Target detection method and system of power transmission line based on Internet of Things

Abstract

The invention relates to the technical field of power transmission line detection, in particular to a target detection method and system for a power transmission line based on the Internet of Things, and the method comprises the steps: S1, data standardization: after original data is input, carrying out the standardization processing of the original data, subtracting the mean value from each variable, dividing the mean value by the standard deviation, and carrying out the calculation; s2, calculating a covariance matrix, and calculating the covariance matrix of the standardized data according to the covariance matrix of the standardized data, so as to ensure that the contribution of each variable in the data set is balanced; according to the system, analysis is performed through the data analysis module, the collected mass data can be stored, analyzed and processed to extract useful information and generate a report, a scientific basis is provided for decision making of a user, the processed data is analyzed through the machine learning module and a deep learning module algorithm, and the user experience is improved. And identifying an abnormal state or a potential fault of the power transmission line, identifying the potential fault through data analysis, and triggering an early warning mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission line detection, and in particular to a method and system for detecting a target of a power transmission line based on the Internet of Things. Background Art

[0002] The target detection method and system for power transmission lines based on the Internet of Things is a solution that integrates multiple advanced technologies and aims to improve the safety and reliability of power transmission lines. The Internet of Things refers to a network that connects any object to the Internet through information sensing devices such as radio frequency identification, infrared sensors, global positioning systems, laser scanners, etc., according to agreed protocols, to exchange information and communicate, so as to realize intelligent identification, positioning, tracking, monitoring and management of objects.

[0003] The patent specification with the announcement number CN116245871A discloses a method for detecting abnormal targets in power transmission lines based on the improved YOLOX algorithm, which includes the following steps: 1) inputting an abnormal image of the power transmission line to be identified, adjusting it to a standard input image after preprocessing, and training it based on the improved YOLOX algorithm; 2) using the Focus model to split the high-resolution image into multiple low-resolution images; 3) extracting the features of the input image through the feature selection module as abnormal features; 4) continuously extracting the abnormal features and dividing them into three effective feature layers of P3, P4 and P5; 5) inputting the feature map of the effective feature layer into the PAFPN network for feature extraction, efficiently fusing the feature maps of different feature layers, and outputting the feature maps of different feature layers; 6) detecting and locating abnormal targets on the effective feature maps of different feature layers. Compared with the prior art, the present invention has the advantages of high accuracy in detecting abnormal targets and reduced algorithm complexity.

[0004] The patent specification with the announcement number CN111444809A discloses a method for detecting abnormal targets in power transmission lines based on improved YOLOv3, which includes the collection and annotation of scene pictures under the power transmission lines; using online data enhancement means to expand the data volume; making targeted improvements to the network structure according to the characteristics of the targets in the pictures and the project requirements; improving the coordinate loss function according to the multi-scale characteristics of the targets; inputting the training data into the improved network for training to obtain the best prediction model; collecting the scene pictures in real time; using the obtained best prediction model to forward deduce the scene pictures; screening out the detection frames with a confidence higher than the set threshold; performing non-maximum suppression processing on the detection frames to reduce repeated detection; drawing the target detection frame on the scene pictures to display the target position damaged by external forces. The present invention can detect the targets (including excavators, tower cranes, and cranes) under the power transmission lines in real time and accurately detect and locate the targets (including excavators, tower cranes, and cranes) that may potentially damage the power transmission lines.

[0005] Although the above technical solution has the advantages of high accuracy in detecting abnormal targets and reducing algorithm complexity, and can detect targets under the transmission line that may potentially cause damage to the transmission line in real time, the above method can only detect and locate external target features and cannot monitor the line target status in real time, resulting in the inability to timely discover and handle potential faults, and the inability to ensure the safety and reliability of the transmission line. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a target detection method and system for a power transmission line based on the Internet of Things.

[0007] In order to achieve the above-mentioned object, the present invention adopts the following technical scheme: a target detection method for a power transmission line based on the Internet of Things, including S1, data standardization, after the original data is input, the original data needs to be standardized first, and each variable is subtracted from its mean and divided by its standard deviation to ensure that the contribution of each variable in the data set is balanced;

[0008] S2. Calculate the covariance matrix. For the standardized data, calculate its covariance matrix. If the covariance of two variables is large, it indicates that there is a close relationship between them. If the covariance is close to 0, it indicates that there is almost no relationship between them.

[0009] S3, eigenvalue calculation, perform eigenvalue decomposition on the covariance matrix to obtain the eigenvalue, which represents the variance of the principal component and reflects the amount of information contained. The larger the eigenvalue, the more original data information it contains;

[0010] S4, eigenvector calculation, perform eigenvalue decomposition on the covariance matrix to obtain the eigenvector, which represents the direction of the principal element, that is, the direction of the coordinate axis of the data in the new coordinate system;

[0011] S5. Select the principal component, select the number of principal components to be retained according to the size of the eigenvalue, and retain the first k principal components in the order of eigenvalue from large to small, where k is a user-defined dimension. These principal components represent the main change direction in the data and are orthogonal to each other. By setting a variance explanation ratio, such as 95%, the original data is projected into a new feature space composed of the selected principal components to obtain the reduced-dimensional data, which is realized by matrix multiplication. The original data matrix is ​​multiplied by the matrix composed of the selected principal components, i.e., the eigenvectors, to obtain a new feature matrix;

[0012] S6. Process monitoring is used to reduce the dimensions of multiple variables that affect system performance, derive the main process variables that affect the system, and provide information basis for analysis, monitoring, and optimization of the production process.

[0013] As a further description of the above technical solution: the step S5, selecting the principal component, further includes the following steps S501, data compression, reducing the storage and transmission requirements of data by retaining the main principal components;

[0014] S502: Noise filtering, reducing the impact of noise by removing secondary components with smaller variance.

[0015] S503, data dimensionality reduction, converting high-dimensional data into low-dimensional data through mapping or transformation. This transformation can be linear, such as PCA, or nonlinear, such as t-SNE, to reduce the dimensionality of the data and make the data set more concise while trying to maintain the structure or variance of the data.

[0016] The target detection system of the power transmission line based on the Internet of Things includes a target detection platform, the output end of which is respectively connected to a GSM wireless communication module, an application layer and a dynamic data storage module, and the output end of the GSM wireless communication module is connected to a perception data acquisition module.

[0017] As a further description of the above technical solution: the perception data acquisition module includes a temperature and humidity sensor, a current sensor, a vibration sensor, a displacement sensor and a surveillance camera;

[0018] Wherein, the temperature and humidity sensor is used to detect the temperature and humidity of key parts of the transmission line;

[0019] The current sensor is used to detect current at key parts of the power transmission line;

[0020] The vibration sensor is used to detect vibrations at key parts of the power transmission line;

[0021] The displacement sensor is used to detect the displacement of key parts of the power transmission line;

[0022] The monitoring camera is used for visual monitoring of key parts of the power transmission line and its surrounding environment.

[0023] As a further description of the above technical solution: the output end of the application layer is connected to a data processing module, and the output end of the data processing module is connected to a data analysis module.

[0024] As a further description of the above technical solution: the data processing module includes a preprocessing module, a cleaning module and a noise removal module;

[0025] Wherein, the preprocessing module is used to perform standardization processing on the collected data;

[0026] The cleaning module is used to clean the data after standardization processing;

[0027] The noise removal module is used to remove secondary components with smaller variance from the cleaned data to reduce the impact of noise.

[0028] As a further description of the above technical solution: the data analysis module includes a machine learning module and a deep learning module;

[0029] Wherein, the machine learning module is used to perform row analysis on the processed data;

[0030] The deep learning module is used to identify abnormal conditions or potential faults of the power transmission line from the data.

[0031] As a further description of the above technical solution: the output end of the data analysis module is connected to a display module.

[0032] As a further description of the above technical solution: the output end of the display module is respectively connected to the early warning module and the alarm module.

[0033] As a further description of the above technical solution: the output end of the dynamic data storage module is respectively connected to a local storage module and a cloud storage module.

[0034] The present invention has the following beneficial effects:

[0035] The object detection method and system for power transmission lines based on the Internet of Things designed by the present invention can store, analyze and process the collected massive data through the data analysis module to extract useful information and generate reports, providing a scientific basis for users' decision-making. The processed data is analyzed by using the machine learning module and the deep learning module algorithm to identify the abnormal state or potential faults of the power transmission line, identify potential faults through data analysis, and trigger the early warning mechanism;

[0036] The object detection method and system for power transmission lines based on the Internet of Things designed by the present invention can remove abnormal values ​​in the data through the preprocessing module, and the cleaning module preprocesses and cleans the data, and the noise removal module extracts feature information useful for subsequent tasks from the data, so that the data is standardized;

[0037] The object detection method and system of the power transmission line based on the Internet of Things designed by the present invention can collect the environmental parameters and information of the power transmission line in real time through sensors, and convert them into electrical signals or digital signals for transmission. The GSM wireless communication module is set to enable the application layer to receive the detected status data in real time. After the application layer receives the real-time monitoring data, the data processing module processes the data;

[0038] The object detection method and system for power transmission lines based on the Internet of Things designed by the present invention, the dynamic data storage module can dynamically allocate and release storage units. When the local storage module stores data, the new data can be allocated and stored in the cloud storage module. The memory space can be dynamically allocated and released as needed to adapt to different data storage requirements. This flexibility enables dynamic storage to efficiently utilize memory resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0040] Figure 2 It is a schematic diagram of the workflow of step S5 of the present invention;

[0041] Figure 3 It is a schematic diagram of the main structure of the system of the present invention;

[0042] Figure 4 It is a structural schematic diagram of the data processing module of the present invention;

[0043] Figure 5 This is a structural diagram of the perception data acquisition module of the present invention;

[0044] Figure 6 It is a structural schematic diagram of the dynamic data storage module of the present invention.

[0045] Legend:

[0046] 1. Target detection platform; 2. GSM wireless communication module; 3. Perception data acquisition module; 31. Temperature and humidity sensor; 32. Current sensor; 33. Vibration sensor; 34. Displacement sensor; 35. Surveillance camera; 4. Application layer; 5. Data processing module; 51. Preprocessing module; 52. Cleaning module; 53. Noise removal module; 6. Data analysis module; 61. Machine learning module; 62. Deep learning module; 7. Display module; 8. Early warning module; 9. Alarm module; 10. Dynamic data storage module; 11. Local storage module; 12. Cloud storage module. DETAILED DESCRIPTION

[0047] Reference Figure 1-Figure 6 , the object detection method of the power transmission line based on the Internet of Things provided by the present invention includes S1, data standardization. After the original data is input, the original data needs to be standardized first, and each variable is subtracted from its mean and divided by its standard deviation to ensure that the contribution of each variable in the data set is balanced;

[0048] S2. Calculate the covariance matrix. For the standardized data, calculate its covariance matrix. If the covariance of two variables is large, it indicates that there is a close relationship between them. If the covariance is close to 0, it indicates that there is almost no relationship between them.

[0049] S3, eigenvalue calculation, perform eigenvalue decomposition on the covariance matrix to obtain the eigenvalue, which represents the variance of the principal component and reflects the amount of information contained. The larger the eigenvalue, the more original data information it contains;

[0050] S4, eigenvector calculation, perform eigenvalue decomposition on the covariance matrix to obtain the eigenvector, which represents the direction of the principal element, that is, the direction of the coordinate axis of the data in the new coordinate system;

[0051] S5. Select the principal component, select the number of principal components to be retained according to the size of the eigenvalue, and retain the first k principal components in the order of eigenvalue from large to small, where k is a user-defined dimension. These principal components represent the main change direction in the data and are orthogonal to each other. By setting a variance explanation ratio, such as 95%, the original data is projected into a new feature space composed of the selected principal components to obtain the reduced-dimensional data, which is realized by matrix multiplication. The original data matrix is ​​multiplied by the matrix composed of the selected principal components, i.e., the eigenvectors, to obtain a new feature matrix;

[0052] S6. Process monitoring is used to reduce the dimensions of multiple variables that affect system performance, derive the main process variables that affect the system, and provide information basis for analysis, monitoring, and optimization of the production process.

[0053] Reference Figure 2 , the object detection method of the power transmission line based on the Internet of Things provided by the present invention:

[0054] Step S5, selecting the pivot element includes the following steps:

[0055] S501, data compression, reducing data storage and transmission requirements by retaining the main principal components;

[0056] S502: Noise filtering, reducing the impact of noise by removing secondary components with smaller variance.

[0057] S503, data dimensionality reduction, converting high-dimensional data into low-dimensional data through mapping or transformation. This transformation can be linear, such as PCA, or nonlinear, such as t-SNE, to reduce the dimensionality of the data and make the data set more concise while trying to maintain the structure or variance of the data.

[0058] The target detection system of the power transmission line based on the Internet of Things includes a target detection platform 1, the output end of the target detection platform 1 is respectively connected to a GSM wireless communication module 2, an application layer 4 and a dynamic data storage module 10, and the output end of the GSM wireless communication module 2 is connected to a perception data acquisition module 3.

[0059] As a further implementation scheme of the above technical scheme: the perception data acquisition module 3 includes a temperature and humidity sensor 31, a current sensor 32, a vibration sensor 33, a displacement sensor 34 and a monitoring camera 35;

[0060] Among them, the temperature and humidity sensor 31 is used to detect the temperature and humidity of key parts of the transmission line;

[0061] The current sensor 32 is used to detect the current at the key parts of the transmission line;

[0062] The vibration sensor 33 is used to detect vibrations at key parts of the power transmission line;

[0063] The displacement sensor 34 is used to detect the displacement of key parts of the transmission line;

[0064] The monitoring camera 35 is used for visual monitoring of the key parts of the transmission line and its surrounding environment; the temperature and humidity sensor 31 detects the temperature and humidity of the key parts of the transmission line, the current sensor 32 can detect the current at the key parts of the transmission line, the vibration sensor 33 and the displacement sensor 34 detect the operating status of the line, and the monitoring camera 35 can monitor the environment of the key parts of the line. These sensors can collect the environmental parameters and information of the transmission line in real time, and convert them into electrical signals or digital signals for transmission.

[0065] As a further implementation scheme of the above technical scheme: the output end of the application layer 4 is connected to the data processing module 5, and the output end of the data processing module 5 is connected to the data analysis module 6; the application layer 4 can receive the detected status data in real time through the set GSM wireless communication module 2.

[0066] As a further implementation of the above technical solution: the data processing module 5 includes a pre-processing module 51, a cleaning module 52 and a noise removal module 53;

[0067] Among them, the preprocessing module 51 is used to perform standardization processing on the collected data;

[0068] The cleaning module 52 is used to clean the data after the standardization process;

[0069] The noise removal module 53 is used to remove the secondary components with smaller variance from the cleaned data to reduce the impact of noise; after the application layer 4 receives the real-time monitoring data, the data processing module 5 processes the data, and the preprocessing module 51 is set to remove the abnormal values ​​in the data, while the cleaning module 52 preprocesses and cleans the data. The noise removal module 53 extracts feature information useful for subsequent tasks from the data, so that the data is standardized, and the principal components are arranged in descending order according to the size of the eigenvalues, and the original data is projected into a new feature space composed of the selected principal components to obtain the reduced-dimensional data.

[0070] As a further implementation of the above technical solution: the data analysis module 6 includes a machine learning module 61 and a deep learning module 62;

[0071] Among them, the machine learning module 61 is used to perform row analysis on the processed data;

[0072] The deep learning module 62 is used to identify abnormal conditions or potential faults of the transmission line from the data; the machine learning module 61 uses the AdaBoost algorithm, which predicts the target variable by combining the results of multiple weak classifiers, trains a weak classifier for each subset, and adjusts the weight of each weak classifier so that misclassified samples receive a larger weight, identifies abnormal conditions or potential faults of the transmission line, identifies potential faults through data analysis, and triggers an early warning mechanism.

[0073] In specific implementation, the target detection platform 1 of the target detection method and system for the power transmission line based on the Internet of Things is provided with a GSM wireless communication module 2, a perception data acquisition module 3 and an application layer 4. The perception data acquisition module 3 can detect the key target parts of the power transmission line, the temperature and humidity sensor 31 detects the temperature and humidity of the key parts of the power transmission line, the current sensor 32 can detect the current of the key parts of the power transmission line, the vibration sensor 33 and the displacement sensor 34 detect the operation status of the line, and the monitoring camera 35 can monitor the environment of the key parts of the line. These sensors can collect the environmental parameters and information of the power transmission line in real time and convert them into electrical signals or digital signals for transmission. The application layer 4 can receive the detected status data in real time through the set GSM wireless communication module 2. After the application layer 4 receives the real-time monitoring data, the data processing module 5 processes the data. The preprocessing module 51 can remove the abnormal values ​​in the data, and the cleaning module 52 preprocesses and cleans the data. The noise removal module 53 extracts feature information useful for subsequent tasks from the data, so that the data is standardized. Standardization usually includes reducing each variable. Remove its mean and divide it by its standard deviation to ensure that the contribution of each variable in the data set is balanced. After the standardized data, calculate its covariance matrix. The covariance matrix describes the correlation and variance between variables. Perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​and eigenvectors. According to the size of the eigenvalue, arrange the principal components in descending order, project the original data into a new feature space composed of the selected principal components, and obtain the data after dimensionality reduction. The processed data is sent to the data analysis module 6 for analysis, which can store, analyze and process the collected massive data to extract useful information and generate reports. To provide a scientific basis for users' decisions, the machine learning module 61 uses the AdaBoost algorithm, which predicts the target variable by combining the results of multiple weak classifiers, trains a weak classifier for each subset, and adjusts the weight of each weak classifier so that misclassified samples get a larger weight, and then trains the weak classifier again. This process is repeated over and over again, so that the processed data can be analyzed by using the machine learning module 61 and the deep learning module 62 algorithms to identify abnormal conditions or potential faults of the transmission line, identify potential faults through data analysis, and trigger an early warning mechanism.

[0074] As a further implementation of the above technical solution: the output end of the data analysis module 6 is connected to the display module 7; the analyzed data is displayed through the display module 7.

[0075] As a further implementation scheme of the above technical scheme: the output ends of the display module 7 are respectively connected to the early warning module 8 and the alarm module 9; the analyzed data can trigger the early warning module 8 or the alarm module 9, and can trigger the early warning or alarm mechanism according to the abnormal state or potential fault of the transmission line.

[0076] As a further implementation scheme of the above technical scheme: the output ends of the dynamic data storage module 10 are respectively connected to the local storage module 11 and the cloud storage module 12; the analyzed data will be sent to the dynamic data storage module 10 for storage, which is used to provide a historical data query function, so that the operation and maintenance personnel can analyze the line operation trends and rules, and can dynamically allocate and release memory space as needed to adapt to different data storage requirements and improve the data transmission efficiency.

[0077] During specific implementation, the analyzed data is displayed through the display module 7, and the analyzed data can trigger the early warning module 8 or the alarm module 9, which can trigger the early warning or alarm mechanism according to the abnormal state or potential fault of the transmission line. The analyzed data will be sent to the dynamic data storage module 10 for storage, which is used to provide a historical data query function, so that the operation and maintenance personnel can analyze the line operation trends and rules. The dynamic data storage module 10 can dynamically allocate and release storage units. When the local storage module 11 stores data, the new data can be allocated and stored in the cloud storage module 12. The memory space can be dynamically allocated and released as needed to adapt to different data storage requirements. This flexibility enables dynamic storage to efficiently utilize memory resources to improve the efficiency of data transmission.

[0078] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A target detection method for a power transmission line based on the Internet of Things, characterized in that: The following steps are involved: S1. Data standardization. After the original data is input, it is necessary to standardize the original data first. Each variable is subtracted from its mean and divided by its standard deviation to ensure that the contribution of each variable in the data set is balanced. S2. Calculate the covariance matrix. For the standardized data, calculate its covariance matrix. If the covariance of two variables is large, it indicates that there is a close relationship between them. If the covariance is close to 0, it indicates that there is almost no relationship between them. S3, eigenvalue calculation, perform eigenvalue decomposition on the covariance matrix to obtain the eigenvalue, which represents the variance of the principal component and reflects the amount of information contained. The larger the eigenvalue, the more original data information it contains; S4, eigenvector calculation, perform eigenvalue decomposition on the covariance matrix to obtain the eigenvector, which represents the direction of the principal element, that is, the direction of the coordinate axis of the data in the new coordinate system; S5. Select the principal component, select the number of principal components to be retained according to the size of the eigenvalue, and retain the first k principal components in the order of eigenvalue from large to small, where k is a user-defined dimension. These principal components represent the main change direction in the data and are orthogonal to each other. By setting a variance explanation ratio, such as 95%, the original data is projected into a new feature space composed of the selected principal components to obtain the reduced-dimensional data, which is realized by matrix multiplication. The original data matrix is ​​multiplied by the matrix composed of the selected principal components, i.e., the eigenvectors, to obtain a new feature matrix; S6. Process monitoring is used to reduce the dimensions of multiple variables that affect system performance, derive the main process variables that affect the system, and provide information basis for analysis, monitoring, and optimization of the production process.

2. The target detection method for power transmission lines based on the Internet of Things according to claim 1 is characterized in that: The step S5, selecting the pivot, also includes the following steps: S501, data compression, reducing data storage and transmission requirements by retaining the main principal components; S502, noise filtering, reducing the influence of noise by removing secondary components with smaller variance; S503, data dimensionality reduction, converting high-dimensional data into low-dimensional data through mapping or transformation. This transformation can be linear, such as PCA, or nonlinear, such as t-SNE, to reduce the dimensionality of the data and make the data set more concise while trying to maintain the structure or variance of the data.

3. A target detection system for a power transmission line based on the Internet of Things, comprising a target detection platform (1), wherein the output end of the target detection platform (1) is respectively connected to a GSM wireless communication module (2), an application layer (4) and a dynamic data storage module (10), and characterized in that: The output end of the GSM wireless communication module (2) is connected to a sensing data collection module (3).

4. The target detection system for power transmission lines based on the Internet of Things according to claim 3 is characterized in that: The sensing data acquisition module (3) comprises a temperature and humidity sensor (31), a current sensor (32), a vibration sensor (33), a displacement sensor (34) and a monitoring camera (35); Wherein, the temperature and humidity sensor (31) is used to detect the temperature and humidity of key parts of the power transmission line; The current sensor (32) is used to detect current at key parts of the power transmission line; The vibration sensor (33) is used to detect vibrations at key parts of a power transmission line; The displacement sensor (34) is used to detect the displacement of key parts of the power transmission line; The monitoring camera (35) is used for visually monitoring the key parts of the power transmission line and its surrounding environment.

5. The target detection system for power transmission lines based on the Internet of Things according to claim 3 is characterized in that: The output end of the application layer (4) is connected to a data processing module (5), and the output end of the data processing module (5) is connected to a data analysis module (6).

6. The target detection system for power transmission lines based on the Internet of Things according to claim 5 is characterized in that: The data processing module (5) comprises a preprocessing module (51), a cleaning module (52) and a noise removal module (53); Wherein, the preprocessing module (51) is used to perform standardization processing on the collected data; The cleaning module (52) is used to clean the data after the standardization process; The noise removal module (53) is used to remove secondary components with smaller variance from the cleaned data to reduce the impact of noise.

7. The target detection system for power transmission lines based on the Internet of Things according to claim 5, characterized in that: The data analysis module (6) comprises a machine learning module (61) and a deep learning module (62); Wherein, the machine learning module (61) is used to perform row analysis on the processed data; The deep learning module (62) is used to identify abnormal conditions or potential faults of the power transmission line from the data.

8. The target detection system for power transmission lines based on the Internet of Things according to claim 7, characterized in that: The output end of the data analysis module (6) is connected to a display module (7).

9. The target detection system for power transmission lines based on the Internet of Things according to claim 8, characterized in that: The output end of the display module (7) is respectively connected to the early warning module (8) and the alarm module (9).

10. The target detection system for power transmission lines based on the Internet of Things according to claim 3, characterized in that: The output end of the dynamic data storage module (10) is respectively connected to a local storage module (11) and a cloud storage module (12).

Images