Low-voltage Distribution Network Fault Early Warning and Fault Location Method, Device, Medium and System Based on Edge Computing

Abstract

The present invention relates to a low-voltage distribution network fault early warning and fault location method, device, medium and system based on edge computing. The method includes: a monitoring terminal collects status information data of the low-voltage distribution network and sends the status information data packet to the corresponding distributed edge computing terminal; the distributed edge computing terminal uses a fault prediction model based on a neural network to perform fault determination and fault location according to the status information data packet, and sends the fault determination result and the fault location result to the monitoring center; the monitoring center generates a fault early warning information according to the fault determination result and the fault location result and pushes it. The present invention has low latency, and the computing power is deployed near the side of the distribution network grid unit of the low-voltage distribution network, and the device request is responded in real time; it has the characteristic of running with low bandwidth, migrates the work closer to the monitoring terminal, and can reduce the impact brought by the bandwidth limitation of the monitoring center.

Description

Technical Field

[0001] The present invention relates to the technical field of power detection, and particularly to a method, device, medium and system for low-voltage distribution network fault early warning and fault location based on edge computing. Background Art

[0002] A distribution network refers to a power network that receives electric energy from a transmission network or a regional power plant and distributes it locally through distribution facilities or step by step according to voltage levels to various users. It is composed of overhead lines, cables, poles, distribution transformers, disconnect switches, reactive power compensators and some auxiliary facilities, etc., and plays an important role in distributing electric energy in the power network.

[0003] The low-voltage distribution network directly provides power distribution and supply for users and is an organic part of the distribution network. There are a large number of facilities in the low-voltage distribution network and they are widely distributed, which are the difficulties and blind spots in power operation and maintenance supervision. With the development of smart grids, more and more information needs to be processed in the low-voltage distribution network, including data information, video monitoring, security information, etc. The situation in the low-voltage distribution network needs to be mastered at all times to ensure safety, so relevant personnel are required to monitor the conditions of various parts inside the distribution network in real time.

[0004] In addition to the internal operation and maintenance data of the low-voltage distribution network, when transmitting electric power from the low-voltage distribution network to users, the electricity consumption at each time period every day and the electricity consumption required in different regions also belong to large-flow data, resulting in more troublesome data processing and a large workload. In modern distribution networks, due to the addition of intelligent inspection devices, it is not necessary to frequently monitor the site manually as before, reducing certain safety hazards and improving efficiency. However, the amount of data transmitted by intelligent inspection devices to the monitoring center is constantly expanding, consuming a large amount of traffic resources. Due to the limitations of the monitoring center's resource conditions, it is inevitably affected by high latency and network instability. Due to the instant response problems caused by latency, it is impossible to process the data in the low-voltage distribution network, detect faults in time and take corresponding measures. Summary of the Invention

[0005] In order to achieve the above objects and other advantages of the present invention, the first object of the present invention is to provide a method for low-voltage distribution network fault early warning and fault location based on edge computing, including the following steps:

[0006] A monitoring terminal collects status information data of the low-voltage distribution network and sends the status information data packet to the corresponding distributed edge computing terminal;

[0007] The distributed edge computing terminal uses a fault prediction model based on a neural network to perform fault determination and fault location according to the status information data packet, and sends the fault determination result and the fault location result to the monitoring center;

[0008] The monitoring center generates a fault warning message based on the fault determination result and the fault location result and pushes it.

[0009] Further, the construction of the neural network-based fault prediction model includes the following steps:

[0010] Taking the status information data packet as the input layer and the low-voltage distribution network fault determination result and the fault location result as the output layer, a feedforward neural network fault prediction model is established, and the fault prediction model is initialized;

[0011] Using a sparse multiple linear regression model to sort the nodes in the hidden layer of the fault prediction model according to their contribution degrees to the prediction result;

[0012] Starting from the nodes with low contribution degrees, the sorted hidden layer nodes are pruned and the model prediction error is calculated;

[0013] According to the prediction error, new hidden layer nodes are added to obtain a new fault prediction model.

[0014] Further, the status information data packet sent by the monitoring terminal contains identification information for marking the data packet.

[0015] Further, the status information data in the status information data packet includes any one or more of loop current data, voltage data, residual current data, and temperature data.

[0016] Further, the input nodes of the input layer of the neural network-based fault prediction model include any one or more of loop current data, voltage data, residual current data, and temperature data, and the identification information.

[0017] Further, the distributed edge computing terminals are deployed in a distributed manner in several distribution network grid units of the low-voltage distribution network, and each distributed edge computing terminal correspondingly controls several monitoring terminals.

[0018] Further, it also includes receiving the node operation status data reported by the distributed edge computing terminal through the monitoring center, judging whether the operation status of the distributed edge computing terminal is normal according to the node operation status data, and if not, alarming the distributed edge computing terminal.

[0019] The second object of the present invention is to provide an electronic device, including: a memory on which program code is stored; a processor connected to the memory, and when the program code is executed by the processor, implementing a low-voltage distribution network fault warning and fault location method based on edge computing.

[0020] The third object of the present invention is to provide a computer-readable storage medium, on which program instructions are stored, and when the program instructions are executed, a fault warning and fault location method for a low-voltage distribution network based on edge computing is realized.

[0021] The fourth object of the present invention is to provide a fault warning and fault location system for a low-voltage distribution network based on edge computing, including a number of monitoring terminals, a number of distributed edge computing terminals, and a monitoring center; the low-voltage distribution network is divided into a number of distribution network grid units, and the distributed edge computing terminals are deployed in a distributed manner in the corresponding distribution network grid units, and each distributed edge computing terminal controls a number of monitoring terminals;

[0022] The monitoring terminals collect the status information data of the low-voltage distribution network and send the status information data packets to the corresponding distributed edge computing terminals;

[0023] The distributed edge computing terminals use a fault prediction model based on a neural network to perform fault determination and fault location according to the status information data packets, and send the fault determination results and fault location results to the monitoring center;

[0024] The monitoring center generates a fault warning information according to the fault determination result and the fault location result and pushes it.

[0025] Compared with the prior art, the beneficial effects of the present invention are:

[0026] The present invention provides a fault warning and fault location method, device, medium and system for a low-voltage distribution network based on edge computing, which has low latency, and the computing power is deployed near the side of the distribution network grid unit of the low-voltage distribution network, and the device requests real-time response; it has the characteristics of low-bandwidth operation, migrates the work closer to the monitoring terminals, and can reduce the impact brought by the bandwidth limitation of the monitoring center; it has the advantage of privacy protection, collects data locally, analyzes locally, and processes locally, effectively reducing the chance of data exposure in the public network and protecting data privacy.

[0027] The present invention can migrate a large amount of data storage and preliminary processing to the distributed edge computing terminals, and then send the fault determination information and fault location information to the monitoring center, and the monitoring center generates and pushes a warning signal, so that the supervisors can timely discover the fault and take corresponding measures to improve the fault handling efficiency.

[0028] The present invention can timely sense and respond to the faults in the low-voltage distribution network, provide a decision-making basis for the low-voltage distribution fault handling of the power grid operation and maintenance part, speed up the fault handling speed, shorten the power outage time, and improve the power supply reliability and user satisfaction.

[0029] The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and be able to implement it according to the content of the description, the following takes the preferred embodiments of the present invention and combines with the accompanying drawings to describe in detail as follows. The specific implementation manners of the present invention are given in detail by the following embodiments and their accompanying drawings. Description of the Drawings

[0030] The drawings described herein are used to provide a further understanding of the present invention and form a part of this application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention. In the drawings:

[0031] Figure 1 It is a flowchart of a method for fault warning and fault location of a low-voltage distribution network based on edge computing in Embodiment 1;

[0032] Figure 2 It is a schematic diagram of an electronic device in Embodiment 2;

[0033] Figure 3 It is a schematic diagram of a system for fault warning and fault location of a low-voltage distribution network based on edge computing in Embodiment 4. Detailed Description of the Invention

[0034] Next, in combination with the accompanying drawings and specific implementation manners, the present invention will be further described. It should be noted that, on the premise of no conflict, the following-described embodiments or technical features can be combined arbitrarily to form new embodiments.

[0035] Embodiment 1

[0036] A method for fault warning and fault location of a low-voltage distribution network based on edge computing, as Figure 1 shown, includes the following steps:

[0037] S1. The monitoring terminal collects the status information data of the low-voltage distribution network and sends the status information data packet to the corresponding distributed edge computing terminal.

[0038] In order to perform refined monitoring on the low-voltage distribution network, the low-voltage distribution network is divided into distribution network grid units. The distributed edge computing terminals are deployed in a distributed manner in several distribution network grid units of the low-voltage distribution network, and each distributed edge computing terminal corresponds to controlling several monitoring terminals.

[0039] In this embodiment, the status information data in the status information data packet includes, but is not limited to, any one or more of loop current data, voltage data, residual current data, and temperature data. The status information data packet sent by the monitoring terminal contains identification information for marking the data packet, which is used for fault location.

[0040] S2. The distributed edge computing terminal uses a neural network-based fault prediction model to perform fault determination and fault location based on the status information data packet, and sends the fault determination result and the fault location result to the monitoring center. Among them, the construction of the neural network-based fault prediction model includes the following steps:

[0041] Taking the status information data packet as the input layer and the low-voltage distribution network fault determination result and the fault location result as the output layer, a feedforward neural network fault prediction model is established and the fault prediction model is initialized. Specifically, the input nodes of the input layer of the neural network-based fault prediction model include any one or more of loop current data, voltage data, residual current data, and temperature data, as well as the identification information.

[0042] By training the feedforward neural network fault prediction model, the trained feedforward neural network fault prediction model can predict the fault determination result and the fault location result that meet the prediction accuracy for the actual data of the above input nodes. The feedforward neural network fault prediction model has the advantages of fast operation speed, strong generalization performance, and simple principle.

[0043] Since the computing power of the distributed edge computing terminal is limited, but there are redundant hidden layer nodes in the feedforward neural network fault prediction model, this embodiment improves this model, reduces the computational amount of the model, and dynamically adjusts the number of hidden layer nodes when new real-time data enters the model to improve the prediction accuracy. Specifically, it includes the following steps:

[0044] Using a sparse multiple linear regression model to sort the nodes in the hidden layer of the fault prediction model according to their contribution degrees to the prediction results;

[0045] The sorted hidden layer nodes are trimmed starting from the nodes with low contribution degrees and the model prediction error is calculated; first, according to the contribution degree sorting result, delete the last m nodes with the lowest contribution degrees, where m is set according to the prediction error. When the prediction error is small, m takes a small value and only the nodes with contribution degrees close to 0 are updated. When the prediction error is large, increase m until only the core nodes are retained.

[0046] Supplement new hidden layer nodes according to the prediction error to obtain a new fault prediction model.

[0047] S3. The monitoring center generates a fault warning message based on the fault determination result and the fault location result and pushes it.

[0048] In one embodiment, it further includes S4, receiving the node operation status data reported by the distributed edge computing terminal through the monitoring center; S5, judging whether the operation status of the distributed edge computing terminal is normal according to the node operation status data; S6, if it is not normal, alarming the distributed edge computing terminal; S7, if it is normal, continuously monitoring the fault determination result and fault location result reported by the distributed edge computing terminal. This ensures that faults can be detected in a timely manner and responded to in real time.

[0049] Embodiment 2

[0050] An electronic device 200, as Figure 2 shown, includes but is not limited to: a memory 201, on which program code is stored; a processor 202, which is connected to the memory, and when the program code is executed by the processor, a multi-window spectral peak recognition method based on signal-to-noise ratio is implemented. For a detailed description of the method, reference can be made to the corresponding description in the above method embodiments, which will not be elaborated here.

[0051] Embodiment 3

[0052] A computer-readable storage medium, on which program instructions are stored, and when the program instructions are executed, a multi-window spectral peak recognition method based on signal-to-noise ratio is implemented. For a detailed description of the method, reference can be made to the corresponding description in the above method embodiments, which will not be elaborated here.

[0053] Embodiment 4

[0054] A low-voltage distribution network fault warning and fault location system 100 based on edge computing, as Figure 3 shown, includes a monitoring center 101, a plurality of distributed edge computing terminals 102, and a plurality of monitoring terminals 103; in order to perform refined monitoring on the low-voltage distribution network, the low-voltage distribution network is divided into a plurality of distribution network grid units, and the distributed edge computing terminals are deployed in a distributed manner in the corresponding distribution network grid units, and each distributed edge computing terminal controls a plurality of monitoring terminals correspondingly.

[0055] In this embodiment, the status information data in the status information data packet includes but is not limited to any one or more of loop current data, voltage data, residual current data, and temperature data. The status information data packet sent by the monitoring terminal contains identification information for marking the data packet, which is used for fault location.

[0056] The monitoring terminal collects the status information data of the low-voltage distribution network and sends the status information data packet to the corresponding distributed edge computing terminal;

[0057] The distributed edge computing terminal uses a neural network-based fault prediction model to perform fault determination and fault location based on the status information data packet, and sends the fault determination result and the fault location result to the monitoring center. Among them, the construction of the neural network-based fault prediction model includes the following steps:

[0058] Take the status information data packet as the input layer, and the low-voltage distribution network fault determination result and the fault location result as the output layer to establish a feedforward neural network fault prediction model, and initialize the fault prediction model. Specifically, the input nodes of the input layer of the neural network-based fault prediction model include any one or more of loop current data, voltage data, residual current data, and temperature data, as well as the identification information.

[0059] By training the feedforward neural network fault prediction model, the trained feedforward neural network fault prediction model can predict the fault determination result and the fault location result that meet the prediction accuracy for the actual data of the above input nodes. The feedforward neural network fault prediction model has the advantages of fast operation speed, strong generalization performance, and simple principle.

[0060] Since the computing power of the distributed edge computing terminal is limited, but there are redundant hidden layer nodes in the feedforward neural network fault prediction model, this embodiment improves this model, reduces the computational amount of the model, and dynamically adjusts the number of hidden layer nodes when new real-time data enters the model to improve the prediction accuracy. Specifically, it includes the following steps:

[0061] Use a sparse multiple linear regression model to sort the nodes in the hidden layer of the fault prediction model according to their contribution degrees to the prediction results;

[0062] Start pruning the sorted hidden layer nodes from the nodes with low contribution degrees and calculate the model prediction error; first, according to the contribution degree sorting result, delete the last m nodes with the lowest contribution degrees, where m is set according to the prediction error. When the prediction error is small, m takes a small value, and only the nodes with contribution degrees close to 0 are updated. When the prediction error is large, increase m until only the core nodes are retained.

[0063] Supplement new hidden layer nodes according to the prediction error to obtain a new fault prediction model.

[0064] The monitoring center generates and pushes fault warning information based on the fault determination result and the fault location result.

[0065] The monitoring center receives the node operation status data reported by the distributed edge computing terminals, determines whether the operation status of the distributed edge computing terminals is normal based on the node operation status data. If it is not normal, an alarm is issued for the distributed edge computing terminals; if it is normal, the fault determination results and fault location results reported by the distributed edge computing terminals are continuously monitored. Ensure that faults can be promptly detected and responded to in real time.

[0066] It should also be noted that the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, commodity or device including the element.

[0067] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and reference can be made to the relevant parts of the method embodiment for the relevant content.

[0068] The above description is only for the embodiments of this specification and is not used to limit one or more embodiments of this specification. For those skilled in the art, one or more embodiments of this specification can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included within the scope of the claims of one or more embodiments of this specification. One or more embodiments of this specification, one or more embodiments of this specification, one or more embodiments of this specification, one or more embodiments of this specification.

Claims

1. A method for fault warning and fault location of low-voltage distribution network based on edge computing, characterized in that, it includes the following steps: The monitoring terminal collects the status information data of the low-voltage distribution network and sends the status information data packet to the corresponding distributed edge computing terminal; The distributed edge computing terminal uses a fault prediction model based on neural network to perform fault determination and fault location according to the status information data packet, and sends the fault determination result and the fault location result to the monitoring center; The monitoring center generates a fault warning message according to the fault determination result and the fault location result and pushes it; The construction of the fault prediction model based on neural network includes the following steps: Taking the status information data packet as the input layer and the low-voltage distribution network fault determination result and the fault location result as the output layer, establishing a feedforward neural network fault prediction model, and initializing the fault prediction model; Using a sparse multiple linear regression model to sort the nodes in the hidden layer of the fault prediction model according to the contribution degree of the nodes to the prediction result; Cutting the sorted hidden layer nodes starting from the nodes with low contribution degree and calculating the model prediction error; Supplementing and adding hidden layer nodes according to the prediction error to obtain a new fault prediction model.

2. The method for fault warning and fault location of low-voltage distribution network based on edge computing according to claim 1, characterized in that: The status information data packet sent by the monitoring terminal contains identification information for marking the data packet.

3. The method for fault warning and fault location of low-voltage distribution network based on edge computing according to claim 2, characterized in that: The status information data in the status information data packet includes any one or more of loop current data, voltage data, residual current data, and temperature data.

4. The method for fault warning and fault location of low-voltage distribution network based on edge computing according to claim 3, characterized in that: The input nodes of the input layer of the fault prediction model based on neural network include any one or more of loop current data, voltage data, residual current data, and temperature data, and the identification information.

5. The method for fault warning and fault location of low-voltage distribution network based on edge computing according to claim 1, characterized in that: The distributed edge computing terminal is deployed in a distributed manner in several distribution network grid units of the low-voltage distribution network, and each distributed edge computing terminal corresponds to and controls several monitoring terminals.

6. The method for fault warning and fault location of low-voltage distribution network based on edge computing according to claim 1, characterized in that: It also includes receiving the node operation status data reported by the distributed edge computing terminal through the monitoring center, judging whether the operation status of the distributed edge computing terminal is normal according to the node operation status data, and if it is not normal, alarming the distributed edge computing terminal.

7. An electronic device, characterized in that, it includes: A memory, on which program code is stored; A processor, which is connected to the memory, and when the program code is executed by the processor, the method described in any one of claims 1 to 6 is implemented.

8. A computer-readable storage medium, characterized in that, program instructions are stored thereon, and when the program instructions are executed, the method described in any one of claims 1 to 6 is implemented.

9. A low-voltage distribution network fault warning and fault location system based on edge computing, which implements the method described in any one of claims 1 to 6, characterized in that: it includes a number of monitoring terminals, a number of distributed edge computing terminals, and a monitoring center; the low-voltage distribution network is divided into a number of distribution network grid units, and the distributed edge computing terminals are deployed in a distributed manner in the corresponding distribution network grid units, and each of the distributed edge computing terminals controls a number of monitoring terminals; the monitoring terminals collect the status information data of the low-voltage distribution network and send the status information data packets to the corresponding distributed edge computing terminals; the distributed edge computing terminals use a fault prediction model based on a neural network to perform fault determination and fault location according to the status information data packets, and send the fault determination results and fault location results to the monitoring center; the monitoring center generates fault warning information according to the fault determination results and the fault location results and pushes it.

Images