Two-way Risk Early Warning System and Method for the Operation of the Backbone Optical Communication System of the Large Power Grid

Abstract

The present invention belongs to the technical field of power automation, and discloses a two-way risk early warning system and method for the operation of a large power grid backbone optical communication system. The system includes: a data acquisition layer for collecting device data and actual environmental information of the large power grid network in the physical entity layer and uploading them to the digital twin layer; a digital twin layer for constructing a power-communication two-way coupled network twin by using the device data and actual environmental information of the large power grid network uploaded by the data acquisition layer; and an application layer for using the power-communication two-way coupled network twin, the device data of the large power grid network, and the actual environmental information to conduct two-way risk early warning for the operation of the large power grid backbone optical communication system. The present invention discovers potential faults in a timely manner and issues early warnings by analyzing the real-time operation data of the large power grid and its supporting backbone optical communication system.

Description

Technical Field

[0001] The present invention belongs to the technical field of power automation, and particularly relates to a two-way risk early warning system and method for the operation of a backbone optical communication system of a large power grid. Background Art

[0002] Under the background of the new power system, with the large-scale development and high-proportion grid connection of new energy, and it gradually becomes the main body of power supply, the power balance, security and stability control of the system will face unprecedented challenges. It has become an urgent need to build a large power grid with more powerful functions, more flexible operation and greater resilience. With the geometric growth of terminals connected to the large power grid, the demand for power communication is also increasing day by day. The large power grid construction puts forward higher requirements for the bearing capacity of the backbone optical communication system. The backbone optical communication system supporting the large power grid is of great significance for its safe and stable operation. However, due to the mutual coupling and correlation between the two, the large power grid supplies power to the backbone optical communication system. Once a grid fault leads to a corresponding communication fault, and the communication bears services such as relay protection, the communication fault will exacerbate the grid fault. Moreover, the geographical location has a wide span and the natural geographical environment is complex, facing various risk hazards. Once a fault occurs, it will lead to the risk of the collapse of the entire large power grid, causing serious economic losses. Therefore, it is necessary to conduct risk early warning analysis on the backbone optical communication system of the large power grid, avoid risks in time, improve the reliability of the two-way operation of the backbone optical communication system of the large power grid, and ensure its operation safety.

[0003] The risk early warning analysis of the traditional backbone optical communication system of the large power grid mainly relies on manual inspection. The method of manual inspection has the disadvantages of high cost, great difficulty, insufficient ability to master and control the information of the large power grid and its supporting backbone optical communication system with large span and complex geographical environment, and slow response. In addition, for the risk monitoring of the backbone optical communication system, the Optical Time Domain Reflectometer (OTDR) technology is mostly used for prediction, which can detect faults in optical fiber communication. In the backbone optical communication system of the large power grid, the optical cable is set up overhead and is affected by more external disturbances. The OTDR technology cannot effectively sense the external events of the optical cable and cannot monitor the external abnormalities of the optical cable. When a communication fault is detected, damage has often occurred, and it is impossible to achieve efficient early warning in the power communication network. Summary of the Invention

[0004] The purpose of the present invention is to provide a two-way risk early warning system and method for the operation of a backbone optical communication system of a large power grid, so as to solve the technical problem that the prior art cannot perform efficient early warning.

[0005] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0006] In a first aspect, the present invention provides a two-way risk early warning system for the operation of a large power grid backbone optical communication system, including:

[0007] A data acquisition layer, configured to collect device data of the large power grid network and actual environment information within the physical entity layer, and upload them to the digital twin layer;

[0008] A digital twin layer, configured to construct a power-communication two-way coupling network twin by using the device data of the large power grid network and the actual environment information uploaded by the data acquisition layer;

[0009] An application layer, configured to perform two-way risk early warning on the operation of the large power grid backbone optical communication system by using the power-communication two-way coupling network twin, the device data of the large power grid network, and the actual environment information.

[0010] A further improvement of the present invention is that: the device data of the large power grid network includes: synchronous digital hierarchy network management, optical fiber transmission network network management, production management system, energy management system, and power communication management system data; the actual environment information includes meteorological monitoring system and hydrological monitoring system information.

[0011] A further improvement of the present invention is that: the power-communication two-way coupling network twin is composed of nodes and branches; the nodes include power nodes, backbone optical communication system communication nodes, and control nodes; the branches include power branches, communication branches, and coupling branches. Under the action of the coupling branches, the power nodes and the backbone optical communication system communication nodes interact with each other to jointly form a coupling body.

[0012] A further improvement of the present invention is that: performing two-way risk early warning on the operation of the large power grid backbone optical communication system by using the power-communication two-way coupling network twin, the device data of the large power grid network, and the actual environment information specifically includes:

[0013] Randomly select a power grid fault event twin Bayesian network BN within the twin layer a , and determine whether there is a power grid fault event twin Bayesian network that needs to be merged currently. If there is still an unmerged power grid fault event twin Bayesian network BN b , then perform the merger of the power grid fault event twin Bayesian network until there is no unmerged power grid fault event twin Bayesian network, and obtain the finally merged concurrent fault event chain network of the large power grid backbone optical communication system;

[0014] According to the finally merged concurrent fault event chain network of the large power grid backbone optical communication system within the twin layer, input the initial fault cause twin data for monitoring the backbone optical communication operation system, and obtain the risk probability F of different types of faults occurring a , F ais a real number between [0, 1], and the risk probability F of different types of faults occurring a is compared with the preset threshold of fault risk to conduct two-way risk early warning for the operation of the backbone optical communication system of the large power grid.

[0015] A further improvement of the present invention lies in: The merging of the twin Bayesian network of power grid fault events includes: judging the interaction relationship between the twin Bayesian network BN b and BN a If there is a causal relationship between BN b and BN a then conduct causal association for BN b and BN a If there is a coupling relationship between BN b and BN a then conduct coupling association for BN b and BN a After association, the network BN is obtained.

[0016] A further improvement of the present invention lies in: The preset threshold of fault risk is a dynamic value, and the adjustment method of the preset threshold of fault risk specifically includes:

[0017] Based on historical operation and maintenance data, for a certain fault event O a , the risk probability output by the concurrent fault event chain network of the backbone optical communication system of the large power grid in the digital twin layer for the zth time is F a,z , and the self-similarity variance of the continuous H outputs of the concurrent fault event chain network is defined as:

[0018]

[0019] where F a,h is the risk probability output by the concurrent fault event chain network of the backbone optical communication system of the large power grid for the hth time;

[0020] If after maintenance, it is determined that the output result of this time is a misjudgment, then increase the preset threshold of the fault risk of the fault event O a at the next early warning moment The adjustment formula of the preset threshold of the fault risk is as follows:

[0021]

[0022] W represents the importance of the communication equipment at the node associated with the fault event O a .

[0023] A further improvement of the present invention lies in: The calculation method of the importance of communication equipment specifically includes:

[0024] Using the twin data of delay, bandwidth, reliability, bit error rate, service type, service quantity, degree centrality, betweenness centrality, site level, site scale, load level, and load size stored in the digital twin layer, input into the pre-established service importance evaluation model, topology importance evaluation model, and power grid node importance evaluation model in the digital twin layer to obtain service importance evaluation indicators, topology importance evaluation indicators, and power grid node importance evaluation indicators. Input into the pre-established three artificial neural networks, select the sigmoid function as the transfer function of the hidden layer of the three artificial neural networks, calculate the output results in sequence according to the initialized weights and thresholds of each neuron network, use the output results of the three artificial neural networks as the input of the pre-established comprehensive equipment importance neural network, and output the communication equipment importance W at the node associated with the fault event O according to the network weights and thresholds of the comprehensive equipment importance neural network. a associated node.

[0025] In a second aspect, the present invention provides a method for bidirectional risk early warning of the operation of a large power grid backbone optical communication system, including:

[0026] Collecting equipment data and actual environment information of the large power grid network;

[0027] Using the pre-established power-communication bidirectional coupled network twin, the equipment data of the large power grid network, and the actual environment information to conduct bidirectional risk early warning on the operation of the large power grid backbone optical communication system;

[0028] The pre-established power-communication bidirectional coupled network twin is constructed through the historical equipment data and historical actual environment information of the large power grid network.

[0029] In a third aspect, the present invention provides an electronic device, including a processor and a memory, where the processor is used to execute a computer program stored in the memory to implement the method for bidirectional risk early warning of the operation of the large power grid backbone optical communication system.

[0030] In a fourth aspect, the present invention provides a computer-readable storage medium, where the computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by a processor, the method for bidirectional risk early warning of the operation of the large power grid backbone optical communication system is implemented.

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] The present invention provides a two-way risk early warning system and method for the operation of a large power grid backbone optical communication system. By carrying out data interface interconnection with SDH, OTN professional network management, integrated network management (TMS system), regulation cloud, and equipment management system (PMS system), historical operation and maintenance alarm data of the large power grid and the supporting backbone optical communication system are widely collected, a two-way interactive digital twin architecture of the large power grid backbone optical communication system is constructed, and the service importance, topological importance, and power grid impact importance of communication equipment are comprehensively considered. A multi-channel artificial neural network is used to analyze the impact degree of communication equipment; based on the obtained risk impact degree of communication equipment, combined with historical maintenance data, and based on Bayesian learning, the causality and relevance of different fault cause events to the occurrence of faults are analyzed, a concurrent fault event chain network is constructed, and through the real-time operation data analysis of the large power grid and its supporting backbone optical communication system, potential faults are timely discovered (when exceeding a certain threshold, it is considered to have a fault risk, and the more important the equipment, the lower the threshold requirement), and early warnings are issued.

[0033] Furthermore, based on the analysis of the risk early warning results and the operation and maintenance results, the present invention obtains the risk early warning accuracy result of the current concurrent fault event chain network, and then adaptively updates the risk assessment threshold according to the result to improve the early warning accuracy.

[0034] Furthermore, the present invention uses the twin data stored in the digital twin layer combined with the service, topology, and power grid node importance evaluation models in the twin layer to obtain the corresponding importance evaluation indicators as the input of each channel artificial neural network. Subsequently, the output results of the multi-channel artificial neural network are input into the comprehensive equipment importance neural network, and the comprehensive equipment importance result is output with comprehensive weights and thresholds.

[0035] Furthermore, based on the power grid-communication network coupled digital twin in the digital twin layer, the present invention obtains the fault event twin data as the nodes in the Bayesian network to construct a power grid concurrent operation fault event chain twin network, obtains the risk probabilities of different fault events, and issues risk early warnings in a timely manner by comparing with the preset threshold to guide operation and maintenance.

[0036] Furthermore, the present invention proposes an adaptive adjustment method for the preset threshold of risk determination. According to the fault risk early warning situation, combined with the actual operation and maintenance results of the power grid, when misjudgment is found, combined with the importance of the communication equipment of the fault risk early warning node, the preset threshold of the fault risk determination is adaptively adjusted, so as to further effectively improve the risk early warning accuracy, reduce the misjudgment phenomenon, and reduce the power grid operation and maintenance cost. Description of the Drawings

[0037] The specification drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0038] Figure 1 It is a schematic flow chart of a two-way risk early warning method for the operation of a backbone optical communication system of a large power grid according to the present invention;

[0039] Figure 2 It is a structural block diagram of a two-way risk early warning system for the operation of a backbone optical communication system of a large power grid according to the present invention;

[0040] Figure 3 It is a schematic flow chart of the importance evaluation of communication equipment based on digital twin-assisted multi-channel artificial neural network;

[0041] Figure 4 It is a schematic diagram of a grid concurrent operation fault event chain;

[0042] Figure 5 It is a schematic diagram of a risk early warning method based on digital twin-assisted Bayesian network;

[0043] Figure 6 It is a schematic flow chart of another two-way risk early warning method for the operation of a backbone optical communication system of a large power grid according to the present invention;

[0044] Figure 7 It is a structural block diagram of an electronic device according to the present invention. Detailed implementation manners

[0045] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0046] The following detailed descriptions are all exemplary descriptions, aiming to provide further detailed descriptions of the present invention. Unless otherwise specified, all technical terms adopted by the present invention have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs. The terms used in the present invention are only for describing specific implementation manners, and are not intended to limit the exemplary embodiments according to the present invention.

[0047] Glossary:

[0048] Large power grid: A power grid with a large total capacity, a large power supply coverage area, a large number and diverse forms of power plants feeding power into the grid, a high highest voltage level of the grid itself and many voltage levels, both AC and DC, and a large number of users with complex demands is a large power grid.

[0049] Backbone optical communication system: The backbone optical communication system is a high-speed network used to connect multiple regions or areas in the power communication network. The backbone optical communication system covers the communication between power grid substations above 35 kV and various production and office places, and is divided into a transmission network, a service network and a support network according to functions.

[0050] Digital Twin: A digital twin is a simulation process that makes full use of data such as physical models, sensor updates, and operation history, integrates multiple disciplines, multiple physical quantities, multiple scales, and multiple probabilities, and completes a mapping in the virtual space to reflect the entire life cycle process of the corresponding entity in real time.

[0051] Embodiment 1

[0052] Please refer to Figure 1 As shown in the figure, the present invention provides a two-way risk warning method for the operation of a large power grid backbone optical communication system, including: firstly, constructing a two-way interactive digital twin architecture for the large power grid backbone optical communication system; secondly, evaluating the importance of communication devices based on the twin data and multi-channel artificial neural networks in the digital twin layer; then, using the twin data and Bayesian network to form a device operation fault event chain, setting an initial risk warning threshold through the obtained comprehensive device importance for risk warning; finally, considering the important role of the selection of the discrimination threshold of the fault risk in the warning accuracy, proposing an adaptive threshold adjustment feedback mechanism combining device importance in the mechanism model of the digital twin layer to further improve the risk warning accuracy.

[0053] (1) Analysis of the influence of digital twin and convolutional neural network on the importance of communication devices

[0054] 1) Two-way interactive digital twin architecture of the large power grid backbone optical communication system

[0055] Please refer to Figure 2 As shown in the figure, the two-way interactive digital twin architecture of the large power grid backbone optical communication system proposed by the present invention includes four layers: an application layer, a digital twin layer, a data acquisition layer, and a physical entity layer, which are specifically introduced as follows:

[0056] a. Physical entity layer: The physical entity layer is the actual network of the large power grid, which covers all links of power generation, transmission, distribution, transformation, and utilization of the large power grid and the power backbone optical communication system for large power grid communication. The actual network of the large power grid in the physical entity layer integrates power, communication, sensing, computing, and control technologies. This complexity and heterogeneity make its operation and control extremely complex. Therefore, risk warning is required based on the digital twin of the large power grid backbone optical communication system constructed by a multi-service system.

[0057] b. Data Acquisition Layer: The data acquisition layer is used to collect data of numerous devices and actual environmental information within the physical entity layer. It includes seven parts: Synchronous Digital Hierarchy (SDH) network management, Optical Transport Network (OTN) network management, Power Production Management System (PMS), Energy Management System (EMS), Telecommunication Management System (TMS), meteorological monitoring system, and hydrological monitoring system. Through the data acquisition layer, numerous information is obtained from the physical entity layer and uploaded to the digital twin layer in real time, realizing the real-time iterative update of the power-communication bidirectional coupling network twin in the digital twin layer. Digital twin technology is data-driven. As an application platform, digital twin can obtain operation data and environmental data of power grid and communication network devices from multiple systems, construct a digital model that can reflect the operation status of the entity network through data analysis and processing, and be applied in the power grid; the twin integrates all data, providing data support and a calculation platform for subsequent risk early warning.

[0058] c. Digital Twin Layer: In the digital twin layer, a power-communication bidirectional coupling network twin is constructed through information of numerous devices in the physical entity layer uploaded by the data acquisition layer. Among them, the power-communication bidirectional coupling network twin mainly consists of nodes and branches. Nodes include power nodes, communication nodes of the backbone optical communication system, and control nodes; branches include power branches, communication branches, and coupling branches. Under the action of the coupling branches, power nodes and communication nodes of the backbone optical communication system interact with each other to jointly form a coupling body. Based on the constructed digital twin layer, instructions can be issued to personnel, equipment, etc. through the output results of the twin layer to optimize the operation of the current physical entity layer, realizing the importance analysis and risk early warning of large power grid communication equipment.

[0059] d. Application layer: The application layer includes a device management module, a risk warning module, and an operation and maintenance module. The device management module is used to create files and models for the device information and graphics of communication devices, establish a communication device ledger, and realize the full-life cycle management of the large power grid optical fiber backbone communication system. Based on the twin data and the power-communication bidirectional coupling network twin body, the operation state of the communication system is monitored and controlled in real time, providing strong support for risk warning, risk diagnosis, predictive maintenance decision-making, etc. Based on the real-time information interaction of the device status between the digital twin layer and the physical entity layer, online device monitoring can be realized. According to the device operation state, maintenance tasks are dynamically arranged. Devices with more problems are preferentially maintained, and those with fewer problems are maintained later, improving the flexibility of operation and maintenance.

[0060] 2) Communication device importance analysis method based on digital twin-assisted multi-channel artificial neural network

[0061] The importance of communication devices is mainly related to three factors, namely the importance of the services carried by the devices, the importance of the devices in the network topology, and the importance of the power grid nodes coupled with the communication devices. The service importance is determined by the security partition of the service in the power grid and the performance requirements of the service for delay, bandwidth, reliability, and bit error rate. The higher the security partition of the service and the higher the performance requirements, the higher the service importance. The topology importance is determined by the topological position and connection of the nodes in the network topology. The commonly used indicators in evaluating the topology importance are the degree centrality and betweenness centrality of the topological nodes. The greater the degree centrality and betweenness centrality of the topological nodes, the higher the topology importance. The importance of power grid nodes is described by power grid sites. Different site levels and scales have different importance in the network. The higher the site voltage level and the greater the load level, the higher the importance of the power grid nodes.

[0062] Since different indicators have different impacts on the device importance, it is necessary to realize the weight analysis of different indicators to quantify the device importance. Using the digital twin-assisted multi-channel artificial neural network to realize the discrimination of the influence weights of different indicators on the device importance, the advantage of separate training is to evaluate in different fields respectively, reduce the interference of non-professional field indicators, and can reduce the data training dimension, enhance the iteration convergence speed of the artificial neural network, and obtain the final comprehensive device importance. The specific steps are as follows:

[0063] Using the twin data of delay, bandwidth, reliability, bit error rate, service type, service quantity, degree centrality, betweenness centrality, site level, site scale, load level, and load size stored in the digital twin layer (the twin body reflects the structure and operating conditions of the physical entity in real time, and the twin data is the data input into the twin body and extracted from the twin body when used), combined with the service, topology, and power grid node importance evaluation models in the twin layer to obtain the corresponding importance evaluation indicators as the input of the artificial neural network for each channel, selecting the sigmoid function as the transfer function of the hidden layer, calculating the output successively according to the weights and thresholds of each initialized neuron network, taking the results output by the output layers of the three channels as the input of the comprehensive equipment importance neural network, and outputting the comprehensive equipment importance result according to the network weights and thresholds, comparing the result output by the comprehensive equipment importance neural network with the expected result, calculating the error and putting it into the comprehensive equipment importance neural network for backpropagation until the error is within the limit value, where the error is obtained according to the mean square error formula.

[0064] In a specific embodiment, the service importance is related to factors such as delay, bandwidth, reliability, bit error rate, service type, and service quantity; the topology importance is related to degree centrality and betweenness centrality; the power grid node importance is related to site level, site scale, load level, and load size.

[0065] In a specific embodiment, as Figure 3 shown, the comprehensive equipment importance neural network adopts the general architecture of the artificial neural network, including an input layer, a hidden layer, and an output layer.

[0066] As Figure 3 shown, calculating the error and putting it into the comprehensive equipment importance neural network for backpropagation until the error is within the limit value specifically includes: taking Channel 1 evaluated by service importance as an example, there are I inputs in the artificial neural network in Channel 1, there are M neurons in the hidden layer of Channel 1, and the actual output of the neural network in Channel 1 is y 1 , and its error calculation formula is where is the expected output result. If the error is greater than the preset error range, then adjust the connection weights between the hidden layer and the output layer, the connection weights between the input layer and the hidden layer, the output layer threshold, and the hidden layer threshold. The generalized error of the output layer of Channel 1 is The generalized error of each unit in the hidden layer of Channel 1 is is the weight of the current neuron in the hidden layer of Channel 1, is the output of the corresponding neuron in the hidden layer of Channel 1. Adjust the connection weight between the hidden layer of Channel 1 and the output layer of Channel 1 and the output layer threshold β 1 is expressed as:

[0067]

[0068] where η 1 is the learning rate of the neural network.

[0069] Further adjust the connection weights between the input layer and the hidden layer of Channel 1

[0070]

[0071] where r i 1 represents the output of the i-th input layer neuron.

[0072] When the errors of each channel within the digital twin layer and the overall neural network converge within the expected values, it indicates that the comprehensive evaluation network training of the equipment importance is completed, and the output result is the comprehensive importance of the corresponding equipment.

[0073] (2) Risk warning method based on digital twin-assisted Bayesian network

[0074] Figure 4 represents the concurrent operation fault event chain of the power grid. Among them, C represents the set of input attributes of the power grid fault event, that is, the set of original fault inducing events, which are twin data such as continuous precipitation, continuous high temperature, continuous snowfall, continuous exposure to the sun, continuous strong wind, short circuit, lightning strike, etc. obtained from the digital twin layer; S represents the set of state attributes of the power grid fault event, that is, the set composed of the derivative fault events Ev, such as transformer water ingress, cable trench water accumulation, etc.; O is the set of output attributes of the power grid fault, that is, the set of fault events that ultimately affect the operation of the power grid, such as cable leakage, terminal fault, etc. Finally, the set of output attributes of the power grid fault is further input into the risk warning, risk diagnosis, and guiding predictive maintenance decision-making in the power grid operation risk warning platform.

[0075] Based on the above analysis, in the mechanism model of the digital twin layer of the present invention, a Bayesian chain network of concurrent fault events is constructed. The twin data of the input attributes, state attributes, and output attributes of the fault events are obtained from the power grid-communication network coupled digital twin in the digital twin layer as the nodes in the Bayesian network, and the interaction relationships between the attributes are used as the directed edges in the network to construct a Bayesian network of fault events based on digital twins, which is represented as BN=(V, R, P). Among them, V={v|v∈C∪S∪O} is the variable set composed of the attributes related to the power grid fault events, which is composed of the set of fault cause events C, the set of state variables S, and the output set O. R={r(v, v')|v, v'∈V} is the set of directed edges composed of the interaction relationships between the variables in the network; P={p(v|v')|v, v'∈V} is the set of uncertain information about the interaction relationships between the variables in the network; the nodes in the Bayesian network represent events, and the edges connecting the nodes represent the association relationships between the events. The edge connecting event v and v’ is represented as r, and all r constitute the set R.

[0076] The present invention comprehensively considers the loop probability convergence property of the Bayesian network with loops and the association form between the Bayesian networks of fault events, and merges and constructs a Bayesian chain network of concurrent fault events. The method flow for merging and constructing the Bayesian chain network of concurrent fault events in the backbone optical communication system of the large power grid based on digital twins is as Figure 5 shown, and the specific steps are as follows:

[0077] Step 1: The initial event chain obtained based on historical experience analysis may still have coupling or causal relationships and needs to be further merged. The present invention randomly selects a Bayesian network BN of the twin of a power grid fault event in the digital twin layer a , and determines whether there is a Bayesian network of the twin of a power grid fault event that needs to be merged currently. If there is still an unmerged network BN b , then go to Step 2 to merge the Bayesian networks of the twins of the power grid fault events; otherwise, obtain the finally merged Bayesian chain network of concurrent fault events in the backbone optical communication system of the large power grid and go to Step 6.

[0078] Step 2: Analyze the causal relevance between the Bayesian networks BN b and BN a ; for example, in a specific embodiment, a qualitative judgment is made. Continuous precipitation will cause waterlogging, and waterlogging may cause transformer leakage, which is a causal relevance event chain.

[0079] Step 3: Judge the interaction relationship between the Bayesian networks BN b and BN a . If there is a causal relationship between BN b and BN a , then for BN band BN a carry out causal association; if BN b and BN a have a causal coupling relationship, then perform coupling association on BN b and BN a ; after association, obtain the network BN;

[0080] Step 4: For the network BN obtained through association, determine the conditional probability information between the network variable nodes by statistically analyzing the acquired historical data. If there is a directed cycle in the network, use the limiting probability as the prior probability information of the variable nodes on the cycle during probability inference to ensure the feasibility of probability inference. The finally constructed network should be a one-way propagation from the fault cause to the fault event. In case of a cyclic network, there will be an infinite loop in the cycle, and the deduction of the fault event cannot be completed, and the final risk probability cannot be obtained. Step 4 can effectively avoid this phenomenon.

[0081] Step 5: Determine whether there is a twin Bayesian network of power grid fault events that needs to be merged currently. If so, go to Step 2 to merge the twin Bayesian network of power grid fault events; otherwise, obtain the finally merged concurrent fault event chain network of the backbone optical communication system of the large power grid and go to Step 6.

[0082] Step 6: Based on the concurrent fault event chain network of the backbone optical communication system of the large power grid finally generated within the twin layer, input the initial fault cause twin data for monitoring the backbone optical communication operation system, and obtain the risk probability F a , F a is a real number between [0, 1]. The closer this value is to 1, the greater the probability of the occurrence of the fault, and the higher the operation risk of the backbone optical communication system of the large power grid; the closer this value is to 0, the lower the risk. Compare the fault risk probability with the preset threshold for comparison, The initial value of is determined by the comprehensive equipment importance W of the corresponding faulty equipment. If the fault occurrence probability exceeds the threshold, upload the corresponding fault information to the application layer for risk warning, and adjust the threshold based on the following adaptive adjustment method of the preset risk determination threshold.

[0083] (3) Adaptive adjustment method of the preset risk determination threshold

[0084] In the risk warning of the backbone optical communication system of the large power grid, the discrimination threshold The selection of has a crucial role in the accuracy of early warning. Whether this value is appropriately selected will directly affect the final result. If there are continuous misjudgments, it will lead to an increase in the grid operation and maintenance costs. Therefore, based on the actual maintenance results after risk early warning, an adaptive threshold adjustment feedback method combining equipment importance is designed in the mechanism model of the digital twin layer to further improve the accuracy of risk early warning.

[0085] Based on historical operation and maintenance data, for a certain fault event O a , the risk probability of the z-th output of the concurrent fault event chain network of the large power grid backbone optical communication system in the twin layer is F a,z , and the self-similarity variance of the continuous H outputs of the concurrent fault event chain network is defined as:

[0086]

[0087] where F a,h is the risk probability of the h-th output of the concurrent fault event chain network of the large power grid backbone optical communication system; if after the inspection by grid-related staff, it is determined that the output result of this time is a misjudgment, it means that the self-similarity variance of the current event chain network is too small, resulting in insufficient response of risk early warning to sudden events. The preset threshold of this fault risk should be increased at the next early warning moment to reduce the misjudgment probability. The specific adjustment formula is as follows:

[0088]

[0089] W represents the importance of the communication equipment at the node associated with the fault event O a . If the importance of the associated equipment is higher, the increased threshold range is smaller. The greater the importance, the lower the tolerance for the occurrence of faults. Appropriately reducing the threshold increases the probability of early warning once there is a risk, so as to carry out operation and maintenance in a timely manner and reduce potential losses.

[0090] The two-way interaction architecture of the large power grid backbone optical communication system based on digital twin proposed by the present invention uses the multi-dimensional index twin data stored in the twin layer and the importance evaluation model to conduct importance evaluation, and obtains a more objective comprehensive equipment importance evaluation result based on the comprehensive weights and thresholds of the digital twin-assisted multi-channel artificial convolutional neural network.

[0091] The present invention constructs a twin network of grid concurrent operation fault events based on the Bayesian network mechanism model and the fault event attribute twin data in the power grid-communication network coupled digital twin body, and issues early warning risks for possible faults of the large power grid backbone optical communication system in a timely manner to guide operation and maintenance.

[0092] According to the risk early warning and the results of operation and maintenance inspection, the present invention proposes an adaptive threshold adjustment feedback mechanism combining equipment importance in the mechanism model of the digital twin layer. By combining the importance of risk early warning node communication equipment evaluated in the twin layer, the preset threshold for fault risk determination is adaptively adjusted to improve the accuracy of risk early warning.

[0093] Embodiment 2

[0094] Please refer to Figure 2 As shown, the present invention provides a two-way risk early warning system for the operation of a large power grid backbone optical communication system, including:

[0095] A data acquisition layer for collecting equipment data and actual environment information of the large power grid network in the physical entity layer and uploading them to the digital twin layer;

[0096] A digital twin layer for constructing a power-communication two-way coupled network twin by using the equipment data and actual environment information of the large power grid network uploaded by the data acquisition layer;

[0097] An application layer for performing two-way risk early warning on the operation of the large power grid backbone optical communication system by using the power-communication two-way coupled network twin, the equipment data of the large power grid network, and the actual environment information.

[0098] In a specific embodiment: the equipment data of the large power grid network includes: synchronous digital hierarchy network management, optical fiber transmission network network management, production management system, energy management system, and power communication management system data; the actual environment information includes meteorological monitoring system and hydrological monitoring system information.

[0099] In a specific embodiment: the power-communication two-way coupled network twin is composed of nodes and branches; the nodes include power nodes, backbone optical communication system communication nodes, and control nodes; the branches include power branches, communication branches, and coupling branches. Under the action of the coupling branches, the power nodes and the backbone optical communication system communication nodes interact with each other to jointly form a coupled body.

[0100] In a specific embodiment: performing two-way risk early warning on the operation of the large power grid backbone optical communication system by using the power-communication two-way coupled network twin, the equipment data of the large power grid network, and the actual environment information specifically includes:

[0101] Randomly select a power grid fault event twin Bayesian network BN in the twin layer a , and determine whether there is a power grid fault event twin Bayesian network that needs to be merged currently. If there is still an unmerged power grid fault event twin Bayesian network BN b, the merging of the power grid fault event twin Bayesian networks is carried out until there are no unmerged power grid fault event twin Bayesian networks, and the final merged concurrent fault event chain network of the large power grid backbone optical communication system is obtained;

[0102] According to the concurrent fault event chain network of the large power grid backbone optical communication system finally merged within the twin layer, input the initial fault cause twin data for the monitoring of the backbone optical communication operation system, and obtain the risk probability F of different types of faults occurring a , F a is a real number between [0,1]. The risk probability F of different types of faults occurring a is compared with the preset fault risk threshold to conduct two-way risk early warning for the operation of the large power grid backbone optical communication system.

[0103] In a specific embodiment: the merging of the power grid fault event twin Bayesian networks includes: judging the interaction relationship between the power grid fault event twin Bayesian networks BN b and BN a . If there is a causal relationship between BN b and BN a , then conduct causal association for BN b and BN a . If there is a coupling relationship between BN b and BN a , then conduct coupling association for BN b and BN a . After association, the network BN is obtained.

[0104] In a specific embodiment: the preset fault risk threshold is a dynamic value. The adjustment method of the preset fault risk threshold specifically includes:

[0105] According to the historical operation and maintenance data, for a certain fault event O a , the risk probability output by the concurrent fault event chain network of the large power grid backbone optical communication system in the digital twin layer for the z-th time is F a,z . Define the self-similarity variance of the consecutive H outputs of the concurrent fault event chain network as:

[0106]

[0107] where F a,h is the risk probability output by the concurrent fault event chain network of the large power grid backbone optical communication system for the h-th time;

[0108] If it is determined that the output result of this time is a misjudgment after maintenance, increase the preset fault risk threshold of the fault event O a at the next early warning moment Preset threshold for failure risk The adjustment formula is as follows:

[0109]

[0110] W represents the importance of the communication device at the node associated with the failure event O a associated with the node.

[0111] In a specific embodiment: The calculation method of the importance of the communication device specifically includes:

[0112] Using the twin data of delay, bandwidth, reliability, bit error rate, service type, service quantity, degree centrality, betweenness centrality, site level, site scale, load level, and load size stored in the digital twin layer, inputting into the service importance evaluation model, topology importance evaluation model, and power grid node importance evaluation model pre-established in the digital twin layer to obtain service importance evaluation indicators, topology importance evaluation indicators, and power grid node importance evaluation indicators, inputting into three pre-established artificial neural networks, selecting the sigmoid function as the transfer function of the hidden layer of the three artificial neural networks, calculating the output results in sequence according to the initialized weights and thresholds of each neuron network, taking the output results of the three artificial neural networks as the input of the pre-established comprehensive equipment importance neural network, and outputting the importance W of the communication device at the node associated with the failure event O according to the network weights and thresholds of the comprehensive equipment importance neural network a associated with the node.

[0113] Embodiment 3

[0114] Please refer to Figure 6 as shown, the present invention provides a two-way risk early warning method for the operation of a large power grid backbone optical communication system, including:

[0115] S1. Collect the device data and actual environment information of the large power grid network;

[0116] S2. Use the pre-established power-communication two-way coupled network twin, the device data of the large power grid network, and the actual environment information to conduct two-way risk early warning on the operation of the large power grid backbone optical communication system;

[0117] Among them, the pre-established power-communication two-way coupled network twin is constructed by using the historical device data and historical actual environment information of the large power grid network.

[0118] In a specific embodiment: The device data of the large power grid network includes: synchronous digital hierarchy network management, optical fiber transmission network management, production management system, energy management system, and power communication management system data; the actual environment information includes meteorological monitoring system and hydrological monitoring system information.

[0119] In a specific embodiment: the power-communication bidirectional coupling network twin is composed of nodes and branches; the nodes include power nodes, backbone optical communication system communication nodes, and control nodes; the branches include power branches, communication branches, and coupling branches. Under the action of the coupling branches, the power nodes and the backbone optical communication system communication nodes interact with each other to jointly form a coupled body.

[0120] In a specific embodiment: using the power-communication bidirectional coupling network twin, as well as the equipment data and actual environment information of the large power grid network, to conduct two-way risk early warning on the operation of the large power grid backbone optical communication system, specifically including:

[0121] Randomly select a power grid fault event twin Bayesian network BN within the twin layer a , and determine whether there is a power grid fault event twin Bayesian network that needs to be merged currently. If there is still an unmerged power grid fault event twin Bayesian network BN b , then perform the merger of the power grid fault event twin Bayesian networks until there is no unmerged power grid fault event twin Bayesian network, and obtain the final merged concurrent fault event chain network of the large power grid backbone optical communication system;

[0122] Based on the final merged concurrent fault event chain network of the large power grid backbone optical communication system within the twin layer, input the initial fault cause twin data for monitoring the backbone optical communication operation system, and obtain the risk probability F of different types of faults occurring a , F a is a real number between [0, 1]. Compare the risk probability F of different types of faults occurring a with the preset fault risk threshold to conduct two-way risk early warning on the operation of the large power grid backbone optical communication system.

[0123] In a specific embodiment: the merger of the power grid fault event twin Bayesian networks includes: determining the interaction relationship between the power grid fault event twin Bayesian networks BN b and BN a . If there is a causal relationship between BN b and BN a , then conduct causal association on BN b and BN a ; if there is a coupling relationship between BN b and BN a , then conduct coupling association on BN b and BN a ; after association, obtain the network BN.

[0124] In a specific embodiment: the preset fault risk threshold is a dynamic value, the preset threshold of failure risk The adjustment method specifically includes:

[0125] Based on historical operation and maintenance data, for a certain failure event O a , the risk probability of the z-th output of the concurrent failure event chain network of the backbone optical communication system of the large power grid in the digital twin layer is F a,z , define the self-similarity variance of the continuous H outputs of the concurrent failure event chain network as:

[0126]

[0127] Among them, F a,h is the risk probability of the h-th output of the concurrent failure event chain network of the backbone optical communication system of the large power grid;

[0128] If after maintenance, it is determined that the output result of this time is a misjudgment, increase the preset threshold of the failure risk of the failure event O a at the next warning moment The preset threshold of the failure risk The adjustment formula of the preset threshold of the failure risk is as follows:

[0129]

[0130] W represents the importance of the communication equipment at the node associated with the failure event O a

[0131] In a specific embodiment: The calculation method of the importance of communication equipment specifically includes:

[0132] Using the twin data of delay, bandwidth, reliability, bit error rate, service type, service quantity, degree centrality, betweenness centrality, site level, site scale, load level, and load size stored in the digital twin layer, input the pre-established service importance evaluation model, topological importance evaluation model, and power grid node importance evaluation model in the digital twin layer, obtain the service importance evaluation index, topological importance evaluation index, and power grid node importance evaluation index, input the pre-established three artificial neural networks, select the sigmoid function as the transfer function of the hidden layer of the three artificial neural networks, calculate the output results in sequence according to the weights and thresholds of each neuron network initialized, use the output results of the three artificial neural networks as the input of the pre-established comprehensive equipment importance neural network, and output the importance W of the communication equipment at the node associated with the failure event O a according to the network weights and thresholds of the comprehensive equipment importance neural network.

[0133] Example 4

[0134] Please refer to Figure 7 ​As shown in the figure, the present invention further provides an electronic device 100 for implementing a two-way risk warning method for the operation of a large power grid backbone optical communication system; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0135] The memory 101 can be used to store the computer program 103. The processor 102 realizes the steps of the two-way risk warning method for the operation of the large power grid backbone optical communication system described in Embodiment 1 or 3 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the electronic device 100 (such as audio data, etc.). In addition, the memory 101 can include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0136] The at least one processor 102 can be a central processing unit (CPU), or can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 can be a microprocessor or the processor 102 can also be any conventional processor, etc. The processor 102 is the control center of the electronic device 100, and connects various parts of the entire electronic device 100 through various interfaces and lines.

[0137] The memory 101 in the electronic device 100 stores multiple instructions to implement a two-way risk warning method for the operation of a large power grid backbone optical communication system. The processor 102 can execute the multiple instructions to implement:

[0138] Collect device data and actual environmental information of the large power grid network;

[0139] Utilize the pre-established power-communication bidirectional coupled network twin, the device data of the large power grid network, and the actual environment information to conduct bidirectional risk early warning for the operation of the large power grid backbone optical communication system;

[0140] The pre-established power-communication bidirectional coupled network twin is constructed through the historical device data and historical actual environment information of the large power grid network.

[0141] Embodiment 5

[0142] If the modules / units integrated in the electronic device 100 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, and read-only memory (ROM, Read-Only Memory).

[0143] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0144] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for implementing in the process Figure 1 each process or multiple processes and / or blocks Figure 1a device for the functions specified in one or more boxes.

[0145] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, and the instruction device implements the functions specified in one Figure 1 process or multiple processes and / or boxes Figure 1 a box or multiple boxes.

[0146] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 process or multiple processes and / or boxes Figure 1 a box or multiple boxes.

[0147] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: it is still possible to modify the specific implementation manners of the present invention or make equivalent replacements. Any modification or equivalent replacement without departing from the spirit and scope of the present invention shall be covered by the protection scope of the claims of the present invention.

Claims

1. Two-way risk early warning system for the operation of the backbone optical communication system of large power grids, characterized in that Including: A data acquisition layer, which is used to collect the device data of the large power grid network in the physical entity layer and the actual environment information, and upload them to the digital twin layer; A digital twin layer, which is used to construct a power-communication two-way coupled network twin body through the device data of the large power grid network and the actual environment information uploaded by the data acquisition layer; The application layer is used to perform two-way risk early warning on the operation of the backbone optical communication system of the large power grid by using the device data and actual environment information of the power-communication two-way coupled network twin and the large power grid network; specifically including: randomly selecting a twin Bayesian network BN of a power grid fault event in the digital twin layer a , determining whether there is a twin Bayesian network of a power grid fault event that needs to be merged currently. If there is still an unmerged twin Bayesian network BN of a power grid fault event b , then merge the twin Bayesian networks of the power grid fault events until there are no unmerged twin Bayesian networks of the power grid fault events, and obtain the final merged concurrent fault event chain network of the backbone optical communication system of the large power grid; according to the final merged concurrent fault event chain network of the backbone optical communication system of the large power grid in the digital twin layer, input the initial fault cause twin data for monitoring the backbone optical communication operation system, and obtain the risk probability F of different types of faults occurring a , F a is a real number between [0,1]. Compare the risk probability F of different types of faults occurring a with the preset threshold of the fault risk to perform two-way risk early warning on the operation of the backbone optical communication system of the large power grid; Preset threshold for failure risk It is a dynamic value, and the adjustment method specifically includes: based on historical operation and maintenance data, for a certain fault event O a , the risk probability of the z-th output of the concurrent fault event chain network of the backbone optical communication system of the large power grid in the digital twin layer is F a,z , define the self-similarity variance of the continuous H outputs of the concurrent fault event chain network as: Among them, F a,h is the risk probability output for the h-th time of the concurrent fault event chain network of the backbone optical communication system of the large power grid; If it is determined that the output result of this time is a misjudgment after maintenance, increase the preset threshold of the failure risk of the fault event O at the next warning time a of the preset threshold of the failure risk Preset threshold of failure risk The adjustment formula is as follows: W represents the importance of the communication device at the node associated with the fault event O a associated.

2. The two-way risk early warning system for the operation of the backbone optical communication system of the large power grid according to claim 1, wherein The device data of the large power grid network includes: synchronous digital hierarchy network management, optical fiber transmission network network management, production management system, energy management system, and power communication management system data; the actual environment information includes meteorological monitoring system and hydrological monitoring system information.

3. The two-way risk early warning system for the operation of the backbone optical communication system of the large power grid according to claim 1, characterized in that, The power-communication two-way coupled network twin body is composed of nodes and branches; the nodes include power nodes, backbone optical communication system communication nodes, and control nodes; the branches include power branches, communication branches, and coupling branches. Under the action of the coupling branches, the power nodes and the backbone optical communication system communication nodes interact with each other to jointly form a coupled body.

4. The two-way risk early warning system for the operation of the backbone optical communication system of the large power grid according to claim 1, wherein The merging of the twin Bayesian network for power grid fault events includes: judging the interaction relationship between the twin Bayesian networks BN b and BN a If there is a causal relationship between BN b and BN a , then perform causal association on BN b and BN a ; if there is a coupling relationship between BN b and BN a , then perform coupling association on BN b and BN a ; after association, the network BN is obtained.

5. The two-way risk early warning system for the operation of the backbone optical communication system of a large power grid according to claim 1, characterized in that, A calculation method for the importance of communication equipment, specifically including: Using the twin data of delay, bandwidth, reliability, bit error rate, service type, service quantity, degree centrality, betweenness centrality, site level, site scale, load level, and load size stored in the digital twin layer, input them into the pre-established service importance evaluation model, topology importance evaluation model, and power grid node importance evaluation model in the digital twin layer to obtain service importance evaluation indicators, topology importance evaluation indicators, and power grid node importance evaluation indicators. Input them into the pre-established three artificial neural networks, select the sigmoid function as the transfer function of the hidden layer of the three artificial neural networks, calculate the output results in sequence according to the initialized weights and thresholds of each neuron network, use the output results of the three artificial neural networks as the input of the pre-established comprehensive equipment importance neural network, and output the communication equipment importance W at the node associated with the fault event O a according to the network weights and thresholds of the comprehensive equipment importance neural network.

6. Two-way risk early warning method for the operation of the backbone optical communication system of large power grids, characterized in that, Including: Collecting the device data of the large power grid network and the actual environment information; Using the pre-established power-communication two-way coupled network twin, the device data of the large power grid network, and the actual environment information, conduct two-way risk early warning for the operation of the large power grid backbone optical communication system; specifically including: randomly selecting a power grid fault event twin Bayesian network BN in the digital twin layer a , determine whether there is a power grid fault event twin Bayesian network that needs to be merged currently. If there is still an unmerged power grid fault event twin Bayesian network BN b , then merge the power grid fault event twin Bayesian networks until there is no unmerged power grid fault event twin Bayesian network, and obtain the final merged concurrent fault event chain network of the large power grid backbone optical communication system; based on the final merged concurrent fault event chain network of the large power grid backbone optical communication system in the digital twin layer, input the initial fault cause twin data for monitoring the backbone optical communication operation system, and obtain the risk probability F of different types of faults occurring a , F a is a real number between [0, 1]. Compare the risk probability F of different types of faults occurring a with the preset fault risk threshold to conduct two-way risk early warning for the operation of the large power grid backbone optical communication system; The pre-established power-communication two-way coupled network twin body is constructed through the historical device data of the large power grid network and the historical actual environment information; Among them, the preset threshold of failure risk is a dynamic value, and the adjustment method specifically includes: Based on historical operation and maintenance data, for a certain fault event O a , the risk probability of the z-th output of the concurrent fault event chain network of the large power grid backbone optical communication system in the digital twin layer is F a,z , and the self-similarity variance of the continuous H outputs of the concurrent fault event chain network is defined as: Among them, F a,h is the risk probability of the h-th output of the concurrent fault event chain network of the backbone optical communication system of the large power grid; If it is determined that the output result of this time is a misjudgment after maintenance, the failure risk preset threshold of the next warning moment for the failure event O a will be increased The adjustment formula for the preset failure risk threshold is as follows: W represents the importance of the communication device at the node associated with the fault event O a associated.

7. An electronic device, characterized in that, Including a processor and a memory, the processor is used to execute the computer program stored in the memory to implement the two-way risk warning method for the operation of the large power grid backbone optical communication system as described in claim 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, and when the at least one instruction is executed by the processor, it implements the two-way risk warning method for the operation of the large power grid backbone optical communication system as described in claim 6.

Images