Unattended substation monitoring system

Abstract

The invention discloses an unattended transformer substation monitoring system, and relates to the technical field of transformer substation monitoring. Comprising a monitoring area construction unit for performing three-dimensional modeling and area division on a transformer substation, an environment data acquisition unit for acquiring environment data and processing the environment data into environment characteristic values, a power grid data acquisition unit for acquiring power grid parameter data, a video data acquisition unit for acquiring video data, and a monitoring area state construction unit for storing state events, a monitoring area operation condition unit storing operation conditions; and an operation execution unit storing execution events. When the unattended substation is monitored, whether abnormity occurs or not is judged by acquiring the state event corresponding to the single monitoring area, if yes, the specific monitoring area can be quickly positioned, the corresponding operation condition is acquired, and the state event and the operation condition of the abnormal monitoring area are utilized to match and execute the corresponding execution event, so that the unattended substation is monitored. And the monitoring efficiency of the unattended substation is greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of substation monitoring, and in particular to an unattended substation monitoring system. Background Art

[0002] With the continuous expansion of power network capacity, the scale and number of substations are also growing, and the maintenance workload of the monitoring system is also increasing, but the number of maintenance personnel has not increased accordingly. In order to improve the economic benefits and labor productivity of the enterprise, an unattended mode is adopted in the daily production of the substation. By sending alarm information to the monitoring center, remote and real-time video monitoring of the substation, remote fault and accident alarm reception and processing are realized, thereby improving the safety and reliability of substation operation.

[0003] However, restricted by technological development, many video parts in the current substation monitoring system can only provide simple on-site monitoring and other functions. The information that electronic induction testers can provide is relatively limited, and most video and inductance systems are independent. Although the local automation level is constantly improving, the global information is not perfect, and the sharing ability is weak and the linkage mechanism is lacking, which leads to the fragmentation and localization of the substation systems, greatly affecting the monitoring efficiency of the substation. Summary of the invention

[0004] In order to solve the above-mentioned deficiencies in the prior art, the present invention provides an unattended substation monitoring system, comprising: A monitoring area construction unit: obtains a three-dimensional point cloud of a substation area collected by a laser detector and a natural language description of characteristics of the substation area input by a user to obtain a first acquisition result; based on the first acquisition result, constructs a power grid space model of the substation area, and spatially divides the power grid space model to obtain a plurality of monitoring areas; Environmental data collection unit: collects environmental data of each monitoring area and performs preset processing to obtain environmental characteristic values, wherein the environmental data includes temperature data, humidity data and monitoring gas data; A power grid data collection unit: collects power grid parameter data of each monitoring area and processes it using a first method to obtain a first processing result; Video data acquisition unit: collects video data of each monitoring area and processes it using a second method to obtain a second processing result; A monitoring area state construction unit: based on the environmental characteristic value of each monitoring area, the first processing result and the second processing result, obtains the state event of each monitoring area and stores it in the first set; Monitoring area operating condition unit: obtains the operating condition of each monitoring area and stores it in the second set; The monitoring area operation execution unit obtains multiple execution events and stores them in a third set; based on the first set and the second set, matches and executes the execution event of any monitoring area in the third set.

[0005] The present invention is achieved through the following technical solutions: the entire unmanned substation is modeled and divided into areas through a monitoring area construction unit, specifically through the three-dimensional point cloud collected by a laser detector and the natural language description of the features input by the user, to build a three-dimensional model covering the entire substation, and based on the three-dimensional model, the entire substation is divided into several monitoring areas. When an abnormality is found in a certain area of ​​the substation, it can be quickly located. The environmental data acquisition unit collects environmental data (and performs preset processing to obtain environmental characteristic values), the power grid data acquisition unit collects power grid parameter data, and the video data acquisition unit collects video data. For a single monitoring area, multi-dimensional monitoring of environmental monitoring, power grid monitoring and video monitoring is established, and the environmental characteristic values, the first processing results and the second processing results under multi-dimensional monitoring are integrated. The integrated results are stored by the monitoring area state construction unit to generate a state event of a single monitoring area; the operating conditions of each monitoring area (such as switches, alarms, circuit breakers, fire extinguishing systems, etc.) are obtained and stored by the monitoring area operation condition unit; multiple execution events (such as closing switches, activating alarms, activating fire extinguishing systems, etc.) are obtained and stored by the monitoring area operation execution unit, and then the state events and corresponding operating conditions of a single monitoring area are used. When an abnormality occurs in a single monitoring area, the corresponding execution events can be immediately matched and executed, thereby greatly improving the monitoring efficiency of the entire unmanned substation.

[0006] As an optional technical solution, based on the first acquisition result, constructing the power grid space model of the substation area includes: Based on the three-dimensional point cloud, extracting a three-dimensional feature vector of the substation area; Performing semantic coding on the characteristic natural language description of the substation area to obtain a corresponding semantic coding feature vector; Based on the three-dimensional feature vector and the semantic coding feature vector, joint optimization is performed to obtain the power grid space model.

[0007] As an optional technical solution, environmental data of each monitoring area is collected and processed by preset methods to obtain environmental characteristic values ​​including: The first formula is used to calculate the environmental characteristic value of each monitoring area, specifically: is the environmental characteristic value of a single monitoring area, is the temperature characteristic value of a single monitoring area, is the humidity characteristic value of a single monitoring area, is the characteristic value of any monitored gas in a single monitoring area, is the number of all monitored gases in a single monitoring area, is the influence coefficient of temperature in a single monitoring area, is the influence coefficient of humidity in a single monitoring area, is the influence coefficient of gas in a single monitoring area.

[0008] As an optional technical solution, the temperature characteristic value, humidity characteristic value and characteristic value of any gas in a single monitoring area are obtained in the following ways: Acquire any environmental data in a single monitoring area, recorded as first environmental data, where the first environmental data is the temperature data, or the humidity data, or any sub-monitoring gas data in the monitoring gas data; Acquire historical data of the first environmental data to obtain a second acquisition result; Based on the second acquisition result, the first environmental data is divided into data intervals, and a value is assigned to each divided data interval; Based on the assigned data interval, a characteristic value corresponding to the first environmental data is obtained.

[0009] As an optional technical solution, in a single monitoring area , and Ways to obtain include: Acquire environmental data of multiple time scales and status data of corresponding monitoring areas, build a weight model based on analyzing the impact of different environmental data on the status of monitoring areas, and train the weight model. Based on the trained weight model, obtain the status of a single monitoring area. , and .

[0010] As an optional technical solution, the first method includes: De-noising and normalizing the power grid parameter data to obtain standard power grid parameter data; Performing time-frequency transformation on the standard power grid parameter data to obtain a corresponding two-dimensional time-frequency diagram of the signal; Extracting time-frequency features from the two-dimensional time-frequency graph of the signal to obtain a signal time-frequency correlation feature graph; Performing dimensionality reduction processing on the signal time-frequency correlation feature graph; A preset clustering algorithm is used to analyze and process the signal time-frequency correlation feature graph after dimensionality reduction processing, and multiple evaluation indicators are preset; Based on all evaluation indicators, output corresponding analysis results; Based on the analysis results, determine whether a fault occurs in the corresponding monitoring area. If so, output the corresponding fault type. If not, output empty.

[0011] As an optional technical solution, the second method includes: Acquiring video data of a single monitoring area to obtain a third acquisition result; Based on the third acquisition result, a mixed Gaussian model with high and low update rates is used to generate a background image with a high update rate, which is recorded as a first background image, and a background image with a low update rate, which is recorded as a second background image; The probability values ​​of the first background image and the second background image are calculated using the second formula, specifically: is the probability value, which is 0 or 1; is a sequence of pixel values ​​in the first background image or the second background image; for time point; is the number of mixed Gaussian models; is the weight value of the nth mixed Gaussian model; is the variance of the n mixed Gaussian models; is the average value of n mixed Gaussian models; If the probability value of the first background image is 1 and the probability value of the second background image is 1, it is determined that there is a moving target in the corresponding background, and a target type of the corresponding moving target is obtained by a preset method; If the probability value of the first background image is 0 and the probability value of the second background image is 1, it is determined that there is a stationary target in the corresponding background; If the probability value of the first background image is 1, and the probability value of the second background image is 0, it is determined that random noise exists in the corresponding background; If the probability value of the first background image is 0, and the probability value of the second background image is 0, it is determined that the corresponding background is a blank background.

[0012] As an optional technical solution, the preset method includes: Acquire multiple target contour data through offline experiments and establish a class contour database; Based on the first background image and the second background image, the target contour of the moving target is extracted, and matched one by one with the contours in the class contour database to obtain the target type of the corresponding moving target.

[0013] As an optional technical solution, based on the environmental characteristic value of each monitoring area, the first processing result and the second processing result, obtaining the state event of each monitoring area and storing it in the first set includes: For the first set, is the event corresponding to the environmental characteristic value of a single monitoring area, The event corresponding to the first processing result of a single monitoring area, The event corresponding to the second processing result of a single monitoring area; Obtaining the operating condition of each monitoring area and storing it in the second set includes: For the second set, operating conditions for a single monitoring area; Acquiring multiple execution events and storing them in a third set includes: For the third set, For a single execution event.

[0014] As an optional technical solution, based on the first set and the second set, matching the execution event of any monitoring area in the third set and executing includes: Based on the first set, a status event of a single monitoring area is obtained, recorded as a first status event; Based on the second set, obtaining an operating condition of a single monitoring area, recorded as a first operating condition; Based on the first state event and the first operating condition, an execution event of a single monitoring area is matched in the third set, recorded as a first execution event and executed.

[0015] One or more technical solutions provided by the present invention have at least the following technical effects or advantages: The present invention provides an unmanned substation monitoring system, which performs three-dimensional modeling and monitoring area division on the entire substation through a monitoring area construction unit, thereby dividing the monitoring of the substation into monitoring of multiple monitoring areas; constructs multi-dimensional monitoring of a single monitoring area through an environmental data acquisition unit, a power grid data acquisition unit and a video data acquisition unit, and integrates and processes the data of all the above acquisition units to generate corresponding status events; stores the status events of a single monitoring area through a monitoring area status construction unit; stores the operating conditions of a single monitoring area through a monitoring area operating condition unit; stores multiple execution events through a monitoring area operation execution unit; when monitoring the entire substation, determines whether an abnormality occurs by obtaining the corresponding status events of a single monitoring area, and if an abnormality occurs, can quickly locate the specific monitoring area, and obtain the corresponding operating conditions, and use the status events and operating conditions of the abnormal monitoring area to match the corresponding execution events in the third set and execute them, thereby greatly improving the monitoring efficiency of the unmanned substation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the embodiments of the present invention, constitute a part of the present invention, and do not constitute a limitation on the embodiments of the present invention; Figure 1 It is a schematic diagram of the composition of an unattended substation monitoring system in the present invention. DETAILED DESCRIPTION

[0017] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0018] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those within the scope of this description. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0019] Example Please refer to Figure 1 , Figure 1 This is a schematic diagram of the composition of an unattended substation monitoring system in the present invention, including: A monitoring area construction unit: obtains a three-dimensional point cloud of a substation area collected by a laser detector and a natural language description of characteristics of the substation area input by a user to obtain a first acquisition result; based on the first acquisition result, constructs a power grid space model of the substation area, and spatially divides the power grid space model to obtain a plurality of monitoring areas; Environmental data collection unit: collects environmental data of each monitoring area and performs preset processing to obtain environmental characteristic values, wherein the environmental data includes temperature data, humidity data and monitoring gas data; A power grid data collection unit: collects power grid parameter data of each monitoring area and processes it using a first method to obtain a first processing result; Video data acquisition unit: collects video data of each monitoring area and processes it using a second method to obtain a second processing result; A monitoring area state construction unit: based on the environmental characteristic value of each monitoring area, the first processing result and the second processing result, obtains the state event of each monitoring area and stores it in the first set; Monitoring area operating condition unit: obtains the operating condition of each monitoring area and stores it in the second set; The monitoring area operation execution unit obtains multiple execution events and stores them in a third set; based on the first set and the second set, matches and executes the execution event of any monitoring area in the third set.

[0020] The present invention is achieved through the following embodiments: Monitoring area building unit: Acquire a three-dimensional point cloud of the substation area collected by a laser detector and a natural language description of the characteristics of the substation area input by a user to obtain a first acquisition result; based on the first acquisition result, construct a power grid space model of the substation area, and spatially divide the power grid space model to obtain a number of monitoring areas.

[0021] Further, based on the first acquisition result, constructing a power grid space model of the substation area includes: Based on the three-dimensional point cloud, extracting a three-dimensional feature vector of the substation area; Performing semantic coding on the characteristic natural language description of the substation area to obtain a corresponding semantic coding feature vector; Based on the three-dimensional feature vector and the semantic coding feature vector, joint optimization is performed to obtain the power grid space model.

[0022] Among them, through the three-dimensional point cloud of the substation area collected by the laser detector (for common component features, such as plates, beams and columns, etc.), and the natural language description of the features input by the user (for uncommon component features); based on the three-dimensional point cloud, the corresponding three-dimensional feature vector is extracted, the natural language description input by the user is semantically encoded, and the corresponding semantic encoding feature vector is generated, the three-dimensional feature vector and the language encoding feature vector are jointly optimized to generate a power grid space model; spatial division is performed in the power grid space model to obtain multiple monitoring areas, and the monitoring of the unmanned substation is converted into the monitoring of multiple monitoring areas.

[0023] Environmental data collection unit: The environmental data of each monitoring area is collected and pre-processed to obtain environmental characteristic values, wherein the environmental data includes temperature data, humidity data and monitoring gas data.

[0024] Furthermore, the environmental data of each monitoring area is collected and processed by preset methods to obtain environmental characteristic values ​​including: The first formula is used to calculate the environmental characteristic value of each monitoring area, specifically: is the environmental characteristic value of a single monitoring area, is the temperature characteristic value of a single monitoring area, is the humidity characteristic value of a single monitoring area, is the characteristic value of any monitored gas in a single monitoring area, is the number of all monitored gases in a single monitoring area, is the influence coefficient of temperature in a single monitoring area, is the influence coefficient of humidity in a single monitoring area, is the influence coefficient of gas in a single monitoring area.

[0025] Furthermore, the temperature characteristic value, humidity characteristic value and characteristic value of any gas in a single monitoring area are obtained by: Acquire any environmental data in a single monitoring area, recorded as first environmental data, where the first environmental data is the temperature data, or the humidity data, or any sub-monitoring gas data in the monitoring gas data; Acquire historical data of the first environmental data to obtain a second acquisition result; Based on the second acquisition result, the first environmental data is divided into data intervals, and a value is assigned to each divided data interval; Based on the assigned data interval, a characteristic value corresponding to the first environmental data is obtained.

[0026] Furthermore, in a single monitoring area , and Ways to obtain include: Acquire environmental data of multiple time scales and status data of corresponding monitoring areas, build a weight model based on analyzing the impact of different environmental data on the status of monitoring areas, and train the weight model. Based on the trained weight model, obtain the status of a single monitoring area. , and .

[0027] Among them, the environmental data of each monitoring area is collected by the environmental data collection unit and the corresponding environmental characteristic value is calculated based on the first formula. The environmental data includes temperature data, humidity data and gas data (such as hydrogen sulfide, hydrogen monoxide, methane and hydrogen, etc.). The environmental characteristic value is used to unify the environmental data of a single monitoring area, which can help to quickly match and execute execution events later.

[0028] Power grid data acquisition unit: The power grid parameter data of each monitoring area is collected and processed using a first method to obtain a first processing result.

[0029] Furthermore, the first method comprises: De-noising and normalizing the power grid parameter data to obtain standard power grid parameter data; Performing time-frequency transformation on the standard power grid parameter data to obtain a corresponding two-dimensional time-frequency diagram of the signal; Extracting time-frequency features from the two-dimensional time-frequency graph of the signal to obtain a signal time-frequency correlation feature graph; Performing dimensionality reduction processing on the signal time-frequency correlation feature graph; A preset clustering algorithm is used to analyze and process the signal time-frequency correlation feature graph after dimensionality reduction processing, and multiple evaluation indicators are preset; Based on all evaluation indicators, output corresponding analysis results; Based on the analysis results, determine whether a fault occurs in the corresponding monitoring area. If so, output the corresponding fault type. If not, output empty.

[0030] Video data acquisition unit: collects video data of each monitoring area and processes it using the second method to obtain a second processing result.

[0031] Among them, the power grid parameter data is obtained and preprocessed, including denoising and normalization, and then time-frequency transformation is performed to obtain the corresponding signal two-dimensional time-frequency diagram, and feature extraction is performed on the signal two-dimensional time-frequency diagram to obtain the signal time-frequency correlation feature diagram, and the signal time-frequency correlation feature diagram is subjected to dimensionality reduction processing (t-SNE or UMAP algorithm), and then a preset clustering algorithm (such as K-means or DBSCAN) is used to analyze and process the time-frequency correlation feature diagram after dimensionality reduction processing, and multiple evaluation indicators (such as silhouette coefficient and Davies-Bouldin index, etc.) are preset at the same time, and based on all evaluation indicators, the analysis results are output, and through the analysis results, it is judged whether a fault occurs in the monitoring area, and if so, the corresponding fault type is output, if not, the output is empty.

[0032] Furthermore, the second method comprises: Acquiring video data of a single monitoring area to obtain a third acquisition result; Based on the third acquisition result, a mixed Gaussian model with high and low update rates is used to generate a background image with a high update rate, which is recorded as a first background image, and a background image with a low update rate, which is recorded as a second background image; The probability values ​​of the first background image and the second background image are calculated using the second formula, specifically: is the probability value, which is 0 or 1; is a sequence of pixel values ​​in the first background image or the second background image; for time point; is the number of mixed Gaussian models; is the weight value of the nth mixed Gaussian model; is the variance of the n mixed Gaussian models; is the average value of n mixed Gaussian models; If the probability value of the first background image is 1 and the probability value of the second background image is 1, it is determined that there is a moving target in the corresponding background, and a target type of the corresponding moving target is obtained by a preset method; If the probability value of the first background image is 0 and the probability value of the second background image is 1, it is determined that there is a stationary target in the corresponding background; If the probability value of the first background image is 1, and the probability value of the second background image is 0, it is determined that random noise exists in the corresponding background; If the probability value of the first background image is 0, and the probability value of the second background image is 0, it is determined that the corresponding background is a blank background.

[0033] Furthermore, the preset method includes: Acquire multiple target contour data through offline experiments and establish a class contour database; Based on the first background image and the second background image, the target contour of the moving target is extracted, and matched one by one with the contours in the class contour database to obtain the target type of the corresponding moving target.

[0034] Among them, with regard to video monitoring, this embodiment uses a mixed Gaussian model based on high and low update rates to process the video, and uses the mixed Gaussian model to generate foreground binary images with high and low update rates for the same background, which are the first background image and the second background image, respectively. Then, the second formula is used to calculate the probability values ​​of the first background image and the second background image, respectively. By analyzing the probability value distribution of the first background image and the second background image, it is determined whether there is a moving target (i.e., illegal intrusion). If it is determined that there is a moving target in the background, contour matching is performed through the class contour database to output the corresponding target type.

[0035] Monitoring area status building unit: Based on the environmental characteristic value of each monitoring area, the first processing result and the second processing result, the status event of each monitoring area is obtained and stored in the first set.

[0036] Further, based on the environmental characteristic value of each monitoring area, the first processing result and the second processing result, obtaining the state event of each monitoring area and storing it in the first set includes: For the first set, is the event corresponding to the environmental characteristic value of a single monitoring area, The event corresponding to the first processing result of a single monitoring area, The event corresponding to the second processing result of a single monitoring area.

[0037] Monitoring area operating condition unit: The operating conditions of each monitoring area are obtained and stored in the second set.

[0038] Further, obtaining the operating condition of each monitoring area and storing it in the second set includes: For the second set, is the operating condition for a single monitoring area.

[0039] Monitoring area operation execution unit: Acquire multiple execution events and store them in a third set; based on the first set and the second set, match and execute the execution events of any monitoring area in the third set.

[0040] Further, obtaining multiple execution events and storing them in the third set includes: For the third set, For a single execution event.

[0041] Further, based on the first set and the second set, matching the execution event of any monitoring area in the third set and executing includes: Based on the first set, a status event of a single monitoring area is obtained, recorded as a first status event; Based on the second set, obtaining an operating condition of a single monitoring area, recorded as a first operating condition; Based on the first state event and the first operating condition, an execution event of a single monitoring area is matched in the third set, recorded as a first execution event and executed.

[0042] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0043] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An unattended substation monitoring system, characterized in that: include: A monitoring area construction unit: obtains a three-dimensional point cloud of a substation area collected by a laser detector and a natural language description of characteristics of the substation area input by a user to obtain a first acquisition result; based on the first acquisition result, constructs a power grid space model of the substation area, and spatially divides the power grid space model to obtain a plurality of monitoring areas; Environmental data collection unit: collects environmental data of each monitoring area and performs preset processing to obtain environmental characteristic values, wherein the environmental data includes temperature data, humidity data and monitoring gas data; A power grid data collection unit: collects power grid parameter data of each monitoring area and processes it using a first method to obtain a first processing result; Video data acquisition unit: collects video data of each monitoring area and processes it using a second method to obtain a second processing result; A monitoring area state construction unit: based on the environmental characteristic value of each monitoring area, the first processing result and the second processing result, obtains the state event of each monitoring area and stores it in the first set; Monitoring area operating condition unit: obtains the operating condition of each monitoring area and stores it in the second set; The monitoring area operation execution unit obtains multiple execution events and stores them in a third set; based on the first set and the second set, matches and executes the execution event of any monitoring area in the third set.

2. The unattended substation monitoring system according to claim 1, characterized in that: Based on the first acquisition result, constructing the power grid space model of the substation area includes: Based on the three-dimensional point cloud, extracting a three-dimensional feature vector of the substation area; Performing semantic coding on the characteristic natural language description of the substation area to obtain a corresponding semantic coding feature vector; Based on the three-dimensional feature vector and the semantic coding feature vector, joint optimization is performed to obtain the power grid space model.

3. The unattended substation monitoring system according to claim 1, characterized in that: Collect environmental data of each monitoring area and perform preset processing to obtain environmental characteristic values ​​including: The first formula is used to calculate the environmental characteristic value of each monitoring area, specifically: is the environmental characteristic value of a single monitoring area, is the temperature characteristic value of a single monitoring area, is the humidity characteristic value of a single monitoring area, is the characteristic value of any monitored gas in a single monitoring area, is the number of all monitored gases in a single monitoring area, is the influence coefficient of temperature in a single monitoring area, is the influence coefficient of humidity in a single monitoring area, is the influence coefficient of gas in a single monitoring area.

4. The unattended substation monitoring system according to claim 3 is characterized in that: The temperature characteristic value, humidity characteristic value and characteristic value of any gas in a single monitoring area are obtained by: Acquire any environmental data in a single monitoring area, recorded as first environmental data, where the first environmental data is the temperature data, or the humidity data, or any sub-monitoring gas data in the monitoring gas data; Acquire historical data of the first environmental data to obtain a second acquisition result; Based on the second acquisition result, the first environmental data is divided into data intervals, and a value is assigned to each divided data interval; Based on the assigned data interval, a characteristic value corresponding to the first environmental data is obtained.

5. The unattended substation monitoring system according to claim 3, characterized in that: In a single monitoring area , and Ways to obtain include: Acquire environmental data of multiple time scales and status data of corresponding monitoring areas, build a weight model based on analyzing the impact of different environmental data on the status of monitoring areas, and train the weight model. Based on the trained weight model, obtain the status of a single monitoring area. , and .

6. The unattended substation monitoring system according to claim 1, characterized in that: The first method comprises: De-noising and normalizing the power grid parameter data to obtain standard power grid parameter data; Performing time-frequency transformation on the standard power grid parameter data to obtain a corresponding two-dimensional time-frequency diagram of the signal; Extracting time-frequency features from the two-dimensional time-frequency graph of the signal to obtain a signal time-frequency correlation feature graph; Performing dimensionality reduction processing on the signal time-frequency correlation feature graph; A preset clustering algorithm is used to analyze and process the signal time-frequency correlation feature graph after dimensionality reduction processing, and multiple evaluation indicators are preset; Based on all evaluation indicators, output corresponding analysis results; Based on the analysis results, determine whether a fault occurs in the corresponding monitoring area. If so, output the corresponding fault type. If not, output empty.

7. The unattended substation monitoring system according to claim 1, characterized in that: The second method comprises: Acquiring video data of a single monitoring area to obtain a third acquisition result; Based on the third acquisition result, a mixed Gaussian model with high and low update rates is used to generate a background image with a high update rate, which is recorded as a first background image, and a background image with a low update rate, which is recorded as a second background image; The probability values ​​of the first background image and the second background image are calculated using the second formula, specifically: is the probability value, which is 0 or 1; is a sequence of pixel values ​​in the first background image or the second background image; for time point; is the number of mixed Gaussian models; is the weight value of the nth mixed Gaussian model; is the variance of the n mixed Gaussian models; is the average value of n mixed Gaussian models; If the probability value of the first background image is 1 and the probability value of the second background image is 1, it is determined that there is a moving target in the corresponding background, and a target type of the corresponding moving target is obtained by a preset method; If the probability value of the first background image is 0 and the probability value of the second background image is 1, it is determined that there is a stationary target in the corresponding background; If the probability value of the first background image is 1, and the probability value of the second background image is 0, it is determined that random noise exists in the corresponding background; If the probability value of the first background image is 0, and the probability value of the second background image is 0, it is determined that the corresponding background is a blank background.

8. The unattended substation monitoring system according to claim 7, characterized in that: The preset method includes: Acquire multiple target contour data through offline experiments and establish a class contour database; Based on the first background image and the second background image, the target contour of the moving target is extracted, and matched one by one with the contours in the class contour database to obtain the target type of the corresponding moving target.

9. The unattended substation monitoring system according to claim 1, characterized in that: In the monitoring area state construction unit, based on the environmental feature value of each monitoring area, the first processing result and the second processing result, obtaining the state event of each monitoring area and storing it in the first set includes: For the first set, is the event corresponding to the environmental characteristic value of a single monitoring area, The event corresponding to the first processing result of a single monitoring area, The event corresponding to the second processing result of a single monitoring area; Obtaining the operating condition of each monitoring area and storing it in the second set includes: For the second set, operating conditions for a single monitoring area; Acquiring multiple execution events and storing them in a third set includes: For the third set, For a single execution event.

10. The unattended substation monitoring system according to claim 1, characterized in that: Based on the first set and the second set, matching the execution event of any monitoring area in the third set and executing includes: Based on the first set, a status event of a single monitoring area is obtained, recorded as a first status event; Based on the second set, obtaining an operating condition of a single monitoring area, recorded as a first operating condition; Based on the first state event and the first operating condition, an execution event of a single monitoring area is matched in the third set, recorded as a first execution event and executed.

Images