Power failure intelligent monitoring method and system of power system

Abstract

The invention relates to the field of intelligent monitoring, in particular to a power failure intelligent monitoring method and system for a power system, and the system comprises a plurality of collection modules which are used for transmitting real-time voltage values and collecting minute-level high-frequency voltage collection data, a receiving module which is used for receiving the minute-level high-frequency voltage collection data, carrying out the pre-learning, generating corresponding pre-learning data, and transmitting the pre-learning data to a server. The learning module is used for selecting a plurality of index characteristics for learning and generating a corresponding voltage curve graph, and the processing module is used for receiving the voltage curve graph, judging whether the power system is powered off or not according to whether a breakpoint occurs in the voltage curve graph, and sending power failure information to a power supply bureau and repairing the power system if the power system is powered off. Intelligent monitoring and quick response of power failure of the power system are achieved, the influence of power failure on users is reduced in an automatic mode, the reliability and operation efficiency of the power system are improved, and a power company is helped to better monitor and maintain stable operation of a power grid.

Description

Technical Field

[0001] The present invention relates to the field of intelligent monitoring, and in particular to a method and system for intelligently monitoring power outages in an electric power system. Background Art

[0002] With the advancement of science and technology, the connection between the normal operation of society and the power system is becoming increasingly close. Power outages will lead to interruptions in industrial production and obstruction of commercial activities, causing huge losses to society and the economy. When a power outage occurs, the intelligent monitoring system can quickly locate the fault point, promptly notify the emergency repair personnel, speed up the fault handling speed, shorten the power outage time, and reduce economic losses. Therefore, the power system's power outage intelligent monitoring method is of great significance. Through real-time monitoring and intelligent analysis, it can timely detect abnormal situations in the power system, reduce the occurrence of power outages, improve the continuity and reliability of power supply, and protect people's lives and property.

[0003] Chinese patent application publication number: CN111431265A discloses a smart power distribution operation and maintenance system applied to subway stations. The deployment of emergency power supply for each monitoring area is realized through the regional division module and the backup power supply module, and various types of lighting equipment are marked and classified in conjunction with the equipment marking module. The abnormalities of the lighting equipment are graded, allocated and processed through the abnormality handling module, so that when the lighting equipment in the subway station has power supply abnormalities, the abnormalities are quickly and reasonably allocated and processed according to the actual situation to ensure the safe operation of the subway station. After an alarm occurs, the emergency power distribution module generates evacuation routes and power distribution instructions, and controls the backup power supply module to execute the power distribution instructions, so that when an alarm such as a regional power outage occurs in the subway station, the load of the backup power supply is reduced, the use time is delayed, and sufficient lighting on the evacuation route is ensured, so that personnel can be evacuated quickly and safely.

[0004] China Patent Authorization Announcement No.: CN205648127U This utility model relates to a smart city single-lamp lighting management system, including an LED lighting lamp, a power supply management circuit, a circuit on-off detection circuit, a charging management circuit, a monitoring host, a driving circuit and a wireless communication module. The charging management circuit and the circuit on-off detection circuit are both connected to the AC power supply management circuit, the monitoring host and the driving circuit are both connected to the charging management circuit, the monitoring host is also connected to the LED lighting lamp through the driving circuit, and the circuit on-off detection circuit and the wireless communication module are both connected to the monitoring host; the charging management circuit is used to process the power from the AC power supply management circuit and output it to the main control circuit and the driving circuit. By detecting a single LED lighting lamp, it is possible to timely understand power outages and power failures; at the same time, according to the driving circuit, a constant current can be output to the LED lighting lamp to extend the service life of the LED lighting lamp.

[0005] However, the above method has the following problems: it is impossible to intelligently monitor the power system in real time and the operation efficiency is low. Summary of the invention

[0006] To this end, the present invention provides a method and system for intelligently monitoring power outages in an electric power system, so as to overcome the problems in the prior art of being unable to intelligently monitor the electric power system in real time and having low operating efficiency.

[0007] To achieve the above object, the present invention provides a power outage intelligent monitoring system for an electric power system, comprising:

[0008] Several acquisition modules, which are used to read real-time voltage values ​​and collect minute-level high-frequency voltage data;

[0009] A receiving module, which is connected to the plurality of acquisition modules, and is used to receive the minute-level high-frequency voltage acquisition data, and pre-learn the minute-level high-frequency voltage acquisition data to generate corresponding pre-learning data;

[0010] A learning module, connected to the receiving module, for receiving the pre-learning data, selecting several indicator features of the pre-learning data, learning the pre-learning data, and generating a corresponding voltage curve graph;

[0011] a processing module, which is connected to the acquisition module and the learning module respectively, and is used to receive the voltage curve graph, and judge whether a power outage occurs in the power system according to whether a breakpoint appears in the voltage curve graph; if a power outage occurs, the power outage information is sent to the power supply bureau and the power system is repaired;

[0012] The pre-learning is to perform standardization processing on the minute-level high-frequency voltage acquisition data, generate a number of pre-learning data with a learning rate of a standard learning rate, and generate corresponding indicator features according to each pre-learning data;

[0013] The standard learning rate is a learning rate that the learning module can recognize, and is inversely proportional to the resolution of the minute-level high-frequency voltage acquisition data. The higher the resolution, the lower the corresponding standard learning rate. For a single learning, the corresponding standard learning rate is a single learning rate.

[0014] Furthermore, the acquisition module includes:

[0015] A timer, which is set with a fixed duration and is used to issue a collection instruction at intervals of the fixed duration;

[0016] A high-frequency voltage sensor, which is connected to the timer and is arranged at a corresponding position of the electric energy meter, and is used to read the real-time voltage value displayed by the electric energy meter when receiving the acquisition instruction;

[0017] A converter is connected to the high-frequency voltage sensor and is used to convert the real-time voltage value into the minute-level high-frequency voltage acquisition data.

[0018] Furthermore, the receiving module includes:

[0019] A receiver, which is used to receive the minute-level high-frequency voltage acquisition data;

[0020] A filter connected to the receiver for filtering the minute-level high-frequency voltage acquisition data and generating corresponding voltage filtering data;

[0021] A pre-learner, which is used to pre-learn the voltage filtering data to generate corresponding pre-learning data,

[0022] Wherein, a filter interval threshold is provided in the filter;

[0023] The filtering interval threshold is an interval formed by the minimum value and the maximum value of the receiving range of the pre-learner.

[0024] Furthermore, the learning module includes:

[0025] A selector, which is used to receive the pre-learning data and select a number of indicator features of the pre-learning data;

[0026] A learner, connected to the selector, receives the pre-learning data and the indicator feature, and learns the pre-learning data to generate a corresponding voltage curve graph;

[0027] A synthesizer, connected to the learner, for marking the voltage curve according to the indicator characteristics,

[0028] Wherein, the learner is provided with a voltage learning model;

[0029] The voltage learning model is generated through training based on a voltage training set formed by pre-learning data corresponding to the minute-level high-frequency voltage acquisition data.

[0030] Furthermore, the high-frequency voltage sensor located at the corresponding position of the electric energy meter receives the collection instruction, copies the real-time voltage value, and transmits the real-time voltage value to the converter. The converter initializes the real-time voltage value to obtain the corresponding original voltage value, performs data calibration on the original voltage value, and obtains the corresponding minute-level high-frequency voltage collection data.

[0031] Furthermore, the filter removes the minute-level high-frequency voltage acquisition data that is not within the filter interval threshold in the minute-level high-frequency voltage acquisition data and forms corresponding voltage filtered data. The pre-learner divides the voltage filtered data according to the standard learning rate based on the voltage training set to form corresponding pre-learning data.

[0032] Further, the learning module selects several indicator features of the pre-learning data, the voltage learning model generates a corresponding voltage curve graph, the synthesizer marks the voltage curve graph in combination with the indicator features, the processing module receives the voltage curve graph, and if the voltage curve graph does not have continuous voltage curve breakpoints, it is determined that the power system has not experienced a power outage.

[0033] The indicator characteristics include voltage level and / or fixed duration.

[0034] Furthermore, if voltage curve breakpoints appear at two consecutive points in the voltage curve graph, the acquisition module transparently copies the real-time voltage value, and if there is no result of the transparent copy or the real-time voltage value is less than the power-on threshold, the processing module determines that a power outage occurs in the power system;

[0035] If the real-time voltage value is greater than the power-on threshold, the processing module determines that no power outage occurs in the power system.

[0036] Furthermore, when the processing module determines that a power outage occurs in the power system, the processing module sends the power outage information to the power supply bureau.

[0037] The power outage information includes the power outage time, the power outage location and / or the real-time voltage value before the power outage.

[0038] In another aspect, the present invention provides a method for intelligently monitoring power outages in an electric power system, comprising:

[0039] The learner iteratively optimizes the voltage learning model according to the voltage training set until the loss value converges to the preset value;

[0040] Test the current voltage learning model through the voltage training set;

[0041] When the test results meet the preset test results, the voltage learning model parameters are saved.

[0042] Compared with the prior art, the present invention sets several acquisition modules for transparently copying real-time voltage values ​​and collecting minute-level high-frequency voltage acquisition data, a receiving module for receiving minute-level high-frequency voltage acquisition data, and performing pre-learning to generate corresponding pre-learning data, a learning module for selecting several indicator features for learning, and generating a corresponding voltage curve graph, and a processing module for receiving the voltage curve graph, and judging whether a power outage occurs in the power system according to whether a breakpoint appears in the voltage curve graph. If a power outage occurs, the power outage information is sent to the power supply bureau and the power system is repaired, thereby realizing intelligent monitoring and rapid response of power outages in the power system, reducing the impact of power outages on users in an automated manner, and improving the reliability and operation efficiency of the power system, helping power companies to better monitor and maintain the stable operation of the power grid.

[0043] Furthermore, by setting a timer and a converter in the acquisition module, the real-time voltage value is converted into minute-level high-frequency voltage acquisition data, thereby realizing high-frequency and real-time acquisition of the electric energy meter voltage. Through the periodic triggering of the timer, the data reading of the sensor and the data conversion of the converter, it is ensured that the system can obtain continuous and accurate voltage data, providing basic data support for the monitoring and analysis of the power system, and helping to promptly detect abnormal conditions in the power system, such as voltage fluctuations or power outages, thereby achieving rapid response and effective management.

[0044] Furthermore, by filtering the minute-level high-frequency voltage acquisition data in the receiving module and pre-learning the voltage filtering data to obtain the corresponding pre-learning data, the collected high-frequency voltage data is preliminarily processed and analyzed to provide high-quality data support for the monitoring and early warning of the power system, which helps to improve the response speed and accuracy of the system and ensure that timely measures can be taken when abnormalities occur in the power system.

[0045] Furthermore, by setting up learning modules, selecting key features, in-depth learning and data labeling, and conducting in-depth analysis of pre-learning data, a voltage curve graph that can reflect the changes in the power system state is generated, which helps to improve the system's prediction ability and response speed, and ensure that timely measures can be taken when an abnormality occurs in the power system. Through the training and application of the voltage learning model, the system can more intelligently identify and respond to power outages, thereby improving the stability and reliability of the power system.

[0046] Furthermore, by judging whether there are continuous voltage curve breakpoints in the voltage curve diagram and comparing the real-time voltage value with the power-on threshold, it is determined whether a power outage has occurred in the power system, thereby realizing real-time monitoring and intelligent judgment of the power system status, and improving the response speed and processing efficiency of power outage events in an automated way, which can not only reduce the impact of power outages on users, but also improve the reliability and safety of the power system. Through precise data analysis and intelligent judgment logic, the system can more effectively manage and maintain the power system and ensure the stability of power supply.

[0047] Furthermore, by training and testing the voltage training set, an optimized voltage learning model is obtained, which can ensure that the voltage learning model provides accurate power outage prediction and identification in practical applications, help improve the monitoring efficiency and response speed of the power system, reduce the impact of power outages on users, and improve the reliability and safety of power supply. Through continuous model optimization and verification, it can ensure that the monitoring system is always in the best state and adapt to possible changes in the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 A schematic diagram of the structure of a power outage intelligent monitoring system for a power system according to the present invention;

[0049] Figure 2 It is a structural schematic diagram of the acquisition module of the present invention;

[0050] Figure 3 It is a structural schematic diagram of a receiving module of the present invention;

[0051] Figure 4 It is a structural diagram of the learning module of the present invention;

[0052] Figure 5 The present invention is a flow chart of a method for intelligently monitoring power outages in an electric power system. DETAILED DESCRIPTION

[0053] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0054] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0055] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0056] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0057] See also Figure 1 As shown, it is a structural schematic diagram of a power outage intelligent monitoring system for a power system of the present invention, comprising:

[0058] Several acquisition modules, which are used to read real-time voltage values ​​and collect minute-level high-frequency voltage data;

[0059] A receiving module, which is connected to a plurality of acquisition modules, is used to receive minute-level high-frequency voltage acquisition data, and pre-learn the minute-level high-frequency voltage acquisition data to generate corresponding pre-learning data;

[0060] A learning module, which is connected to the receiving module, is used to receive pre-learning data, select several indicator features of the pre-learning data, learn the pre-learning data, and generate a corresponding voltage curve graph;

[0061] The processing module is connected to the acquisition module and the learning module respectively, and is used to receive the voltage curve graph, and judge whether a power outage occurs in the power system according to whether a breakpoint appears in the voltage curve graph. If a power outage occurs, the power outage information is sent to the power supply bureau and the power system is repaired;

[0062] Among them, pre-learning is to perform standardization processing based on minute-level high-frequency voltage collection data, and generate a number of pre-learning data with a learning rate of a standard learning rate, and generate corresponding indicator features based on each pre-learning data;

[0063] The standard learning rate is the learning rate that the learning module can recognize. It is inversely proportional to the resolution of the minute-level high-frequency voltage acquisition data. The higher the resolution, the lower the corresponding standard learning rate. For a single learning, the corresponding standard learning rate is a single learning rate.

[0064] Preferably, setting the learning rate to 700 / second has a better learning effect on the pre-learning data, and for the system described in this application, the corresponding analysis effect is optimal.

[0065] By setting up several acquisition modules, it is used to copy the real-time voltage value and collect minute-level high-frequency voltage acquisition data, a receiving module is used to receive minute-level high-frequency voltage acquisition data, and pre-learn to generate corresponding pre-learning data, a learning module is used to select several indicator features for learning, and generate a corresponding voltage curve graph, a processing module is used to receive the voltage curve graph, and judge whether the power system has a power outage based on whether there are breakpoints in the voltage curve graph. If a power outage occurs, the power outage information is sent to the power supply bureau and the power system is repaired. Intelligent monitoring and rapid response to power outages in the power system are realized, the impact of power outages on users is reduced through automation, and the reliability and operation efficiency of the power system are improved, which helps power companies better monitor and maintain the stable operation of the power grid.

[0066] See also Figure 2 As shown, it is a schematic diagram of the structure of the acquisition module of the present invention, including:

[0067] A timer, which has a fixed duration and is used to issue collection instructions at fixed intervals;

[0068] A high-frequency voltage sensor is connected to a timer and is arranged at a corresponding position of the electric energy meter to read the real-time voltage value displayed by the electric energy meter when receiving a collection instruction;

[0069] The converter is connected to the high-frequency voltage sensor and is used to convert the real-time voltage value into minute-level high-frequency voltage acquisition data.

[0070] In a specific implementation, the high-frequency voltage sensor is connected to the timer and installed at the corresponding position of the electric energy meter. When receiving the collection instruction sent by the timer, the high-frequency voltage sensor will read the real-time voltage value displayed by the electric energy meter.

[0071] By setting a timer and a converter in the acquisition module, the real-time voltage value is converted into minute-level high-frequency voltage acquisition data, thus realizing high-frequency and real-time acquisition of the electric energy meter voltage. Through the periodic triggering of the timer, the data reading of the sensor and the data conversion of the converter, it is ensured that the system can obtain continuous and accurate voltage data, providing basic data support for the monitoring and analysis of the power system, and helping to timely discover abnormal conditions in the power system, such as voltage fluctuations or power outages, so as to achieve rapid response and effective management.

[0072] See also Figure 3 As shown, it is a schematic diagram of the structure of the receiving module of the present invention, including:

[0073] A receiver, which is used to receive minute-level high-frequency voltage acquisition data;

[0074] A filter, which is connected to the receiver, is used to filter the minute-level high-frequency voltage acquisition data and generate corresponding voltage filtering data;

[0075] A pre-learner is used to pre-learn the voltage filtering data and generate corresponding pre-learning data.

[0076] Among them, a filter interval threshold is set in the filter;

[0077] The filtering interval threshold is an interval formed by the minimum and maximum values ​​of the pre-learner receiving range.

[0078] In the specific implementation, the main task of the receiver is to receive the minute-level high-frequency voltage acquisition data transmitted from the acquisition module. As the first station for data to flow into the receiving module, the receiver ensures the integrity and timeliness of the data, laying the foundation for subsequent data preprocessing and analysis. The filter is connected to the receiver, and its main task is to filter the received high-frequency voltage acquisition data, remove noise and outliers, and generate purer voltage filtered data. Through filtering, the quality of the data can be improved, and errors and interference in subsequent analysis can be reduced, ensuring that the pre-learner can receive more accurate data. The pre-learner pre-learns the voltage filtered data processed by the filter to generate pre-learning data.

[0079] The pre-learner may use some preliminary data analysis techniques, such as statistical analysis or simple machine learning algorithms, to identify trends and patterns in the data, providing a basis for subsequent in-depth learning.

[0080] The filter has a filter interval threshold, which defines the range of data received by the pre-learner.

[0081] Composition: The filter interval threshold is formed by the minimum and maximum values ​​of the pre-learner receiving range, ensuring that only data within this interval will be processed by the pre-learner.

[0082] Setting the filter interval threshold helps to further optimize the data, exclude extreme values ​​and irrelevant data, and enable the pre-learner to focus more on analyzing data within the normal voltage range.

[0083] By filtering the minute-level high-frequency voltage acquisition data in the receiving module and pre-learning the voltage filtering data to obtain the corresponding pre-learning data, the collected high-frequency voltage data is preliminarily processed and analyzed to provide high-quality data support for the monitoring and early warning of the power system, which helps to improve the response speed and accuracy of the system and ensure that timely measures can be taken when abnormalities occur in the power system.

[0084] See also Figure 4 As shown, it is a schematic diagram of the structure of the learning module of the present invention, including:

[0085] A selector, which is used to receive pre-learning data and select several indicator features of the pre-learning data;

[0086] A learner, which is connected to the selector, receives pre-learning data and indicator features, learns the pre-learning data, and generates a corresponding voltage curve graph;

[0087] A synthesizer, which is connected to the learner, is used to annotate the voltage curve according to the indicator characteristics.

[0088] Among them, a voltage learning model is provided in the learner;

[0089] The voltage learning model is trained and generated based on a voltage training set formed by pre-learning data corresponding to minute-level high-frequency voltage acquisition data.

[0090] In the specific implementation, the main task of the selector is to receive the pre-learning data and select several key indicator features from it. By selecting key features, the selector can identify the most important data points for the power system state analysis and provide a concise and effective data set for the subsequent learning process. The learner is connected to the selector and receives the pre-learning data and indicator features. Its main task is to conduct in-depth learning of the pre-learning data and generate a voltage curve. The learner may use complex machine learning algorithms, such as neural networks, decision trees, or support vector machines, to analyze patterns and trends in the data to predict the state changes of the power system. The synthesizer is connected to the learner and annotates the voltage curve according to the indicator features. Through annotation, the synthesizer can provide richer information to help operation and maintenance personnel better understand the meaning of the voltage curve and identify potential power outage risks. The voltage learning model is trained and generated according to the voltage training set formed by the pre-learning data corresponding to the minute-level high-frequency voltage acquisition data. It is the core of the learner. It improves the accuracy of the system's prediction of power outage events by training and learning the inherent laws of voltage data.

[0091] By setting up learning modules, selecting key features, in-depth learning and data labeling, and conducting in-depth analysis of pre-learning data, a voltage curve graph that can reflect changes in the power system state is generated, which helps to improve the system's prediction ability and response speed, and ensure that timely measures can be taken when an abnormality occurs in the power system. Through the training and application of the voltage learning model, the system can more intelligently identify and respond to power outages, thereby improving the stability and reliability of the power system.

[0092] Specifically, the high-frequency voltage sensor located at the corresponding position of the electric energy meter receives the collection instruction, copies the real-time voltage value, and transmits the real-time voltage value to the converter. The converter initializes the real-time voltage value to obtain the corresponding original voltage value, calibrates the original voltage value, and obtains the corresponding minute-level high-frequency voltage collection data.

[0093] Specifically, the filter removes the minute-level high-frequency voltage acquisition data that is not within the filter interval threshold and forms corresponding voltage filtered data. The pre-learner divides the voltage filtered data according to the standard learning rate based on the voltage training set to form corresponding pre-learning data.

[0094] Embodiment 1:

[0095] If the filter interval threshold is 210V-230V, and the A-minute high-frequency voltage collection data is 200V, which is less than the filter interval threshold, the A-minute high-frequency voltage collection data will be cleared.

[0096] Embodiment 2:

[0097] If the filter interval threshold is 210V-230V, and the B-minute high-frequency voltage collection data is 240V, which is greater than the filter interval threshold, the B-minute high-frequency voltage collection data will be cleared.

[0098] Embodiment 3:

[0099] If the filter interval threshold is 210V-230V, and the C-minute high-frequency voltage collection data is 220V, which is within the filter interval threshold, the C-minute high-frequency voltage collection data is retained and the corresponding voltage filtering data is formed.

[0100] In the specific implementation, the workflow of the filter is:

[0101] Set filter interval threshold: The filter first determines which data is valid based on the preset filter interval threshold (eg, 210V-230V).

[0102] Data removal: The filter removes data that is not within the filter interval threshold. This means that any data below the minimum threshold (210V) or above the maximum threshold (230V) will be excluded.

[0103] Forming voltage filtering data: After clearing, the remaining data (ie, data within the filter interval threshold) is retained and forms voltage filtering data, which will be used in the subsequent pre-learning process.

[0104] The workflow of the pre-learner:

[0105] Receiving voltage filtered data: The pre-learner receives the voltage filtered data after being processed by the filter.

[0106] Split by standard learning rate: The pre-learner splits the voltage filtered data into several pre-learning data according to the standard learning rate (e.g., processing 700 data points per second).

[0107] Forming pre-learning data: The segmented data is used as pre-learning data, which will be further used for learning and analysis by the learner in the learning module.

[0108] Specifically, the learning module selects several indicator features of the pre-learning data, the voltage learning model generates the corresponding voltage curve graph, the synthesizer marks the voltage curve graph based on the indicator features, and the processing module receives the voltage curve graph. If the voltage curve graph does not have continuous voltage curve breakpoints, it is determined that the power system has not experienced a power outage.

[0109] The indicator characteristics include voltage level and / or fixed duration.

[0110] Embodiment 4:

[0111] If in a voltage curve graph, the voltage is 220V at 1:00, the voltage is 0V at 1:01, and the voltage is 220V at 1:02, and there are no consecutive voltage curve breakpoints, it is determined that there is no power outage in the power system.

[0112] Specifically, if a voltage curve breakpoint appears at two consecutive points in the voltage curve graph, the acquisition module will transparently copy the real-time voltage value. If there is no result from the transparent copy or the real-time voltage value is less than the power-on threshold, the processing module will determine that a power outage has occurred in the power system;

[0113] If the real-time voltage value is greater than the power-on threshold, the processing module determines that no power outage has occurred in the power system.

[0114] Embodiment 5:

[0115] If a voltage curve shows that the voltage is 220V at 1:00, 0V at 1:01, and 0V at 1:02, and there are voltage curve breakpoints at two consecutive points, the acquisition module will read the real-time voltage value. If there is no result, the processing module will determine that a power outage has occurred in the power system.

[0116] Embodiment 6:

[0117] If the power-on threshold is 220V, a voltage curve Figure 1 At 1:00, the voltage is 220V, at 1:01, the voltage is 0V, and at 1:02, the voltage is 0V. There are two consecutive voltage curve breakpoints. The acquisition module transparently copies the real-time voltage value. If the transparently copied real-time voltage value is 110V, and the real-time voltage value is less than the power-on threshold, the processing module determines that a power outage has occurred in the power system.

[0118] Embodiment 7:

[0119] If the power-on threshold is 220V, a voltage curve Figure 1At 1:00, the voltage is 220V, at 1:01, the voltage is 0V, and at 1:02, the voltage is 0V. There are two consecutive voltage curve breakpoints. The acquisition module transparently copies the real-time voltage value. If the real-time voltage value copied is 221V, and the real-time voltage value is greater than the power-on threshold, the processing module determines that there is no power outage in the power system.

[0120] In the specific implementation, the workflow of the learning module is:

[0121] Selecting indicator features: The learning module selects key indicator features from the pre-learning data. These features may include the voltage level and / or the voltage value within a fixed time period.

[0122] Generate voltage curve graphs: The voltage learning model generates voltage curve graphs based on these indicator features. These graphs reflect the changes in power system voltage over time.

[0123] Annotate voltage graphs: The synthesizer annotates voltage graphs based on indicator features to more intuitively identify voltage changes and potential power outages.

[0124] Processing module workflow:

[0125] Receiving voltage curve graph: The processing module receives the voltage curve graph generated by the learning module.

[0126] Determine power outage: The processing module checks whether the voltage curve graph has continuous voltage curve breakpoints to determine whether a power outage has occurred in the power system.

[0127] Through these examples, we can see how the system can intelligently determine whether a power outage has occurred in the power system by detecting the changes in the voltage curve and the real-time voltage value. This method improves the accuracy and response speed of power outage detection, helps to take emergency repair measures in a timely manner, and reduces the impact of power outages on users.

[0128] Specifically, when the processing module determines that a power outage occurs in the power system, it sends a power outage message to the power supply bureau.

[0129] The power outage information includes the power outage time, the power outage location and / or the real-time voltage value before the power outage.

[0130] By judging whether there are continuous voltage curve breakpoints in the voltage curve graph and comparing the real-time voltage value with the power-on threshold, it is determined whether a power outage has occurred in the power system, thereby realizing real-time monitoring and intelligent judgment of the power system status, and improving the response speed and processing efficiency of power outage events through automation, which can not only reduce the impact of power outages on users, but also improve the reliability and safety of the power system. Through precise data analysis and intelligent judgment logic, the system can more effectively manage and maintain the power system and ensure the stability of power supply.

[0131] In the specific implementation, the processing module:

[0132] Power outage judgment: Analyze the voltage curve and real-time voltage value to judge whether a power outage has occurred in the power system. This is achieved by checking whether there are continuous breakpoints in the voltage curve and comparing the real-time voltage value with the power-on threshold.

[0133] Information transmission: Once a power outage is detected, the processing module will send detailed power outage information to the power supply bureau. This information includes:

[0134] Power outage time: Recording the specific time when the power outage occurs helps the power supply bureau respond quickly.

[0135] Power outage location: Provides the geographical location where the power outage occurred to facilitate positioning and emergency repairs.

[0136] Real-time voltage value before power outage: Provides the voltage reading at the last moment before power outage, which helps to analyze the cause of power outage.

[0137] See also Figure 5 As shown, it is a flow chart of a power outage intelligent monitoring method for a power system of the present invention, comprising:

[0138] Step S1, the learner iteratively optimizes the voltage learning model according to the voltage training set until the loss value converges to a preset value;

[0139] Step S2, testing the current voltage learning model through the voltage training set;

[0140] Step S3: When the test result meets the preset test result, the voltage learning model parameters are saved.

[0141] Step S1: The learner iteratively optimizes the voltage learning model based on the voltage training set

[0142] In a specific implementation, the learner iteratively trains the voltage learning model using the voltage training set. This process usually involves adjusting the model parameters to minimize a loss function such as mean square error or cross entropy.

[0143] Loss value convergence: Iterative training will continue until the model's loss value converges to a preset value, which means that the model's performance has reached a stable state and can better fit the training data.

[0144] The current voltage learning model is tested through the voltage training set. After the model is trained to a certain extent, the current voltage learning model is tested with an independent voltage test set to evaluate the generalization ability of the model. The purpose of evaluating the performance is to ensure that the model can also show good performance on unseen data, avoid overfitting, and ensure the practicality and accuracy of the model. When the test results meet the preset test results, the voltage learning model parameters are saved. If the test results meet the preset performance indicators (such as accuracy, recall rate, etc.), it is considered that the model has been trained well enough and can be used for actual power system monitoring. When the test results meet the preset conditions, the system will save the parameters of the voltage learning model, which will be used for subsequent power system monitoring and prediction.

[0145] This method embodies the application of machine learning in power system monitoring. Through automation and intelligent means, it improves the management and maintenance level of the power system and ensures the stability and continuity of power supply.

[0146] By training and testing the voltage training set, an optimized voltage learning model is obtained, which can ensure that the voltage learning model can provide accurate power outage prediction and identification in practical applications, help improve the monitoring efficiency and response speed of the power system, reduce the impact of power outages on users, and improve the reliability and safety of power supply. Through continuous model optimization and verification, it can ensure that the monitoring system is always in the best state and adapt to possible changes in the power system.

[0147] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0148] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A power outage intelligent monitoring system for a power system, characterized in that: include: Several acquisition modules, which are used to read real-time voltage values ​​and collect minute-level high-frequency voltage data; A receiving module, which is connected to the plurality of acquisition modules, and is used to receive the minute-level high-frequency voltage acquisition data, and pre-learn the minute-level high-frequency voltage acquisition data to generate corresponding pre-learning data; A learning module, connected to the receiving module, for receiving the pre-learning data, selecting several indicator features of the pre-learning data, learning the pre-learning data, and generating a corresponding voltage curve graph; a processing module, which is connected to the acquisition module and the learning module respectively, and is used to receive the voltage curve graph, and judge whether a power outage occurs in the power system according to whether a breakpoint appears in the voltage curve graph; if a power outage occurs, the power outage information is sent to the power supply bureau and the power system is repaired; The pre-learning is to perform standardization processing on the minute-level high-frequency voltage acquisition data, generate a number of pre-learning data with a learning rate of a standard learning rate, and generate corresponding indicator features according to each pre-learning data; The standard learning rate is a learning rate that the learning module can recognize, and is inversely proportional to the resolution of the minute-level high-frequency voltage acquisition data. The higher the resolution, the lower the corresponding standard learning rate. For a single learning, the corresponding standard learning rate is a single learning rate.

2. The power outage intelligent monitoring system of the power system according to claim 1 is characterized in that: The acquisition module comprises: A timer, which is set with a fixed duration and is used to issue a collection instruction at intervals of the fixed duration; A high-frequency voltage sensor, which is connected to the timer and is arranged at a corresponding position of the electric energy meter, and is used to read the real-time voltage value displayed by the electric energy meter when receiving the acquisition instruction; A converter is connected to the high-frequency voltage sensor and is used to convert the real-time voltage value into the minute-level high-frequency voltage acquisition data.

3. The power outage intelligent monitoring system of the power system according to claim 2 is characterized in that: The receiving module comprises: A receiver, which is used to receive the minute-level high-frequency voltage acquisition data; A filter connected to the receiver, for filtering the minute-level high-frequency voltage acquisition data and generating corresponding voltage filtering data; A pre-learner, which is used to pre-learn the voltage filtering data to generate corresponding pre-learning data, Wherein, a filter interval threshold is provided in the filter; The filtering interval threshold is an interval formed by the minimum value and the maximum value of the receiving range of the pre-learner.

4. The power outage intelligent monitoring system of the power system according to claim 3 is characterized in that: The learning module includes: A selector, which is used to receive the pre-learning data and select a number of indicator features of the pre-learning data; A learner, connected to the selector, receives the pre-learning data and the indicator feature, and learns the pre-learning data to generate a corresponding voltage curve graph; a synthesizer, connected to the learner, for marking the voltage curve graph according to the indicator characteristics, Wherein, the learner is provided with a voltage learning model; The voltage learning model is generated through training based on a voltage training set formed by pre-learning data corresponding to the minute-level high-frequency voltage acquisition data.

5. The power outage intelligent monitoring system of the power system according to claim 4 is characterized in that: The high-frequency voltage sensor located at the corresponding position of the electric energy meter receives the collection instruction, copies the real-time voltage value, and transmits the real-time voltage value to the converter. The converter initializes the real-time voltage value to obtain the corresponding original voltage value, performs data calibration on the original voltage value, and obtains the corresponding minute-level high-frequency voltage collection data.

6. The power outage intelligent monitoring system of the power system according to claim 5 is characterized in that: The filter removes the minute-level high-frequency voltage acquisition data that is not within the filter interval threshold in the minute-level high-frequency voltage acquisition data and forms corresponding voltage filtering data. The pre-learner divides the voltage filtering data according to the standard learning rate based on the voltage training set to form corresponding pre-learning data.

7. The power outage intelligent monitoring system of the power system according to claim 6 is characterized in that: The learning module selects several indicator features of the pre-learning data, the voltage learning model generates a corresponding voltage curve graph, the synthesizer marks the voltage curve graph in combination with the indicator features, the processing module receives the voltage curve graph, and if the voltage curve graph does not have continuous voltage curve breakpoints, it is determined that the power system has not experienced a power outage, The indicator characteristics include voltage level and / or fixed duration.

8. The power outage intelligent monitoring system of the power system according to claim 7 is characterized in that: If voltage curve breakpoints appear at two consecutive points in the voltage curve graph, the acquisition module transparently copies the real-time voltage value. If there is no result in the transparent copy or the real-time voltage value is less than the power-on threshold, the processing module determines that a power outage occurs in the power system; If the real-time voltage value is greater than the power-on threshold, the processing module determines that no power outage occurs in the power system.

9. The power outage intelligent monitoring system of the power system according to claim 8, characterized in that: When the processing module determines that a power outage occurs in the power system, the processing module sends the power outage information to the power supply bureau. The power outage information includes the power outage time, the power outage location and / or the real-time voltage value before the power outage.

10. A method for intelligently monitoring power outages in an electric power system according to any one of claims 1 to 9, characterized in that: include: The learner iteratively optimizes the voltage learning model according to the voltage training set until the loss value converges to the preset value; Test the current voltage learning model through the voltage training set; When the test results meet the preset test results, the voltage learning model parameters are saved.

Images