figure 1 As shown, Embodiment 1 of the present invention proposes a low voltage area line loss diagnosis method, which is characterized by comprising the following contents: Survey the nodes contained in the branch boxes and meter boxes at each level from the transformer outlet to the meter box step by step; The voltage and current of the node are collected at the set time interval, and the electrical quantity, phase angle, fundamental current amplitude and the set number of odd harmonic current amplitudes of the node are calculated based on the voltage and current; It should be noted that the power is calculated by thinking of voltage and current, and the electric quantity is calculated by integrating the power over time. The phase angle calculation formula is:

[0026] in, is power; , are the effective values ​​of voltage and current, is the phase angle.

[0027] The fundamental current amplitude and the amplitude of a set number of odd harmonic currents are obtained by performing Fourier analysis on the measured current signal to obtain the amplitudes of the fundamental wave and each harmonic component.

[0028] Specifically, the number set in this embodiment is 2, that is, the third and fifth harmonic current amplitudes are calculated; The weight of each node is calculated according to the phase angle of the node, the fundamental current amplitude and the amplitude of a set number of odd harmonic currents, and the weight is normalized; Calculate the total line loss of the substation area and all segment line losses according to the power and the corresponding normalized weights, and calculate all comprehensive segment line loss rates according to the relative change rates of the total line loss of the substation area, all segment line losses and all line losses; Construct and train a line loss time series prediction model, input all the comprehensive segmented line loss rates at this moment into the line loss time series prediction model to predict all the comprehensive segmented line loss rates at the next moment, and issue an early warning when the predicted comprehensive segmented line loss rate at the next moment exceeds the set line loss rate threshold; add all the comprehensive segmented line loss rates at this moment to the training set to update the line loss time series prediction model.

[0029] Preferably, the nodes included in the branch boxes and meter boxes at all levels from the transformer outlet to the meter box are specifically: Each level of branch box contains several branch boxes, and the A-phase, B-phase, and C-phase cables in front of the outlet terminals of each branch box serve as a node respectively; The A-phase, B-phase, and C-phase cables in each meter box each serve as a node.

[0030] Specifically, the low voltage distribution station area of ​​this embodiment is as follows Figure 2 As shown, each level of branch boxes includes primary and secondary branch boxes.

[0031] Preferably, the weight of each node is calculated according to the phase angle of the node, the fundamental current amplitude and the set number of odd harmonic current amplitudes, and the weight is normalized, specifically: Weight The calculation formula is:

[0032] in, , is the number of branch box levels; For the x indivual k The first branch boxj The weight of the node, when k is equal to m +1, k The level branch box is a metering box; , j 1, 2, and 3 correspond to the nodes of phase A, phase B, and phase C respectively; is the fundamental current amplitude of the node; is the 2i+1 harmonic current amplitude of the node; N is the number of odd harmonic currents set; is the phase angle of the node; , These are all set sensitivity parameters; is the importance ratio of the set 2i+1 harmonic current; Specifically, in this embodiment Set to 0.2; Set to 0.5; , Set to 0.8 and 0.6 respectively.

[0033] The normalized weight is obtained by dividing the weight of each node by the sum of the weights of all nodes.

[0034] Preferably, the calculation of the area total line loss and the segment line loss is performed by multiplying the power by the corresponding normalized weight, specifically: This moment x indivual k Sectional line loss of branch box Specifically:

[0035] in, For the x indivual k The first branch box j The power of each node; After normalization, x indivual k The first branch box j The weight of each node; For the x indivual k The first branch box connected to the indivual k +1 level branch box j The power of each node, is the normalized x indivual k The first branch box connected to the indivual k +1 level branch box j The weight of each node; For the xindivual k Connected to the branch box k +Number of level 1 branch boxes; The total line loss is the sum of all segment line losses.

[0036] Preferably, the total line loss and segment line loss of the substation area are calculated according to the power and the corresponding normalized weights, and the comprehensive segment line loss rate is calculated according to the relative change rate of the total line loss of the substation area and all segment line losses and all line losses, specifically: This moment x indivual k Comprehensive segment line loss rate of branch box The calculation formula is:

[0037] in, is the total bus loss at this moment; For this moment x indivual k The relative change rate of the segment line loss of the branch box, Equal to the difference between this moment and the previous moment x indivual k The difference of the segment line loss of the branch box at the first level is divided by the x indivual k Sectional line loss of the branch box; is the relative change rate of the total bus loss at this moment, It is equal to the difference between the bus loss at this moment and the previous moment divided by the bus loss at this moment.

[0038] Preferably, the line loss time series prediction model adopts an echo state network, which includes an input layer, a dynamic reservoir and an output layer, and the dynamic reservoir is composed of neurons connected by random sparse connections; The dynamic reservoir state update function of the echo state network is:

[0039] in, is the dynamic reservoir state vector at time step t; is the vector output by the input layer at time step t; is the dynamic reservoir state vector at time step t-1; is the weight matrix from the input layer to the dynamic reservoir; is the weight matrix inside the dynamic reservoir; is the bias term inside the dynamic reservoir; is a non-linear activation function.

[0040] Preferably, when adding all the comprehensive segmented line loss rates at the present moment to the training set to update the line loss time series prediction model, all weight matrices of the echo state network are not updated and trained, and only the weight matrix of the output layer of the network is updated.

[0041] Embodiment 2 of the present invention proposes a low voltage area line loss diagnosis system using the method described in Embodiment 1 of the present invention, comprising a plurality of node measurement units, a weight calculation unit, a comprehensive segmented line loss rate calculation unit and a prediction unit, characterized in that: The node measurement unit is installed on each node contained in the branch box and the meter box at each level from the outlet of the transformer in the substation area to the meter box; it is used to collect the voltage and current of the node at a set time interval, and calculate the electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes of the node according to the voltage and current; Weight calculation unit: used to calculate the weight of each node according to the phase angle of the node, the fundamental current amplitude and the set number of odd harmonic current amplitudes, and normalize the weight; Comprehensive section line loss rate calculation unit: used to calculate the total line loss of the substation area and all section line losses according to the power and the corresponding normalized weights, and calculate all comprehensive section line loss rates according to the total line loss of the substation area and all section line losses and the relative change rate corresponding to all line losses; Prediction unit: used to build and train the line loss time series prediction model. All the comprehensive segmented line loss rates at this moment are input into the line loss time series prediction model to predict all the comprehensive segmented line loss rates at the next moment. When the predicted comprehensive segmented line loss rate at the next moment exceeds the set line loss rate threshold, an early warning is issued; all the comprehensive segmented line loss rates at this moment are added to the training set to update the line loss time series prediction model.

[0042] Preferably, if Figure 3 As shown, the node measurement unit is composed of a voltage measurement module, a current measurement module, a data storage module, a data communication module, a synchronous clock module, a power supply module and a microprocessor; The voltage measurement module and the current measurement module synchronously measure the voltage and current at the node under the control of the microprocessor; the synchronous clock module provides clock information for the microprocessor; the microprocessor calculates the electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes based on the measured voltage and current; the data storage module is used to store the measured voltage and current values, as well as the calculated electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes; the data communication module is used for the node measurement unit to communicate with the weight calculation unit and the comprehensive segmented line loss rate calculation unit.

[0043] like Figure 4As shown, preferably, the voltage measurement module adopts a direct access mode, and the voltage measurement module is directly connected to the measured node through a wire, and the connection between the wire and the voltage measurement module adopts a banana head or a BNC interface, and the connection between the wire and the measured node adopts an alligator clip or a neodymium iron boron magnet; The current measurement module adopts an open-type current clamp access mode, the open-type current clamp is clamped on the phase line, and the lead-out line of the open-type current clamp is connected to the current measurement module by using a BNC interface.

[0044] Embodiment 3 of the present invention proposes a device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the steps of the low-voltage substation line loss diagnosis method described in Embodiment 1 of the present invention.

[0045] Embodiment 4 of the present invention proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the low-voltage substation line loss diagnosis method described in Embodiment 1 of the present invention are used.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for diagnosing line loss in a low voltage area, characterized in that: Includes the following: Survey the nodes contained in the branch boxes and meter boxes at each level from the transformer outlet to the meter box step by step; The voltage and current of the node are collected at the set time interval, and the electrical quantity, phase angle, fundamental current amplitude and the set number of odd harmonic current amplitudes of the node are calculated based on the voltage and current; The weight of each node is calculated according to the phase angle of the node, the fundamental current amplitude and the amplitude of a set number of odd harmonic currents, and the weight is normalized; Calculate the total line loss of the substation area and all segment line losses according to the power and the corresponding normalized weights, and calculate all comprehensive segment line loss rates according to the relative change rates of the total line loss of the substation area, all segment line losses and all line losses; Construct and train a line loss time series prediction model, input all the comprehensive segmented line loss rates at this moment into the line loss time series prediction model to predict all the comprehensive segmented line loss rates at the next moment, and issue an early warning when the predicted comprehensive segmented line loss rate at the next moment exceeds the set line loss rate threshold; add all the comprehensive segmented line loss rates at this moment to the training set to update the line loss time series prediction model.

2. A low voltage area line loss diagnosis method according to claim 1, characterized in that: The nodes included in the branch boxes and meter boxes at all levels from the transformer outlet to the meter box are specifically: Each level of branch box contains several branch boxes, and the A-phase, B-phase, and C-phase cables in front of the outlet terminals of each branch box serve as a node respectively; The A-phase, B-phase, and C-phase cables in each meter box each serve as a node.

3. A low voltage area line loss diagnosis method according to claim 2, characterized in that: The weight of each node is calculated according to the phase angle of the node, the fundamental current amplitude and the set number of odd harmonic current amplitudes, and the weight is normalized, specifically: Weight The calculation formula is: in, , is the number of branch box levels; For the x indivual k The first branch box j The weight of the node, when k is equal to m +1, k The level branch box is a metering box; , j 1, 2, and 3 correspond to the nodes of phase A, phase B, and phase C respectively; is the fundamental current amplitude of the node; is the 2i+1 harmonic current amplitude of the node; N is the number of odd harmonic currents set; is the phase angle of the node; , These are all set sensitivity parameters; is the importance ratio of the set 2i+1 harmonic current; The normalized weight is obtained by dividing the weight of each node by the sum of the weights of all nodes.

4. A low voltage area line loss diagnosis method according to claim 3, characterized in that: The total line loss and segment line loss of the substation area are calculated by multiplying the power by the corresponding normalized weight, specifically: This moment x indivual k Sectional line loss of branch box Specifically: in, For the x indivual k The first branch box j The power of each node; After normalization, x indivual k The first branch box j The weight of each node; For the x indivual k The first branch box connected to the indivual k +1 level branch box j The power of each node, is the normalized x indivual k The first branch box connected to the indivual k +1 level branch box j The weight of each node; For the x indivual k Connected to the branch box k +Number of level 1 branch boxes; The total line loss is the sum of all segment line losses.

5. A low voltage area line loss diagnosis method according to claim 4, characterized in that: The total line loss and segment line loss of the substation area are calculated according to the power and the corresponding normalized weights, and the comprehensive segment line loss rate is calculated according to the relative change rate of the total line loss of the substation area and all segment line losses and all line losses, specifically: This moment x indivual k Comprehensive segment line loss rate of branch box The calculation formula is: in, is the total bus loss at this moment; For this moment x indivual k The relative change rate of the segment line loss of the branch box, Equal to the difference between this moment and the previous moment x indivual k The difference of the segment line loss of the branch box at the current moment is divided by x indivual k Sectional line loss of the branch box; is the relative change rate of the total bus loss at this moment, It is equal to the difference between the bus loss at this moment and the previous moment divided by the bus loss at this moment.

6. A low voltage area line loss diagnosis method according to claim 1, characterized in that: The line loss time series prediction model adopts an echo state network, which includes an input layer, a dynamic reservoir and an output layer, and the dynamic reservoir is composed of neurons connected by random sparse connections; The dynamic reservoir state update function of the echo state network is: in, is the dynamic reservoir state vector at time step t; is the vector output by the input layer at time step t; is the dynamic reservoir state vector at time step t-1; is the weight matrix from the input layer to the dynamic reservoir; is the weight matrix inside the dynamic reservoir; is the bias term inside the dynamic reservoir; is a non-linear activation function.

7. A low voltage area line loss diagnosis method according to claim 6, characterized in that: When adding all the comprehensive segmented line loss rates at this moment into the training set to update the line loss time series prediction model, all weight matrices of the echo state network are not updated and trained, and only the weight matrix of the output layer of the network is updated.

8. A low voltage area line loss diagnosis system using the method according to any one of claims 1 to 7, comprising a plurality of node measurement units, a weight calculation unit, a comprehensive segmented line loss rate calculation unit and a prediction unit; characterized in that: The node measurement unit is installed on each node contained in the branch box and the meter box at each level from the outlet of the transformer in the substation area to the meter box; it is used to collect the voltage and current of the node at a set time interval, and calculate the electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes of the node according to the voltage and current; Weight calculation unit: used to calculate the weight of each node according to the phase angle of the node, the fundamental current amplitude and the set number of odd harmonic current amplitudes, and normalize the weight; Comprehensive section line loss rate calculation unit: used to calculate the total line loss of the substation area and all section line losses according to the power and the corresponding normalized weights, and calculate all comprehensive section line loss rates according to the total line loss of the substation area and all section line losses and the relative change rate corresponding to all line losses; Prediction unit: used to build and train the line loss time series prediction model. All the comprehensive segmented line loss rates at this moment are input into the line loss time series prediction model to predict all the comprehensive segmented line loss rates at the next moment. When the predicted comprehensive segmented line loss rate at the next moment exceeds the set line loss rate threshold, an early warning is issued; all the comprehensive segmented line loss rates at this moment are added to the training set to update the line loss time series prediction model.

9. A low voltage area line loss diagnosis system according to claim 8, characterized in that: The node measurement unit is composed of a voltage measurement module, a current measurement module, a data storage module, a data communication module, a synchronous clock module, a power supply module and a microprocessor; The voltage measurement module and the current measurement module synchronously measure the voltage and current at the node under the control of the microprocessor; the synchronous clock module provides clock information for the microprocessor; The microprocessor calculates the electric quantity, phase angle, fundamental current amplitude and the amplitude of a set number of odd harmonic currents according to the measured voltage and current; The data storage module is used to store the measured voltage and current values, as well as the calculated electrical quantity, phase angle, fundamental current amplitude and a set number of odd harmonic current amplitudes; The data communication module is used for the node measurement unit to communicate with the weight calculation unit and the comprehensive segment line loss rate calculation unit.

10. A low voltage area line loss diagnosis system according to claim 9, characterized in that: The voltage measurement module adopts a direct access mode, and the voltage measurement module is directly connected to the measured node through a wire. The connection between the wire and the voltage measurement module adopts a banana plug or a BNC interface, and the connection between the wire and the measured node adopts an alligator clip or a neodymium iron boron magnet; The current measurement module adopts an open-type current clamp access mode, the open-type current clamp is clamped on the phase line, and the lead-out line of the open-type current clamp is connected to the current measurement module by using a BNC interface.

11. A device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the steps of the low-voltage substation line loss diagnosis method according to any one of claims 1 to 9.

12. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, uses the steps of the low-voltage substation line loss diagnosis method according to any one of claims 1 to 9.

Images