Relay protection interlocking control method based on full-process fault prediction

Abstract

The present application discloses a relay protection interlocking control method based on full-process fault prediction, which relates to the field of relay protection technology and includes the following steps: dividing power grid equipment combinations based on fault impact range and equipment complexity; constructing a power grid equipment fault prediction model based on the correlation between single device faults and equipment combination faults according to historical power grid equipment fault data and power grid equipment combinations; obtaining the probability of faults with the same principle and the probability of faults with the same object according to the protection principle and protection object of the relay protection equipment; obtaining the full-process prediction results of power grid equipment faults according to current power grid operation data and the power grid equipment fault prediction model; and obtaining the relay protection interlocking control strategy based on the full-process prediction results of power grid equipment faults in combination with the interlocking relationship of the relay protection equipment, the probability of faults with the same principle and the probability of faults with the same object. The beneficial effects of the present application are: achieving a balance between comprehensive fault handling and rapid system response.

Description

Technical Field

[0001] The present application relates to the field of relay protection technology, and in particular to a relay protection interlocking control method based on full-process fault prediction. Background Art

[0002] Relay protection interlock control is an electrical interlocking mechanism based on relays and logic circuits. Its core function is to detect faults or abnormal conditions by monitoring changes in electrical quantities (such as current, voltage, and power) in the power system and trigger protective actions according to preset logic. For example, if a short circuit or overload is detected, the interlock control will disconnect the faulty equipment or adjust its operating status to prevent the accident from escalating.

[0003] However, as power systems continue to expand in size and complexity, existing relay protection strategies face challenges such as irrational protection device configuration and inaccurate fault identification and location. This has led to a decline in relay protection performance and compromised system security. In particular, a failure in any power device can potentially cause cascading failures in other power devices. Relay protection in related technologies typically outputs relay protection strategies based on the fault prediction results of a single power device, making it difficult to adapt to the dynamic topology of today's integrated power grids. This results in poor accuracy and efficiency in relay protection disconnection during power device failures, making it impossible to promptly and accurately disconnect all faulty power devices.

[0004] The patent, "A Digital Operation and Maintenance System and Method for Relay Protection," published with publication number CN 119919127 A and date of publication May 2, 2025, specifically discloses predicting whether a relay protection device will experience a fault in the future. If so, determining a set of robots capable of operation and maintenance based on a collection of fault types; determining the number of robots from the set of robots capable of operation and maintenance to be dispatched to the relay protection device for operation and maintenance based on the preset importance value of the relay protection device; calculating the priority value for each robot in the set of robots capable of operation and maintenance to be dispatched to the relay protection device for operation and maintenance, and selecting a corresponding number of robots from the set of robots capable of operation and maintenance for operation and maintenance based on the priority values, in descending order. This solution only compensates for potential faults in the relay protection device. When a cascading fault occurs in power equipment, it relies entirely on the initial disconnection settings of the relay protection device and cannot promptly and accurately disconnect all faulty power equipment.

[0005] The patent, "A Visualized Operation and Maintenance System for Secondary Equipment in Intelligent Substations," with publication number CN 119966075A and publication date May 9, 2025, specifically discloses: a secondary equipment data acquisition module for acquiring warning briefs, diagnostic briefs, fault briefs, and protection action messages from secondary equipment at the substation end; a secondary equipment intelligent analysis module for parsing warning briefs, diagnostic briefs, and fault briefs to obtain basic warning information, basic diagnostic information, basic fault information, and the entire protection action process, and generating warning information, diagnostic links, and fault links; a secondary equipment remote operation and maintenance module for generating safety measure tickets from a constructed typical safety measure library based on the maintenance scope and maintenance type; and a visualization module for visually displaying basic warning information, basic diagnostic information, and basic fault information, and demonstrating the entire protection action process. This solution, while providing a visualization module to demonstrate the entire warning, diagnostic, and fault information, as well as the protection action process, and to inversely analyze the protection action, still fails to guarantee the accurate and timely shutdown of all power equipment in the event of a cascading failure. Summary of the Invention

[0006] In view of the fact that the existing technology cannot take into account both the accuracy and timeliness of disconnection when an interlocking fault occurs in power equipment, the present application provides a relay protection interlocking control method based on the prediction of the entire fault process, and utilizes the relationship between equipment combinations and single equipment failures to construct a power grid equipment fault prediction model. The fault propagation relationship between equipment is reflected through the correlation between single equipment and equipment combinations, and the impact of a single equipment failure on the failures of other equipment is reflected, thereby reducing the limitations of single equipment failure prediction and improving the accuracy of the prediction of the entire power equipment fault process. At the same time, the interlocking relationship of the relay protection equipment, the probability of failures with the same principle and the probability of failures with the same object are combined to dynamically adjust the interlocking control strategy of the relay protection equipment, so as to control the action of the relay protection equipment with optimal logic, shorten the fault processing time, and improve the accuracy of fault power-off protection.

[0007] To achieve the above-mentioned technical objectives, the present application provides a technical solution, which is a relay protection interlocking control method based on the prediction of the entire fault process, comprising the following steps: dividing the power grid equipment combination based on the fault impact range and the equipment complexity; constructing a power grid equipment fault prediction model based on the correlation between the historical fault data of the power grid equipment and the power grid equipment combination and the single equipment failure and the equipment combination failure; obtaining the probability of failures with the same principle and the probability of failures with the same object according to the protection principle of the relay protection equipment and the protection object; obtaining the current power grid operation data, and obtaining the prediction results of the entire power grid equipment failure process according to the current power grid operation data and the power grid equipment fault prediction model; obtaining the relay protection interlocking control strategy according to the prediction results of the entire power grid equipment failure process, combined with the interlocking relationship of the relay protection equipment, the probability of failures with the same principle and the probability of failures with the same object.

[0008] Furthermore, the division of power grid equipment combinations based on the fault impact range and equipment complexity includes: obtaining the fault impact range of each power grid device based on the historical fault data of the power grid equipment; obtaining the equipment complexity of each power equipment based on the line correlation and redundancy of the power grid equipment; obtaining associated equipment associated with each power equipment based on the fault impact range and equipment complexity, and obtaining a power grid equipment combination based on the associated equipment and the corresponding power equipment.

[0009] Furthermore, the method of obtaining associated devices associated with each power device based on the fault impact range and equipment complexity, and obtaining a power grid device combination based on the associated devices and the corresponding power equipment includes: obtaining the fault impact range of each power grid device based on the historical fault data of the power grid equipment based on the flow transfer impact and load loss impact.

[0010] Furthermore, the construction of a power grid equipment failure prediction model based on the correlation between single equipment failure and equipment combination failure according to the historical failure data of power grid equipment and the combination of power grid equipment includes: constructing a single equipment failure prediction model based on the historical failure data of power grid equipment based on a neural network architecture; constructing an equipment combination failure prediction model based on the historical failure data of power grid equipment and the combination of power grid equipment based on a neural network architecture; fusing the single equipment failure prediction model and the equipment combination failure prediction model, and using the output of the single equipment failure prediction model as the input of the equipment combination failure prediction model to obtain the power grid equipment failure prediction model.

[0011] Furthermore, the method of obtaining the probability of failure with the same principle and the probability of failure with the same object according to the protection principle of the relay protection equipment and the protection object includes: dividing the relay protection equipment according to the protection principle of the relay protection equipment and the protection object, and obtaining a set of the same principle and a set of the same object; obtaining the probability of failure with the same principle according to the fault type and the overall fault ratio of the set of the same principle; and obtaining the probability of failure with the same object according to the fault type and the overall fault ratio of the set of the same object.

[0012] Furthermore, the obtaining of current power grid operation data and obtaining a prediction result of the entire process of power grid equipment failure based on the current power grid operation data and the power grid equipment failure prediction model include: obtaining the failure probability of a single power device based on the current power grid operation data and the power grid equipment failure prediction model; when the failure probability of a single power device is greater than a preset probability threshold, obtaining a power grid equipment combination of the power device; obtaining the equipment combination failure probability based on the power grid equipment combination; and outputting the failure probability of a single power device and the equipment combination failure probability as the prediction result of the entire process of power grid equipment failure.

[0013] Furthermore, the construction of a power grid equipment failure prediction model based on historical power grid equipment failure data and the correlation between single equipment failure and equipment combination failure based on the power grid equipment combination also includes: constructing a combined scheduling layer based on a preset probability threshold and a preset combination threshold; executing in the combined scheduling layer: when the failure probability of a single power device is greater than the preset probability threshold, calling the equipment combination failure prediction model; when the failure probability of a single power device is less than or equal to the preset probability threshold, obtaining the failure probability of all single power devices in the corresponding power grid equipment combination; if the sum of the failure probabilities of all single power devices in the power grid equipment combination is greater than the preset combination threshold, calling the equipment combination failure prediction model; if the sum of the failure probabilities of all single power devices in the power grid equipment combination is less than or equal to the preset combination threshold, not calling the equipment combination failure prediction model.

[0014] Furthermore, the method of obtaining the relay protection interlocking control strategy based on the prediction results of the entire process of power grid equipment failure combined with the interlocking relationship of the relay protection equipment, the probability of failures with the same principle, and the probability of failures with the same object includes: obtaining the interlocking relationship of the relay protection equipment based on the historical interlocking control data of the relay protection equipment; obtaining a set of relay protection interlocking control strategies based on the prediction results of the entire process of power grid equipment failure and the interlocking relationship of the relay protection equipment; and obtaining the optimal relay protection interlocking control strategy based on the set of relay protection interlocking control strategies based on the minimum failure probability and the minimum loss.

[0015] Furthermore, the method of obtaining the optimal relay protection interlocking control strategy based on the minimum failure probability and the minimum loss according to the relay protection interlocking control strategy set, the failure probability of the same principle and the failure probability of the same object includes: obtaining the failure probability of the same principle and the failure probability of the same object of each relay protection device in the relay protection interlocking control strategy set; calculating the failure probability of each relay protection interlocking control strategy according to the failure probability of the same principle and the failure probability of the same object; calculating the loss of each relay protection interlocking control strategy according to the action loss of the relay protection device; constructing an optimization objective function with the minimum failure probability and the minimum loss, and obtaining the optimal relay protection interlocking control strategy by optimizing the objective function and the relay protection interlocking control strategy set.

[0016] Furthermore, the method of obtaining associated devices associated with each power device based on the fault impact range and device complexity, and obtaining a power grid device combination based on the associated devices and the corresponding power devices includes: filtering out power devices that match the flow transfer impact based on the number of branches with corresponding functions of the line where each power grid device is located; filtering out power devices that match the load loss impact based on the redundant configuration of each power grid device; and using the power devices in the filtered flow transfer impact and the filtered load loss impact as associated devices of each power device, and obtaining a power grid device combination based on the associated devices and the corresponding power equipment.

[0017] The beneficial effects of this application are as follows: the power grid architecture is divided into multiple equipment combinations according to the fault impact range and equipment complexity, and a power grid equipment fault prediction model is constructed by using the relationship between the equipment combination and the single equipment failure. This not only retains the fault characteristics of a single device, but also reflects the fault propagation relationship between devices through the correlation between a single device and the equipment combination, and reflects the impact of a single device failure on the failures of other devices, reducing the limitations of single device failure prediction, outputting fault prediction results based on the entire fault process, and improving the accuracy of the prediction with a global perspective.

[0018] At the same time, the interlocking control strategy of the relay protection equipment is dynamically adjusted based on the interlocking relationship of the relay protection equipment, the probability of faults with the same principle and the probability of faults with the same object, so as to avoid the impact of false operation or refusal of the relay protection equipment on the safe operation of the power grid. The action of the relay protection equipment is controlled with the optimal logic, the fault handling time is shortened, the accuracy of fault power-off protection is improved, the reliability of the power grid operation is improved, and the comprehensiveness of fault handling and the rapidity of system response are taken into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of the relay protection interlocking control method based on full-process fault prediction in this application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of this application more clear, the application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific implementation method described here is only an optimal embodiment of this application, which is only used to explain this application and does not limit the scope of protection of this application. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0021] like Figure 1 As shown in FIG, the relay protection interlocking control method based on the prediction of the entire fault process includes the following steps:

[0022] Divide the grid equipment portfolio based on the scope of fault impact and equipment complexity;

[0023] Build a power grid equipment failure prediction model based on the historical failure data of power grid equipment and the correlation between single equipment failure and equipment combination failure;

[0024] According to the protection principle of the relay protection equipment and the protection object, the probability of failure with the same principle and the probability of failure with the same object are obtained;

[0025] Obtain current grid operation data, and obtain the prediction results of the entire grid equipment failure process based on the current grid operation data and the grid equipment failure prediction model;

[0026] The relay protection interlocking control strategy is obtained based on the prediction results of the entire process of power grid equipment failure, combined with the interlocking relationship of relay protection equipment, the probability of failures with the same principle, and the probability of failures with the same object.

[0027] In this embodiment, the power grid architecture is divided into multiple equipment combinations based on the fault impact range and equipment complexity, and the relationship between the equipment combination and the single equipment failure is used to construct a power grid equipment failure prediction model. This not only retains the fault characteristics of a single device, but also reflects the fault propagation relationship between devices through the correlation between a single device and the equipment combination, and reflects the impact of a single device failure on the failures of other devices. This reduces the limitations of single device failure prediction, outputs fault prediction results based on the entire fault process, and improves the accuracy of prediction with a global perspective.

[0028] At the same time, the interlocking control strategy of the relay protection equipment is dynamically adjusted based on the interlocking relationship of the relay protection equipment, the probability of faults with the same principle and the probability of faults with the same object, so as to avoid the impact of false operation or refusal of the relay protection equipment on the safe operation of the power grid. The action of the relay protection equipment is controlled with the optimal logic, the fault handling time is shortened, the accuracy of fault power-off protection is improved, the reliability of the power grid operation is improved, and the comprehensiveness of fault handling and the rapidity of system response are taken into account.

[0029] Specifically, the power grid equipment combinations based on the fault impact range and equipment complexity include:

[0030] Obtain the fault impact range of each power grid device based on the historical fault data of the power grid equipment;

[0031] Obtain the equipment complexity of each power device based on the line correlation and redundancy of the power grid equipment;

[0032] The associated devices associated with each power device are obtained based on the fault impact range and device complexity, and the power grid device combination is obtained based on the associated devices and the corresponding power devices.

[0033] The grid topology is segmented based on the fault impact range and device complexity. When a current grid device fails, the remaining grid devices within the impact area are treated as associated devices of the current grid device. At the same time, the device complexity is used to balance the excessive spread of the fault impact range, which not only improves the globality of fault handling but also retains the targetedness of fault handling, thereby improving the accuracy of subsequent fault prediction.

[0034] Among them, obtaining the fault impact range of each power grid device based on the historical fault data of the power grid equipment includes:

[0035] Based on the influence of power flow transfer and load loss, the fault impact range of each power grid device is obtained according to the historical fault data of the power grid equipment.

[0036] The impact range of power grid equipment failure is quantified through two dimensions: flow transfer impact and load loss impact. The flow transfer impact is used to quantify the impact of over-limit on remaining lines caused by flow redistribution after power equipment failure. The load loss impact is used to quantify the load impact of power outage caused by the failure.

[0037] Specifically, the influence of the flow transfer is obtained based on the flow redistribution:

[0038] ;

[0039] ;

[0040] in, represents the current in the branch line where power equipment j is located after power equipment i fails; Indicates the current increment coefficient of the branch line between power equipment i and power equipment j; It represents the current of the branch line where the power equipment i is located when there is no fault; It represents the current of the branch line where power equipment j is located when power equipment i is not faulty; Indicates the maximum current of the branch line where the power equipment j is located; Indicates whether the branch line where the power equipment j is located exceeds the limit due to the power flow transfer caused by the failure of the power equipment i. Indicates that the power flow transfer of the branch line where power equipment j is located exceeds the limit due to the failure of power equipment i. Indicates that the branch line where the power equipment j is located has not exceeded the power flow transfer limit caused by the failure of the power equipment i. The power equipment j is affected by the power flow transfer.

[0041] The power equipment existing in the power outage area caused by the power equipment i fault is used as the load loss impact, and the fault impact range is obtained by the flow transfer impact and the load loss impact.

[0042] The equipment complexity of each power device is obtained based on the line correlation and redundancy of the power grid equipment, including:

[0043] The device complexity of each power device is obtained according to the number of branches corresponding to the function of the line where each power grid device is located and the redundancy configuration of each power grid device.

[0044] The associated devices associated with each power device are obtained based on the fault impact range and device complexity, and the grid device combination is obtained based on the associated devices and the corresponding power devices.

[0045] Eliminate power equipment that is affected by power flow transfer based on the number of branches corresponding to the functions of the lines where each power grid equipment is located;

[0046] Eliminate the matching power equipment affected by load loss based on the redundant configuration of each power grid equipment;

[0047] The power equipment in which the influence of the power flow transfer and the influence of the load loss are eliminated is used as the associated equipment of each power equipment, and the power grid equipment combination is obtained by using the associated equipment and the corresponding power equipment.

[0048] In this embodiment, branches with the same function connected to the line where the power grid equipment resides are obtained. If there are two or more such branches, the power equipment on these branches is excluded from the impact of power flow transfer. Redundancy refers to physical backup equipment at the power equipment level. Through parallel devices, backup lines, or backup power sources, the load can continue to be supplied even after a power grid equipment failure. For example, if the failure of power equipment 1 causes a power outage in area A, but backup power equipment 2 exists to restore power to area A, backup power equipment 2, although located in area A, is affected by the load loss and is therefore excluded. Through the dual screening mechanism of branch line redundancy and equipment redundancy, non-critical equipment affected by power flow transfer and load loss is screened out. When the number of branches corresponding to a certain function exceeds 2, it indicates that a redundant power flow path exists. At this time, although the branch line affected by the transfer will exceed the limit in the power flow transfer impact calculation, it can be transferred through other branches. In other cases, in order to further improve the accuracy of screening, the corresponding number of branches can also be calculated based on the power flow transfer caused by the power equipment failure in historical fault data, so as to prepare to screen out power equipment that will not actually exceed the limit because the other branches share the power flow transfer. When there is a physical backup device, even if the failure of the power equipment causes a power outage in the area and thus causes the shutdown or restriction of other power equipment in the area, the activation of the physical backup device can restore the load in the area to a certain extent, thereby screening out the backup device and the power equipment within the load recovery range of the backup device from the load loss impact. This avoids including non-critical equipment in the interlocking control range, and avoids the complexity of protection coordination caused by the simultaneous interlocking of backup equipment and faulty equipment. It allows for the consideration of the full fault situation caused by the failure of a certain power equipment, and avoids the occurrence of unnecessary interlocking control and repeated interlocking control, thereby improving the accuracy of interlocking control and the reliability of power supply.

[0049] Branches with the same function must be electrically connected in parallel and have overlapping power supply targets. Physical backup equipment includes parallel devices, equipment on backup lines, and backup power sources. Power equipment affected by power flow shifts and load losses after screening are considered associated with the currently assumed faulty power equipment, and the associated equipment and the currently assumed faulty power equipment are considered the grid equipment combination.

[0050] Based on the historical fault data of power grid equipment and the correlation between single equipment failure and equipment combination failure, a power grid equipment failure prediction model is constructed, which includes:

[0051] Build a single equipment failure prediction model based on the historical failure data of power grid equipment based on the neural network architecture;

[0052] Based on the neural network architecture, a device combination fault prediction model is constructed according to the historical fault data of power grid equipment and the combination of power grid equipment;

[0053] The single device fault prediction model and the device combination fault prediction model are integrated, and the output of the single device fault prediction model is used as the input of the device combination fault prediction model to obtain the power grid equipment fault prediction model.

[0054] In this embodiment, the LSTM neural network architecture is used to learn the fault characteristics and feature weights of a single power device in the historical time series, and the correlation relationship of the fault of a single power device is constructed. The heterogeneous graph model is used to learn the fault correlation relationship between the device combinations. The power grid equipment failure prediction model is constructed based on the correlation relationship of the failure of a single power device and the failure relationship between the device combinations. At the same time, the individual failure of the equipment and the combined failure of the equipment are taken into consideration to improve the comprehensiveness of the prediction of the entire fault process.

[0055] Historical fault data includes at least fault time, fault type, fault duration, load data, and fault environment data. This data is preprocessed to remove abnormal and erroneous data. The preprocessed data is then subjected to data screening to obtain training, validation, and test sets. An LSTM network architecture is selected and trained using the training, validation, and test sets. The network parameters are adjusted using a backpropagation algorithm to obtain the characteristic weights of each fault feature. The LSTM network can capture long-term dependencies between equipment faults, identify changes in equipment faults over time, and improve prediction accuracy.

[0056] Using heterogeneous graph models to learn fault associations between device combinations includes:

[0057] Each power device is defined as a node corresponding to the device type according to the device type, and the connection relationship between devices is obtained as the edge according to the combination of power grid devices to construct a heterogeneous graph;

[0058] Historical fault data is used to train heterogeneous graphs to obtain the association relationship between nodes and edges and the fault propagation law.

[0059] The heterogeneous graph model can adopt a heterogeneous graph attention network architecture. The fault propagation law at least includes fault propagation speed, fault propagation order, and fault propagation time.

[0060] According to the protection principle of the relay protection equipment and the protection object, the probability of failure with the same principle and the probability of failure with the same object are obtained, including:

[0061] Classify relay protection devices according to their protection principles and protection objects, and obtain sets with the same principles and sets with the same objects;

[0062] According to the historical failure data of the same principle set, the overall failure ratio corresponding to the common cause failure factor is calculated based on the common cause failure to obtain the same principle failure probability;

[0063] The probability of the same failure of an object is obtained by calculating the overall failure ratio corresponding to the failure type based on the historical failure data of the same set of objects.

[0064] According to the consistency of protection principles, relay protection devices with the same protection principle are divided into the same principle set. The overall failure ratio corresponding to the common cause failure factor is calculated based on the number of failures occurring simultaneously under the same fault factor. The overall failure ratio is used as the probability of failure with the same principle corresponding to the common cause failure factor.

[0065] Among them, the protection principles include at least the longitudinal differential protection principle, the distance protection principle, and the overcurrent protection principle. Correspondingly, the set of identical principles includes the longitudinal differential protection principle set, the distance protection principle set, and the overcurrent protection principle set. All equipment in the power system that adopts the longitudinal differential protection principle, such as the longitudinal differential protection device of the transmission line and the longitudinal differential protection device of the transformer, are classified into the longitudinal differential protection principle set; all equipment that works based on the distance protection principle through the same principle, such as the distance protection device of the transmission line of different voltage levels, are classified into the distance protection principle set; all types of equipment that adopt the overcurrent protection principle, such as the overcurrent protection device of the low-voltage distribution line and the overcurrent protection device of the motor, are classified into the overcurrent protection principle set. The equipment manufacturer, equipment model, and operating environment are used as common cause failure factors to obtain historical fault data corresponding to each identical principle set, calculate the number of equipment that fails simultaneously in the identical principle set, and calculate the overall fault ratio:

[0066] ;

[0067] Where P represents the overall failure rate corresponding to the common cause failure factor, X represents the number of devices that fail simultaneously in the same principle set under the same common cause failure factor, and Y represents the total number of devices in the principle set.

[0068] Devices with different protection principles differ in structure, function, and operational characteristics, and face varying common-cause failure risks. By grouping devices with the same protection principle into similar groups, and then analyzing common-cause failure factors and calculating the probability of similar-principle failures, we can more accurately predict the failure probability of relay protection devices.

[0069] At the same time, the relay protection equipment is divided according to the type of protected objects to obtain the object identity set. The proportion of synchronous faults under the same fault condition in the object identity set is obtained according to the fault type, which is used as the object identity fault probability corresponding to the fault type.

[0070] Obtain current grid operation data, and obtain the prediction results of the entire grid equipment failure process based on the current grid operation data and the grid equipment failure prediction model, including:

[0071] Obtain the failure probability of a single power device based on current power grid operation data and power grid equipment failure prediction models;

[0072] When the failure probability of a single power device is greater than a preset probability threshold, obtaining a power grid device combination of the power device;

[0073] Obtain the combined failure probability of equipment based on the combination of power grid equipment;

[0074] The failure probability of a single power device and the probability of a combined failure of the devices are used as the output of the prediction results for the entire process of power grid equipment failure.

[0075] Current grid operation data includes at least current, voltage, environmental data, switch status, and relay protection device status. By collecting grid operation data in real time, taking into account the interactions between devices and fault propagation paths, and calculating the failure probability of device combinations, the accuracy of predictions for the entire grid fault process is improved. This avoids focusing solely on the failure of a single power device, which can lead to overlooking the propagation of faults to other devices, and improves the accuracy of fault interlocking control.

[0076] Correspondingly, in this embodiment, constructing a power grid equipment failure prediction model based on historical power grid equipment failure data and power grid equipment combinations based on the correlation between single device failures and device combination failures further includes:

[0077] Construct a combined scheduling layer based on preset probability thresholds;

[0078] Execute in the combined scheduling layer:

[0079] When the failure probability of a single power device is greater than a preset probability threshold, the device combination failure prediction model is called; when the failure probability of a single power device is less than or equal to the preset probability threshold, the device combination failure prediction model is not called.

[0080] By setting up the combined scheduling layer, useless scheduling of equipment combined fault prediction can be avoided, computing pressure can be reduced, and computing efficiency can be improved.

[0081] In other embodiments, constructing a power grid equipment failure prediction model based on historical power grid equipment failure data and power grid equipment combinations based on correlation between single device failures and device combination failures further includes:

[0082] Building a combined scheduling layer based on a preset probability threshold and a preset combination threshold;

[0083] Execute in the combined scheduling layer:

[0084] When the failure probability of a single power device is greater than a preset probability threshold, the device combination failure prediction model is called;

[0085] When the failure probability of a single power device is less than or equal to a preset probability threshold, the failure probability of all single power devices in the corresponding power grid device combination is obtained;

[0086] If the sum of the failure probabilities of all individual power devices in the power grid equipment combination is greater than the preset combination threshold, the equipment combination failure prediction model is called; if the sum of the failure probabilities of all individual power devices in the power grid equipment combination is less than or equal to the preset combination threshold, the equipment combination failure prediction model is not called.

[0087] Through the dual judgment mechanism, the mutual influence of faults between power equipment is quantified to avoid inaccurate fault risk assessment caused by ignoring the synergy between equipment. At the same time, when the failure probability of a single power equipment is less than or equal to the preset probability threshold and the sum of the failure probabilities of all single power equipment in the power grid equipment combination is less than or equal to the preset combination threshold, the equipment combination fault prediction model will not be called, which reduces excessive model calls, reduces computing resource consumption, shortens response time, and improves the efficiency of fault prediction throughout the entire process.

[0088] It is understandable that the prediction result of the entire process of power grid equipment failure includes at least the faulty power equipment information, the fault sequence and the fault time sequence.

[0089] The relay protection interlocking control strategy is obtained based on the prediction results of the entire process of power grid equipment failure, combined with the interlocking relationship of relay protection equipment, the probability of failures with the same principle, and the probability of failures with the same object, including:

[0090] Obtaining the interlocking relationship of relay protection equipment based on the historical interlocking control data of relay protection equipment;

[0091] Obtaining a set of relay protection interlocking control strategies based on the prediction results of the entire process of power grid equipment failure and the interlocking relationship of relay protection equipment;

[0092] Based on the minimum failure probability and minimum loss, the optimal relay protection interlocking control strategy is obtained according to the set of relay protection interlocking control strategies, the failure probability of the same principle and the failure probability of the same object.

[0093] The next-generation dispatch management subsystem, the distribution network dispatch management system, acquires historical interlocking control data for relay protection devices. This data includes at least the relay protection device's action time, action sequence, triggering conditions, and information related to other devices. The interlocking relationships between relay protection devices are used to obtain a set of relay protection interlocking control strategies corresponding to the full-process prediction results of power grid equipment failures. For example, when a power equipment failure occurs, which relay protection devices will be activated in sequence, as well as their triggering conditions and action logic.

[0094] For the same protected object, redundant configuration of relay protection devices with different principles can reduce the probability of failure but increase the additional loss of the relay protection devices. Therefore, the minimum loss is constructed based on the action loss of the relay protection devices. Combined with the failure probability of the same principle and the failure probability of the same object, the relay protection interlocking control strategy with the optimal balance between minimum loss and redundant configuration is selected from the relay protection interlocking control strategy set.

[0095] Specifically, obtaining the optimal relay protection interlocking control strategy based on the minimum failure probability and the minimum loss according to the relay protection interlocking control strategy set, the failure probability of the same principle and the failure probability of the same object includes:

[0096] Obtain the probability of failure of the same principle and the probability of failure of the same object of each relay protection device in the relay protection interlocking control strategy set;

[0097] Calculate the failure probability of each relay protection interlocking control strategy based on the same principle failure probability and the same object failure probability;

[0098] Calculate the loss of each relay protection interlocking control strategy based on the operation loss of the relay protection equipment;

[0099] An optimization objective function is constructed with minimum failure probability and minimum loss, and the optimal relay protection interlocking control strategy is obtained by optimizing the objective function and the relay protection interlocking control strategy set.

[0100] The optimal relay protection interlocking control strategy is selected by balancing the fault probability and loss, and a setting list is output according to the optimal relay protection interlocking control strategy to achieve relay protection interlocking control of the entire fault process.

[0101] In some cases, obtaining the optimal relay protection interlocking control strategy based on the minimum failure probability and the minimum loss according to the relay protection interlocking control strategy set, the failure probability of the same principle and the failure probability of the same object further includes:

[0102] Constructing redundancy strategies based on different principles based on failure probability thresholds;

[0103] When the failure probability of each relay protection interlocking control strategy is greater than or equal to the failure probability threshold, the redundant relay protection interlocking control strategy output by the nearest relay protection device corresponding to the full process prediction result of the power grid equipment failure is called according to the different principle redundancy strategy.

[0104] When the failure probability of each relay protection interlocking control strategy is greater than or equal to the failure probability threshold, the first setting order is output based on the optimal relay protection interlocking control strategy. Simultaneously, based on the grid topology, a redundant relay protection interlocking control strategy is output based on the closest relay protection device that can remove each power device in the full-process prediction results of the grid equipment fault. The redundant relay protection interlocking control strategy is then used to output the second setting order. When a power equipment fault occurs, the first setting order corresponding to the optimal relay protection interlocking control strategy is obtained based on the full-process prediction results of the fault. Real-time execution feedback for the first setting order is obtained. If a protection anomaly occurs in the first setting order, relay protection is immediately executed based on the second setting order to ensure timely fault removal and avoid fault expansion due to the failure of a single strategy. Furthermore, redundant relay protection devices are selected based on redundant strategies with different principles, further reducing the risk of both strategies failing.

[0105] The specific implementation method described above is a preferred implementation method of the relay protection interlocking control method based on the prediction of the entire fault process of this application, and is not intended to limit the specific implementation scope of this application. The scope of this application includes but is not limited to this specific implementation method. All equivalent changes made in accordance with the shape and structure of this application are within the protection scope of this application.

Claims

1. A relay protection interlocking control method based on full-process fault prediction is characterized by: The steps include: Divide the grid equipment portfolio based on the scope of fault impact and equipment complexity; Build a power grid equipment failure prediction model based on the historical failure data of power grid equipment and the correlation between single equipment failure and equipment combination failure; According to the protection principle of the relay protection equipment and the protection object, the probability of failure with the same principle and the probability of failure with the same object are obtained; Obtain current grid operation data, and obtain the prediction results of the entire grid equipment failure process based on the current grid operation data and the grid equipment failure prediction model; The relay protection interlocking control strategy is obtained based on the prediction results of the entire process of power grid equipment failure, combined with the interlocking relationship of relay protection equipment, the probability of failures with the same principle, and the probability of failures with the same object; in, Based on the neural network architecture, a single device fault prediction model is constructed based on the historical fault data of power grid equipment. Based on the neural network architecture, a device combination fault prediction model is constructed based on the historical fault data of power grid equipment and the combination of power grid equipment. The single device fault prediction model and the device combination fault prediction model are integrated, and the output of the single device fault prediction model is used as the input of the device combination fault prediction model to obtain the power grid equipment fault prediction model. Obtain the failure probability of a single power device based on current power grid operation data and a power grid equipment failure prediction model; when the failure probability of a single power device is greater than a preset probability threshold, obtain the power grid equipment combination of the power device; obtain the combined failure probability of the devices based on the power grid equipment combination; and output the failure probability of a single power device and the combined failure probability of the devices as the prediction result of the entire power grid equipment failure process. The interlocking relationship of relay protection equipment is obtained based on the historical interlocking control data of relay protection equipment; the relay protection interlocking control strategy set is obtained based on the prediction results of the entire process of power grid equipment failure and the interlocking relationship of relay protection equipment; based on the minimum failure probability and minimum loss, the optimal relay protection interlocking control strategy is obtained according to the relay protection interlocking control strategy set, the failure probability of the same principle and the failure probability of the same object.

2. The relay protection interlocking control method based on full-process fault prediction according to claim 1, characterized in that: The division of power grid equipment combinations based on fault impact range and equipment complexity includes: Obtain the fault impact range of each power grid device based on the historical fault data of the power grid equipment; Obtain the equipment complexity of each power device based on the line correlation and redundancy of the power grid equipment; The associated devices associated with each power device are obtained based on the fault impact range and device complexity, and the power grid device combination is obtained based on the associated devices and the corresponding power devices.

3. The relay protection interlocking control method based on full-process fault prediction according to claim 2, characterized in that: The method of obtaining associated devices associated with each power device based on the fault impact range and device complexity, and obtaining a power grid device combination based on the associated devices and the corresponding power devices includes: Based on the influence of power flow transfer and load loss, the fault impact range of each power grid device is obtained according to the historical fault data of the power grid equipment.

4. The relay protection interlocking control method based on full-process fault prediction according to claim 1, characterized in that: The obtaining of the same-principle fault probability and the same-object fault probability according to the protection principle of the relay protection device and the protection object includes: Classify relay protection devices according to their protection principles and protection objects, and obtain sets with the same principles and sets with the same objects; Obtain the probability of failures with the same principle based on the failure types and overall failure ratio of the same principle set; The probability of the same failure of an object is obtained based on the failure type and overall failure ratio of the same set of objects.

5. The relay protection interlocking control method based on full-process fault prediction according to claim 1, characterized in that: The method of constructing a power grid equipment failure prediction model based on historical power grid equipment failure data and power grid equipment combinations based on the correlation between single equipment failure and equipment combination failures further includes: Building a combined scheduling layer based on a preset probability threshold and a preset combination threshold; Execute in the combined scheduling layer: When the failure probability of a single power device is greater than a preset probability threshold, the device combination failure prediction model is called; When the failure probability of a single power device is less than or equal to a preset probability threshold, the failure probability of all single power devices in the corresponding power grid device combination is obtained; If the sum of the failure probabilities of all individual power devices in the power grid equipment combination is greater than the preset combination threshold, the equipment combination failure prediction model is called; if the sum of the failure probabilities of all individual power devices in the power grid equipment combination is less than or equal to the preset combination threshold, the equipment combination failure prediction model is not called.

6. The relay protection interlocking control method based on full-process fault prediction according to claim 1, characterized in that: The method of obtaining the optimal relay protection interlocking control strategy based on the minimum failure probability and the minimum loss according to the relay protection interlocking control strategy set, the failure probability of the same principle and the failure probability of the same object includes: Obtain the probability of failure of the same principle and the probability of failure of the same object of each relay protection device in the relay protection interlocking control strategy set; Calculate the failure probability of each relay protection interlocking control strategy based on the same principle failure probability and the same object failure probability; Calculate the loss of each relay protection interlocking control strategy based on the operation loss of the relay protection equipment; An optimization objective function is constructed with minimum failure probability and minimum loss, and the optimal relay protection interlocking control strategy is obtained by optimizing the objective function and the relay protection interlocking control strategy set.

7. The relay protection interlocking control method based on full-process fault prediction according to claim 3 is characterized in that: The method of obtaining associated devices associated with each power device based on the fault impact range and device complexity, and obtaining a power grid device combination based on the associated devices and the corresponding power devices includes: Eliminate power equipment that is affected by power flow transfer based on the number of branches corresponding to the functions of the lines where each power grid equipment is located; Eliminate the matching power equipment affected by load loss based on the redundant configuration of each power grid equipment; The power equipment in which the influence of the power flow transfer and the influence of the load loss are eliminated is used as the associated equipment of each power equipment, and the power grid equipment combination is obtained by using the associated equipment and the corresponding power equipment.

Images