A device detection system for a smart grid

Abstract

This invention discloses an equipment detection system for smart grids, relating to the field of smart grid technology. It solves the technical problems of existing technologies, such as difficulty in predicting the operating status of power equipment and the inability to identify the mutual influence between different power devices, leading to low efficiency in fault early warning. This invention first predicts whether a power device will fail based on multi-source power data, then identifies secondary equipment, and mines the impact of multi-source power data from these secondary equipment on the target power equipment, thereby determining the fault type. This invention comprehensively considers the mutual influence between various power devices in the smart grid, enabling more accurate determination of the fault type. This invention determines feature points based on data anomalies, retrieves multi-source power data using these feature points as boundaries, generates a primary prediction sequence, and combines this with an artificial intelligence model to predict whether a power device will fail. This invention predicts whether a fault will occur based on the pattern of data changes before and after the event, thus improving prediction accuracy.

Description

Technical Field

[0001] This invention belongs to the field of smart grids and relates to equipment testing technology for smart grids, specifically an equipment testing system for smart grids. Background Technology

[0002] With the development of power equipment fault detection technology, the shortcomings of traditional detection techniques, such as static shutdown inspection and comprehensive manual mechanical inspection, have been overcome to some extent. With the support of technologies such as machine vision and artificial intelligence, automated system detection and calibration can be achieved to a certain degree, reducing detection costs.

[0003] Existing equipment detection technologies generally collect multi-source data from power equipment and analyze the data based on artificial intelligence to determine whether the power equipment is malfunctioning. Although they can quickly and accurately identify faults in power equipment, they are unable to predict the operating status of power equipment or identify the mutual influence between different power equipment, and therefore cannot provide effective fault warnings for smart grids. Therefore, there is an urgent need for an equipment detection system for smart grids. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an equipment detection system for smart grids to solve the technical problem that the prior art is unable to predict the operating status of power equipment, cannot identify the mutual influence between power equipment, and thus has low efficiency in fault early warning.

[0005] To achieve the above objectives, a first aspect of the present invention provides a device detection system for a smart grid, including a central control module, and a data acquisition module and a detection and display module connected thereto;

[0006] The central control module collects multi-source power data from various power devices through the data acquisition module, periodically overwrites and stores the multi-source power data; it extracts and analyzes the multi-source power data to determine if data anomalies have occurred; if so, it extracts the anomaly points as feature points; otherwise, it continues to determine the anomaly point; and...

[0007] By integrating multi-source power data before and after the feature points as boundaries, a first-level prediction sequence is obtained. This sequence is then combined with an artificial intelligence model to predict whether the power equipment is abnormal. If so, a second-level prediction sequence is generated by combining multi-source power data from the power equipment connected to the power equipment. This sequence is then combined with an artificial intelligence model to determine the fault type.

[0008] The detection and display module establishes a device topology model based on the physical connection of power equipment, and displays the predicted power equipment anomalies and fault types in the device topology model in real time, and issues warnings through smart terminals.

[0009] Preferably, the central control module collects multi-source power data from various power devices through a data acquisition module, and performs periodic overlay storage of the multi-source power data, including:

[0010] Determine the data storage period; the data storage period is set according to the data analysis process of the power equipment.

[0011] The system continuously collects multi-source power data from various power devices through several types of data sensors connected to the data acquisition module; the multi-source power data includes voltage or current.

[0012] The collected multi-source power data is stored and overwritten according to the data storage cycle.

[0013] Preferably, the step of extracting and analyzing multi-source power data to determine whether data anomalies have occurred includes:

[0014] Data change curves are constructed based on multi-source power data; these curves include either voltage change curves or current change curves.

[0015] Determine whether there are data anomalies in the current or voltage of power equipment based on the data change curve; among which, data anomalies include sudden changes in effective value or sudden changes in peak value.

[0016] Preferably, the step of integrating multi-source power data before and after the acquisition of a primary prediction sequence by using feature points as boundaries, and then combining this sequence with an artificial intelligence model to predict whether power equipment is abnormal, includes:

[0017] The characteristic points of the power equipment are determined, and power data for several previous and subsequent cycles are extracted from the multi-source power data based on the time corresponding to the characteristic points; the power data includes voltage or current.

[0018] The target data of the power data is extracted, and the target data is integrated according to the collection order to generate a primary prediction sequence. This sequence is then combined with an artificial intelligence model to predict whether the power equipment is abnormal. The target data includes valid values ​​or peak values.

[0019] Preferably, the step of generating a secondary prediction sequence by combining multi-source power data from power equipment connected to the power equipment, and determining the fault type by combining it with an artificial intelligence model, includes:

[0020] When an abnormality is predicted for a power device, the power devices connected to it will be marked as secondary devices.

[0021] Based on the time corresponding to the feature points, power data from several previous and subsequent cycles are extracted from the power multi-source data of the secondary equipment, and integrated into a secondary prediction sequence according to the collection order.

[0022] The secondary prediction sequence is combined with an artificial intelligence model to determine the fault type.

[0023] Preferably, the artificial intelligence model includes a BP neural network model or an RBF neural network model; and is trained using standard training data obtained from historical experience data or experimental simulation data;

[0024] The standard training data includes model input data and corresponding model output data; wherein, the model input data has the same content attributes as the first-level prediction sequence or the second-level prediction sequence.

[0025] Preferably, the detection and display module establishes a device topology model based on the physical connections of the power equipment, and displays the predicted power equipment anomalies and fault types in the device topology model in real time, including:

[0026] Obtain the connection relationships of various power devices in the smart grid and establish a device topology model;

[0027] The operating status of each power device is updated and marked in real time in the equipment topology model, and a warning is issued for faulty power devices through a smart terminal; the operating status includes normal operation, fault, and fault type.

[0028] Preferably, the central control module is communicatively and / or electrically connected to the data acquisition module and the detection and display module, respectively; and the detection and display module is communicatively and / or electrically connected to the smart terminal.

[0029] The data acquisition module communicates and / or is electrically connected to several types of data sensors; wherein the data sensors include voltage sensors or current sensors.

[0030] Compared with the prior art, the beneficial effects of the present invention are:

[0031] 1. This invention first predicts whether power equipment will fail based on multi-source power data, then identifies secondary equipment, and mines the impact of the secondary equipment's multi-source power data on the target power equipment to determine the type of failure. This invention comprehensively considers the mutual influence between various power devices in the smart grid, and can more accurately determine the type of failure of power equipment.

[0032] 2. This invention determines feature points based on data anomalies, retrieves multi-source power data using these feature points as boundaries, generates a first-level prediction sequence, and combines it with an artificial intelligence model to predict whether power equipment will malfunction. This invention predicts whether a malfunction will occur based on the pattern of data changes before and after, which can improve prediction accuracy. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a schematic diagram of the system principle of the present invention;

[0035] Figure 2 This is a schematic diagram illustrating the working steps of the present invention. Detailed Implementation

[0036] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] Please see Figures 1-2 The first aspect of this invention provides a device detection system for a smart grid, including a central control module, and a data acquisition module and a detection and display module connected thereto. The central control module collects multi-source power data from various power devices through the data acquisition module, and performs periodic overlay storage of the multi-source power data; it extracts and analyzes the multi-source power data to determine whether data anomalies have occurred; if so, it extracts the anomaly points as feature points; if not, it continues to judge; and it integrates the multi-source power data before and after the feature points to obtain a first-level prediction sequence, and combines it with an artificial intelligence model to predict whether the power devices are abnormal; if so, it combines the multi-source power data of the power devices connected to the power devices to generate a second-level prediction sequence, and combines it with an artificial intelligence model to determine the fault type; the detection and display module establishes a device topology model based on the physical connection of the power devices, and displays the predicted power device anomalies and fault types in the device topology model in real time, and issues warnings through smart terminals.

[0038] In this invention, the central control module is communicatively and / or electrically connected to the data acquisition module and the detection and display module, respectively; and the detection and display module is communicatively and / or electrically connected to the smart terminal; and the data acquisition module is communicatively and / or electrically connected to several types of data sensors.

[0039] The central control module is primarily responsible for data processing and constructing the equipment topology model, interacting with the data acquisition module and the detection and display module. The data acquisition module is mainly responsible for data acquisition and interacts with various types of data sensors. The detection and display module is mainly used to display detection results and provide early warnings of abnormal situations via smart terminals (mobile phones or computers, etc.). The data sensors in this invention include voltage sensors or current sensors, mainly used to continuously collect various power data from the power equipment, identifying faults in the power equipment through changes in these multi-source power data.

[0040] In a preferred embodiment, the central control module collects multi-source power data from various power devices through a data acquisition module, and performs periodic overlay storage of the multi-source power data, including: determining the data storage period; continuously collecting multi-source power data from various power devices through several types of data sensors connected to the data acquisition module; and storing and overlaying the collected multi-source power data according to the data storage period.

[0041] The foundation for power equipment monitoring in smart grids is suitable data, which is the multi-source power data discussed in this invention. To reduce the overall system storage cost, a data storage cycle needs to be set to promptly delete historical data that is useless for power equipment monitoring and analysis. This invention continuously collects data and simultaneously updates old data with newly collected data according to the data storage cycle.

[0042] It should be noted that the data storage period in this embodiment is set according to the data analysis process of the power equipment. The multi-source power data stored based on the data storage period can generate at least one complete prediction sequence; however, to ensure prediction accuracy, it is best to generate at least one prediction sequence.

[0043] Anomalies in multi-source power data (such as voltage and current) can help identify potential faults or malfunctions in power equipment. Therefore, a preliminary assessment is made using multi-source power data corresponding to the power equipment. This involves extracting and analyzing the multi-source power data to determine if any anomalies have occurred, including: constructing data change curves based on the multi-source power data; and determining whether there are anomalies in the current or voltage of the power equipment based on the data change curves.

[0044] Data variation curves, such as current variation curves and voltage variation curves, are constructed based on multi-source power data corresponding to the power equipment. When the data variation curves deviate from the standard curves, or exhibit abnormal non-periodic changes, data anomalies are indicated, suggesting a potential malfunction in the power equipment. It's important to note that these anomalies include those of the power equipment itself and the impact of anomalies from other power equipment on that equipment. When a power equipment anomaly persists and affects normal operation, timely warnings can be issued; if it currently does not affect operation, it's necessary to predict whether the power equipment will fail in the future.

[0045] By integrating multi-source power data before and after the acquisition of a primary prediction sequence based on feature points, and combining this with an artificial intelligence model to predict whether power equipment is abnormal, the process includes: identifying feature points of the power equipment; extracting power data from multi-source power data for several cycles before and after the corresponding time points; extracting target data from the power data; integrating the target data according to the acquisition order to generate a primary prediction sequence; and combining this with an artificial intelligence model to predict whether power equipment is abnormal.

[0046] Points where data anomalies occur are used as feature points. Power data for several cycles (data change cycles) before and after the feature point are extracted from the multi-source power data of the corresponding power equipment. After extracting the target data from the power data, a first-level prediction sequence is generated in sequence. For example: after determining the feature point, the effective voltage and current values ​​for the five cycles before the feature point and the effective voltage and current values ​​for the three cycles after the feature point are extracted and integrated to generate a first-level prediction sequence [(QDY1,QDL1), (QDY2,QDL2), (QDY3,QDL3), (QDY4,QDL4), (QDY5,QDL5), (HDY1,HDL1), (HDY2,HDL2), (HDY3,HDL3)].

[0047] The primary prediction sequence is input into a specially trained artificial intelligence model, and the corresponding output can predict whether power equipment will malfunction. In the standard training data for this artificial intelligence model, the model input data has the same attributes as the primary prediction sequence, while the model output data is the result corresponding to the input data. This can be set or adjusted by professional power personnel.

[0048] Based on multi-source power data before and after a feature point, it is possible to quickly predict whether power equipment will malfunction. However, the specific cause of the malfunction is not clear; that is, it is uncertain whether it is due to the power equipment itself or affected by other power equipment. Therefore, it is necessary to further clarify the type of malfunction.

[0049] A secondary prediction sequence is generated by combining multi-source power data from power equipment connected to the power equipment. The fault type is determined by combining the sequence with an artificial intelligence model. This includes: when a power equipment is predicted to be abnormal, the power equipment connected to it is marked as a secondary equipment; power data from several previous and subsequent cycles are extracted from the multi-source power data of the secondary equipment based on the time corresponding to the feature points, and integrated into a secondary prediction sequence according to the collection order; the secondary prediction sequence is combined with the artificial intelligence model to determine the fault type.

[0050] When the prediction results for a power device are abnormal, other power devices directly connected to it are marked as secondary devices. Again, using feature points as boundaries, power data from several preceding and following cycles are extracted and integrated into a secondary prediction sequence. This secondary prediction sequence is then input into a specially trained artificial intelligence model to obtain the corresponding output results. Combined with the output results of the primary prediction sequence, the fault type is determined.

[0051] The mutual influence between power equipment cannot be quantified, so artificial intelligence models are used to explore it. Moreover, secondary equipment includes the upstream and downstream equipment of the faulty power equipment. Power flows from the upstream equipment through the faulty power equipment to the downstream equipment. Obviously, the impact of the upstream equipment and the downstream equipment on the faulty power equipment is different.

[0052] During the generation of the secondary prediction sequence, the power multi-source data corresponding to each secondary device can also be pre-judged. That is, by referring to the primary prediction sequence and the prediction process, the power multi-source data of each power device before and after the feature point is analyzed to determine whether the secondary device will be abnormal. If so, the corresponding power device is marked as 1; otherwise, it is marked as 0. This makes the generated secondary prediction sequence simpler.

[0053] The artificial intelligence model used in the secondary prediction sequence is also specifically trained. The model input data of the standard training data used for training is consistent with the content attributes of the secondary prediction sequence. The model output data is obtained by summarizing historical experience data or experimental simulation data. The model input data can be specifically represented as whether the secondary equipment has an impact on the target power equipment.

[0054] If the secondary equipment does not affect the abnormal power equipment, the fault type is determined to be caused by the power equipment itself, and a solution can be matched based on multi-source power data; if the secondary equipment has a significant impact on the abnormal power equipment, the impact of the secondary equipment should be eliminated, and the abnormal secondary equipment should be treated together with the warning.

[0055] Finally, the detection and display module establishes a device topology model based on the physical connections of power equipment, and displays the predicted power equipment anomalies and fault types in the device topology model in real time. This includes: obtaining the connection relationships of each power device in the smart grid and establishing a device topology model; updating and marking the operating status of each power device in the device topology model in real time, and issuing early warnings for faulty power devices through smart terminals. Power workers can troubleshoot based on the real-time updated device topology model.

[0056] The working principle of this invention is as follows: Multi-source power data from various power devices is collected and periodically overwritten and stored. The multi-source power data is extracted and analyzed to determine if any data anomalies have occurred. If so, the anomaly points are extracted as feature points; otherwise, the process continues. The multi-source power data is integrated with the feature points as boundaries to obtain a primary prediction sequence. This sequence is then combined with an artificial intelligence model to predict whether the power devices are abnormal. If so, a secondary prediction sequence is generated by combining the multi-source power data from the connected power devices, and this sequence is further combined with the artificial intelligence model to determine the fault type. A device topology model is established based on the physical connections of the power devices, and the predicted power device anomalies and fault types are displayed in real time within the topology model, with early warnings issued via smart terminals.

[0057] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A device detection system for a smart grid, comprising a central control module, and a data acquisition module and a detection and display module connected thereto; characterized in that: The central control module collects multi-source power data from various power devices through the data acquisition module, and performs periodic overwrite storage of the multi-source power data; Extract and analyze multi-source power data to determine if any data anomalies have occurred; If yes, then extract the points of variation as feature points; No, continue the assessment; as well as By integrating multi-source power data before and after the feature points as boundaries, a first-level prediction sequence is obtained. This sequence is then combined with an artificial intelligence model to predict whether the power equipment is abnormal. If so, a second-level prediction sequence is generated by combining multi-source power data from the power equipment connected to the power equipment. This sequence is then combined with an artificial intelligence model to determine the fault type. The detection and display module establishes a device topology model based on the physical connection of power equipment, and displays the predicted power equipment anomalies and fault types in the device topology model in real time, and issues warnings through smart terminals; The process of generating a secondary prediction sequence by combining multi-source power data from power equipment connected to the power equipment, and then using an artificial intelligence model to determine the fault type, includes: When an abnormality is predicted for a power device, the power devices connected to it will be marked as secondary devices. Based on the time corresponding to the feature points, power data from several previous and subsequent cycles are extracted from the power multi-source data of the secondary equipment, and integrated into a secondary prediction sequence according to the collection order. The secondary prediction sequence is combined with an artificial intelligence model to determine the fault type.

2. The equipment testing system for a smart grid according to claim 1, characterized in that, The central control module collects multi-source power data from various power devices through the data acquisition module, and performs periodic overlay storage of the multi-source power data, including: Determine the data storage period; the data storage period is set according to the data analysis process of the power equipment. The system continuously collects multi-source power data from various power devices through several types of data sensors connected to the data acquisition module; the multi-source power data includes voltage or current. The collected multi-source power data is stored and overwritten according to the data storage cycle.

3. The equipment testing system for a smart grid according to claim 1, characterized in that, The extraction and analysis of multi-source power data to determine whether data anomalies have occurred includes: Data change curves are constructed based on multi-source power data; these curves include either voltage change curves or current change curves. Determine whether there are data anomalies in the current or voltage of power equipment based on the data change curve; among which, data anomalies include sudden changes in effective value or sudden changes in peak value.

4. The equipment testing system for a smart grid according to claim 1, characterized in that, The process of integrating multi-source power data before and after the acquisition of a primary prediction sequence by using feature points as boundaries, and then combining this sequence with an artificial intelligence model to predict whether power equipment is abnormal, includes: The characteristic points of the power equipment are determined, and power data for several previous and subsequent cycles are extracted from the multi-source power data based on the time corresponding to the characteristic points; the power data includes voltage or current. The target data of the power data is extracted, and the target data is integrated according to the collection order to generate a primary prediction sequence. This sequence is then combined with an artificial intelligence model to predict whether the power equipment is abnormal. The target data includes valid values ​​or peak values.

5. The equipment testing system for a smart grid according to claim 4, characterized in that, The artificial intelligence model includes a BP neural network model or an RBF neural network model; and is trained using standard training data obtained from historical experience data or experimental simulation data. The standard training data includes model input data and corresponding model output data; wherein, the model input data has the same content attributes as the first-level prediction sequence or the second-level prediction sequence.

6. The equipment testing system for a smart grid according to claim 1, characterized in that, The detection and display module establishes a device topology model based on the physical connections of power equipment, and displays the predicted power equipment anomalies and fault types in the device topology model in real time, including: Obtain the connection relationships of various power devices in the smart grid and establish a device topology model; The operating status of each power device is updated and marked in real time in the equipment topology model, and a warning is issued for faulty power devices through a smart terminal; the operating status includes normal operation, fault, and fault type.

7. The equipment testing system for a smart grid according to claim 1, characterized in that, The central control module is communicatively and / or electrically connected to the data acquisition module and the detection and display module, respectively; and the detection and display module is communicatively and / or electrically connected to the smart terminal. The data acquisition module communicates and / or is electrically connected to several types of data sensors; wherein the data sensors include voltage sensors or current sensors.

Images