Wind power field station remote monitoring system based on artificial intelligence

Abstract

The invention relates to the field of wind power field station monitoring, and discloses an artificial intelligence-based wind power field station remote monitoring system, which is characterized in that a field station is divided into a plurality of inspection subareas, an artificial intelligence algorithm is used for automatically planning inspection routes for intelligent inspection equipment in the inspection subareas, and inspection and data acquisition are carried out; through analysis and data fusion of the image, temperature data and vibration data of the wind power equipment in the inspection sub-region, the abnormal problem of the wind power equipment in the inspection sub-region is comprehensively detected. Then fault prediction is carried out on abnormal wind power equipment in the inspection sub-region based on combination of historical power generation amount, voltage and current and real-time power generation amount, voltage and current, and remote monitoring and management are carried out by a remote monitoring platform, so that comprehensive coverage and real-time monitoring of the wind power plant equipment and region can be realized; the inspection efficiency and precision are improved, and the accuracy and reliability of fault prediction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farm station monitoring, and more specifically to a wind farm station remote monitoring system based on artificial intelligence. Background Art

[0002] With the rapid development of wind power, the number of wind turbines has increased dramatically, and the scale of wind farms has continued to expand, which has gradually improved the status of wind power in the energy structure. However, wind farms are usually distributed in remote areas, covering a large area and with a complex and changeable environment, which poses a huge challenge to the operation and maintenance of wind farms. Therefore, the wind farm monitoring system can be used to intelligently monitor the wind farm, collect data by wired or wireless means, and transmit it to the server through the network, conduct in-depth analysis of the collected data, set various alarm and warning rules, and generate various reports and trend charts, so as to conduct real-time monitoring and early warning feedback on the operation of the wind farm;

[0003] However, the above process still has the following disadvantages:

[0004] First, the existing wind farm station monitoring system relies heavily on manual inspection, data analysis and fault diagnosis, which cannot guarantee inspection efficiency and accuracy, and it is difficult to fully cover all equipment and areas;

[0005] Second, the existing wind farm monitoring systems mainly rely on simple data analysis tools and methods, which are often unable to cope with complex data analysis and processing tasks, and lack in-depth analysis of wind farm equipment from multiple aspects of data.

[0006] Third, the existing wind farm station monitoring system lacks the ability to accurately predict and quickly diagnose wind power equipment failures by combining historical data with real-time data, and cannot provide a remote monitoring platform for managers to remotely monitor and manage the equipment operating status of new energy stations. Summary of the invention

[0007] In order to overcome the above-mentioned defects of the prior art, the present invention provides a wind farm station remote monitoring system based on artificial intelligence to solve the problems existing in the above-mentioned background technology.

[0008] The present invention provides the following technical solution: a remote monitoring system for a wind farm station based on artificial intelligence, comprising:

[0009] Inspection area division module: divide the station into several inspection sub-areas according to the equipment type in the new energy station, and number them as 1, 2, 3...n respectively;

[0010] Inspection route planning module: In the divided inspection sub-areas, the inspection route is automatically planned for the intelligent inspection equipment using artificial intelligence algorithms, and then the new energy stations are inspected according to the planned inspection route;

[0011] Inspection data collection module: used to collect images, temperature data and vibration data of wind power equipment in each inspection sub-area. The images, temperature data and vibration data of wind power equipment in the inspection sub-area are collected by drones, and the collected images, temperature data and vibration data of wind power equipment are uploaded to the cloud server for processing;

[0012] Preliminary analysis module: by analyzing the images, temperature data and vibration data of wind power equipment in the inspection sub-area, the blade appearance condition index, tower appearance change index, temperature fluctuation index and equipment wear impact index are obtained, including image analysis unit, temperature analysis unit and vibration analysis unit;

[0013] Data fusion module: By performing data fusion calculation on the data analysis results in the preliminary analysis module, the risk assessment coefficient of abnormal equipment condition is obtained;

[0014] Abnormal detection module: Based on the result of data fusion, the abnormal risk assessment coefficient of the equipment condition is compared with the abnormal risk threshold of the equipment condition, so as to detect abnormalities of the wind power equipment in the inspection sub-area, and transmit the abnormal detection results to the fault prediction module;

[0015] Fault prediction module: Based on the detection results of the anomaly detection module, the historical power generation, voltage and current are combined with the real-time power generation, voltage and current for analysis, and the fault prediction coefficient is calculated to predict the faults of abnormal wind power equipment in the inspection sub-area;

[0016] Fault diagnosis module: performs fault diagnosis on abnormal wind power equipment in the inspection sub-area based on the results of fault prediction;

[0017] Remote monitoring module: By setting up a remote monitoring platform, the wind power equipment inspection and monitoring report is automatically generated, and the fault diagnosis results of the wind power equipment are transmitted to the remote monitoring platform through the cloud server to provide early warning feedback to the management personnel.

[0018] Preferably, the inspection area division module divides the wind farm station into several relatively independent inspection sub-areas according to the type of equipment. Each inspection sub-area should contain a certain number of wind power equipment of the same type, and each divided inspection sub-area is numbered, starting from 1 and increasing sequentially to n.

[0019] Preferably, the inspection route planning module selects drones equipped with high-definition cameras, infrared thermal imagers and sensors as intelligent inspection equipment for new energy stations, utilizes the path optimization algorithm in artificial intelligence, and combines the equipment distribution and road conditions in the inspection sub-area to intelligently plan and adjust the inspection routes of drones in real time, and automatically generates the optimal inspection route, and then enables the drones to inspect the inspection sub-areas in the new energy stations according to the planned inspection routes.

[0020] Preferably, the inspection data acquisition module starts the drone and collects inspection data for each inspection sub-area according to the preset inspection route, and then uploads the collected image, temperature data and vibration data to the cloud server in real time through the wireless transmission module of the drone, and performs denoising, image enhancement and data cleaning on the image, temperature data and vibration data through the cloud server. The specific process of data acquisition includes:

[0021] By using drones equipped with high-resolution cameras to fly over wind farms, detailed visual inspections of wind power equipment can be carried out, capturing images of the equipment's appearance in real time. The thermal imagers and vibration sensors installed on the drones can collect the equipment's temperature and vibration data in real time.

[0022] Preferably, the preliminary analysis module monitors the characteristic change trend of the wind power equipment in each inspection sub-area in real time by collaboratively analyzing the image, temperature data and vibration data of the wind power equipment in each inspection sub-area;

[0023] The image analysis unit uses a deep learning algorithm to perform feature analysis on a single device in the device image when the wind power device is running, and calculates the blade appearance condition index C ij and tower appearance change index I ij , thereby automatically identifying blade appearance damage and tower appearance defects;

[0024] The temperature analysis unit analyzes the temperature change trend of a single device in the wind power equipment in the inspection sub-area and calculates the temperature fluctuation index T ij ′;

[0025] The vibration analysis unit analyzes the vibration data of individual wind turbines in the inspection sub-area, including wind speed, wind direction, vibration frequency and external wind pressure, and calculates the equipment wear impact index D ij .

[0026] Preferably, the data fusion module first performs data standardization processing on the blade appearance condition index, tower appearance change index, temperature fluctuation index, and equipment wear impact index, and then performs fusion calculation to obtain the specific calculation formula of the equipment condition abnormality risk assessment coefficient: Q ij =ln(Cij +a)×ln(I ij +a)×ln(T ij ′+a)×ln(D ij +a), C ij Indicates the leaf appearance index, I ij Represents the tower appearance change index, T ij ′ represents the temperature fluctuation index, D ij It represents the equipment wear impact index, and a represents a constant.

[0027] Preferably, the abnormality detection module sets an abnormal risk threshold value θ for each wind power equipment in the inspection sub-area. i , the equipment condition abnormal risk assessment coefficient Q ij and the abnormal risk threshold of equipment condition θ i Compare and detect abnormal wind power equipment in the inspection sub-area; when the equipment condition is abnormal, the risk assessment coefficient Q ij ≤ Equipment condition abnormal risk threshold θ i When the equipment status abnormality risk assessment coefficient Q ij > Equipment condition abnormal risk threshold θ i When , it indicates that there is a single wind power equipment with abnormal problems in the inspection sub-area, and all wind power equipment with abnormal problems in the inspection sub-area are screened out, and the screening results are transmitted to the fault prediction module.

[0028] Preferably, the fault prediction module collects the historical average power generation, average voltage and average current of each abnormal wind power equipment within a period of time and the real-time power generation, voltage and current for difference analysis, and calculates the fault prediction coefficient as G im，实时 represents the real-time power generation currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, R im，实时 represents the real-time current currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, V im，实时 represents the real-time voltage currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, G im，历史 represents the historical average power generation of the mth abnormal wind power equipment in the i-th inspection sub-area, R im，历史 represents the historical average current of the mth abnormal wind power equipment in the ith inspection sub-area, V im，历史 represents the historical average voltage of the mth abnormal wind power equipment in the i-th inspection sub-area, represents the standard deviation of the historical power generation of the mth abnormal wind power equipment in the i-th inspection sub-area, represents the standard deviation of the historical current of the mth abnormal wind power equipment in the i-th inspection sub-area, It represents the standard deviation of the historical voltage of the mth abnormal wind power equipment in the i-th inspection sub-area.

[0029] Preferably, the fault diagnosis module is configured to calculate the fault prediction coefficient H im With the preset fault threshold λ i Compare the results to determine whether there is a fault in the abnormal wind power equipment in the inspection sub-area. If the fault prediction coefficient H im >Preset fault threshold λ i , then there is no fault problem with the abnormal wind power equipment in the inspection sub-area, and it is necessary to focus on monitoring the abnormal wind power equipment. If the fault prediction coefficient H im >Preset fault threshold λ i , it is determined that a single abnormal wind power equipment in the inspection sub-area has a fault problem, and the fault diagnosis result is immediately fed back to the result feedback module.

[0030] Preferably, the remote monitoring module is used to receive automatically generated wind power equipment inspection and monitoring reports and wind power equipment fault diagnosis results transmitted by the cloud server. Management personnel can view the operating status and inspection results of the new energy station anytime and anywhere through the remote monitoring platform, and receive early warning results, thereby realizing remote monitoring and management.

[0031] Technical effects and advantages of the present invention:

[0032] The present invention divides a station into several inspection sub-areas through an inspection area division module and numbers them respectively. The inspection route planning module uses an artificial intelligence algorithm to automatically plan inspection routes for intelligent inspection devices in the inspection sub-areas and conduct inspections. The inspection data acquisition module collects images, temperature data and vibration data of wind power equipment in each inspection sub-area in real time. The preliminary analysis module analyzes the images, temperature data and vibration data of wind power equipment in the inspection sub-area. The data fusion module performs data fusion calculation on the data analysis results in the preliminary analysis module to obtain an equipment condition abnormal risk assessment coefficient. The abnormal detection module compares the equipment condition abnormal risk assessment coefficient with the equipment condition abnormal risk threshold, thereby performing abnormal detection on the wind power equipment in the inspection sub-area. The fault prediction module compares the historical power generation , voltage and current are combined with real-time power generation, voltage and current for analysis, and fault prediction is carried out on abnormal wind power equipment in the inspection sub-area. Fault diagnosis module is used to diagnose faults of abnormal wind power equipment in the inspection sub-area. Remote monitoring platform is set up through remote monitoring module to remotely monitor and manage wind power equipment inspections. By utilizing drones and combining inspection routes planned by artificial intelligence algorithms, it is beneficial to achieve comprehensive coverage and real-time monitoring of wind farm equipment and areas, and improve inspection efficiency and accuracy. By integrating image analysis unit, temperature analysis unit, vibration analysis unit, data fusion module and anomaly detection module, comprehensive analysis and processing of wind power equipment data can be realized, and the accuracy and reliability of data analysis can be improved. Accurate prediction and rapid diagnosis of wind power equipment faults are carried out by combining historical data with real-time data. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 The present invention is a flow chart of a remote monitoring system for a wind farm station based on artificial intelligence. DETAILED DESCRIPTION

[0034] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. In addition, the forms of the various structures recorded in the following embodiments are only examples. The artificial intelligence-based wind farm station remote monitoring system involved in the present invention is not limited to the various structures recorded in the following embodiments. All other implementations obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0035] like Figure 1 The present embodiment provides a remote monitoring system for a wind farm station based on artificial intelligence, including:

[0036] Inspection area division module: The station is divided into several inspection sub-areas according to the type of equipment in the new energy station, and they are numbered 1, 2, 3...n.

[0037] In this embodiment, the inspection area division module divides the wind farm station into several relatively independent inspection sub-areas according to the type of equipment. Each inspection sub-area should contain a certain number of wind power equipment of the same type, and each divided inspection sub-area is numbered, starting from 1 and increasing sequentially until n.

[0038] It should be specifically noted that according to the type of equipment and its distribution in the new energy station, the station is divided into multiple inspection sub-areas. When dividing the inspection sub-areas, the terrain, climate and traffic need to be considered. As the station operation time increases and the equipment ages, the division of inspection sub-areas needs to be dynamically adjusted.

[0039] Inspection route planning module: By using artificial intelligence algorithms to automatically plan inspection routes for intelligent inspection equipment in the divided inspection sub-areas, the new energy stations are inspected according to the planned inspection routes.

[0040] In this embodiment, the inspection route planning module selects a drone equipped with a high-definition camera, an infrared thermal imager and a sensor as the intelligent inspection equipment for the new energy station, and utilizes the path optimization algorithm in artificial intelligence, combined with the equipment distribution and road conditions in the inspection sub-area, to intelligently plan and adjust the inspection route of the drone in real time, and automatically generate the optimal inspection route, and then enable the drone to inspect the inspection sub-area in the new energy station according to the planned inspection route.

[0041] It should be specifically noted that when planning inspection routes, it is necessary to comprehensively consider the endurance of the inspection equipment, the priority of the inspection tasks, and the historical data of equipment failures. If an equipment failure or road congestion occurs, the inspection route must be adjusted in real time through artificial intelligence algorithms.

[0042] Inspection data acquisition module: used to collect images, temperature data and vibration data of wind power equipment in each inspection sub-area. The images, temperature data and vibration data of wind power equipment in the inspection sub-area are collected by drones, and the collected images, temperature data and vibration data of wind power equipment are uploaded to the cloud server for processing.

[0043] In this embodiment, the inspection data acquisition module starts the drone and collects inspection data for each inspection sub-area according to the preset inspection route, and then uploads the collected image, temperature data and vibration data to the cloud server in real time through the wireless transmission module of the drone. The cloud server performs denoising, image enhancement and data cleaning on the image, temperature data and vibration data. The specific process of data acquisition includes:

[0044] By using drones equipped with high-resolution cameras to fly over wind farms, detailed visual inspections of wind power equipment can be carried out, capturing images of the equipment's appearance in real time. The thermal imagers and vibration sensors installed on the drones can collect the equipment's temperature and vibration data in real time.

[0045] It should be specifically explained that inspection tasks, including inspection routes and data collection, are issued to drones through the remote monitoring system. When the drone receives the task, it automatically takes off and flies along the preset route. After the drone arrives at the designated location, it automatically adjusts its posture, starts sensors and high-definition cameras for data collection, and then the cloud server receives the drone's location information and collected data in real time.

[0046] Preliminary analysis module: By analyzing the images, temperature data and vibration data of wind power equipment in the inspection sub-area, the blade appearance condition index, tower appearance change index, temperature fluctuation index and equipment wear impact index are obtained, including image analysis unit, temperature analysis unit and vibration analysis unit.

[0047] In this embodiment, the preliminary analysis module collaboratively analyzes the image, temperature data and vibration data of the wind power equipment in each inspection sub-area, thereby monitoring the characteristic change trend of the wind power equipment in the inspection sub-area in real time;

[0048] The image analysis unit uses a deep learning algorithm to perform feature analysis on the equipment image when the wind power equipment is running, and calculates the blade appearance condition index and the tower appearance change index, so as to automatically identify blade appearance damage and tower appearance defects;

[0049] The temperature analysis unit analyzes the temperature variation trend of the wind power equipment in the inspection sub-area and calculates the temperature fluctuation index;

[0050] The vibration analysis unit analyzes the vibration data of the wind power equipment in the inspection sub-area, including wind speed, wind direction, vibration frequency and external wind pressure, and calculates the equipment wear impact index.

[0051] It should be specifically stated that the specific analysis method of the image analysis unit using the deep learning algorithm for the leaf appearance condition index is as follows:

[0052] Step S411: Identify the total area A of the blade damage region in the image by training the damage recognition model ij ;

[0053] Step S412: By using the deep learning model to compare the difference between the actual shape of the blade and the standard shape, the blade shape deviation is obtained. (x ijk ,yijk ) represents the kth corresponding point on the blade profile of the jth wind turbine in the i-th inspection sub-area, obtained by actual measurement. (x i ' jk ,y i ' jk ) represents the kth point on the ideal blade contour of the jth wind turbine in the i-th inspection sub-area, and K represents the total number of points on the blade contour;

[0054] Step S413: by measuring the blade damage area A of each wind turbine in the inspection sub-area ij and blade shape deviation SI ij Analyze and calculate the leaf appearance index as C ij =w1×A ij +w2×M×SI ij , M represents the total number of blades of a wind turbine;

[0055] The specific analysis method of the image analysis unit using the deep learning algorithm for the tower appearance change index is as follows:

[0056] Step S421: extracting image features by using a pre-trained deep learning model;

[0057] Step S422: comparing the tower image features at different time points to identify appearance changes;

[0058] Step S423: Assume f ij0 and f ij1 are the feature vectors extracted at time points t0 and t1, respectively, so the tower appearance change index is calculated as ||f ij1 -f ij0 ||1 represents vector f ij0 and f ij1 The L1 Manhattan distance of ||f ij1 -f ij0 ||2 represents vector f ij0 and f ij1 The L2 Manhattan distance of

[0059] The specific calculation formula of temperature fluctuation index is: T ij represents the operating temperature of the jth wind turbine in the i-th inspection sub-area, represents the average operating temperature of all wind power equipment in the i-th inspection sub-area;

[0060] The specific calculation formula of equipment wear impact index is D ij =f(v ij , d ij , rij , p i ), v ij represents the wind speed of the jth wind turbine in the i-th inspection sub-area, d ij represents the wind direction of the jth wind turbine in the i-th inspection sub-area, r ij represents the vibration frequency of the jth wind turbine in the i-th inspection sub-area, p i Represents the external wind pressure on the i-th inspection sub-area.

[0061] Data fusion module: By performing data fusion calculation on the data analysis results in the preliminary analysis module, the risk assessment coefficient of abnormal equipment condition is obtained.

[0062] In this embodiment, the data fusion module first performs data standardization processing on the blade appearance condition index, tower appearance change index, temperature fluctuation index, and equipment wear impact index, and then performs fusion calculation to obtain the specific calculation formula of the equipment condition abnormality risk assessment coefficient: Q ij =ln(C ij +a)×ln(I ij +a)×ln(T′ ij +a)×ln(D ij +a), C ij Indicates the leaf appearance index, I ij Represents the tower appearance change index, T′ ij Indicates the temperature fluctuation index, D ij It represents the equipment wear impact index, and a represents a constant.

[0063] Anomaly detection module: Based on the results of data fusion, the equipment condition abnormal risk assessment coefficient is compared with the equipment condition abnormal risk threshold, so as to perform anomaly detection on the wind power equipment in the inspection sub-area, and transmit the anomaly detection results to the fault prediction module.

[0064] In this embodiment, the abnormality detection module sets an abnormal risk threshold value θ for each wind power equipment in the inspection sub-area. i , the equipment condition abnormal risk assessment coefficient Q ij and the abnormal risk threshold of equipment condition θ i Compare and detect abnormal wind power equipment in the inspection sub-area; when the equipment condition is abnormal, the risk assessment coefficient Q ij ≤ Equipment condition abnormal risk threshold θ i When the equipment status abnormality risk assessment coefficient Q ij > Equipment condition abnormal risk threshold θ iWhen , it indicates that there is a single wind power equipment with abnormal problems in the inspection sub-area, and all wind power equipment with abnormal problems in the inspection sub-area are screened out, and the screening results are transmitted to the fault prediction module.

[0065] Fault prediction module: Based on the detection results of the anomaly detection module, the historical power generation, voltage and current are combined with the real-time power generation, voltage and current for analysis, and the fault prediction coefficient is calculated to predict the faults of abnormal wind power equipment in the inspection sub-area.

[0066] In this embodiment, the fault prediction module collects the historical average power generation, average voltage and average current of each abnormal wind power equipment within a period of time and the real-time power generation, voltage and current for difference analysis, and calculates the fault prediction coefficient: G im，实时 represents the real-time power generation currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, R im，实时 represents the real-time current currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, V im，实时 represents the real-time voltage currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, G im，历史 represents the historical average power generation of the mth abnormal wind power equipment in the i-th inspection sub-area, R im，历史 represents the historical average current of the mth abnormal wind power equipment in the ith inspection sub-area, V im，历史 represents the historical average voltage of the mth abnormal wind power equipment in the i-th inspection sub-area, represents the standard deviation of the historical power generation of the mth abnormal wind power equipment in the i-th inspection sub-area, represents the standard deviation of the historical current of the mth abnormal wind power equipment in the i-th inspection sub-area, It represents the standard deviation of the historical voltage of the mth abnormal wind power equipment in the i-th inspection sub-area.

[0067] Fault diagnosis module: performs fault diagnosis on abnormal wind power equipment in the inspection sub-area based on the results of fault prediction.

[0068] In this embodiment, the fault diagnosis module calculates the fault prediction coefficient H im With the preset fault threshold λ i Compare the results to determine whether there is a fault in the abnormal wind power equipment in the inspection sub-area. If the fault prediction coefficient H im >Preset fault threshold λ i , then there is no fault problem with the abnormal wind power equipment in the inspection sub-area, and it is necessary to focus on monitoring the abnormal wind power equipment. If the fault prediction coefficient H im >Preset fault threshold λ i, it is determined that a single abnormal wind power equipment in the inspection sub-area has a fault problem, and the fault diagnosis result is immediately fed back to the result feedback module.

[0069] Remote monitoring module: By setting up a remote monitoring platform, the wind power equipment inspection and monitoring report is automatically generated, and the fault diagnosis results of the wind power equipment are transmitted to the remote monitoring platform through the cloud server to provide early warning feedback to the management personnel.

[0070] In this embodiment, the remote monitoring module is used to receive the automatically generated wind power equipment inspection and monitoring reports and wind power equipment fault diagnosis results transmitted by the cloud server. Management personnel can view the operating status and inspection results of the new energy station anytime and anywhere through the remote monitoring platform, and receive early warning results, thereby realizing remote monitoring and management.

[0071] It should be specifically noted that the fault diagnosis results and inspection and monitoring reports of wind power equipment are automatically generated by the system and then transmitted to the remote monitoring platform through the cloud server. The remote monitoring platform displays real-time data and statistical analysis results, including power generation trend charts, equipment health status curves, and fault alarm responses. Managers can view the operating status, inspection reports, and fault diagnosis results of the wind farm by logging into the platform, and are allowed to remotely execute control commands, including starting and stopping equipment and adjusting parameters, thereby remotely scheduling inspection tasks and viewing inspection progress and results.

[0072] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. 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.

[0073] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A remote monitoring system for wind farm stations based on artificial intelligence, characterized by: include: Inspection area division module: divide the station into several inspection sub-areas according to the equipment type in the new energy station, and number them as 1, 2, 3...n respectively; Inspection route planning module: In the divided inspection sub-areas, the inspection route is automatically planned for the intelligent inspection equipment using artificial intelligence algorithms, and then the new energy stations are inspected according to the planned inspection route; Inspection data collection module: used to collect images, temperature data and vibration data of wind power equipment in each inspection sub-area. The images, temperature data and vibration data of wind power equipment in the inspection sub-area are collected by drones, and the collected images, temperature data and vibration data of wind power equipment are uploaded to the cloud server for processing; Preliminary analysis module: by analyzing the images, temperature data and vibration data of wind power equipment in the inspection sub-area, the blade appearance condition index, tower appearance change index, temperature fluctuation index and equipment wear impact index are obtained, including image analysis unit, temperature analysis unit and vibration analysis unit; Data fusion module: By performing data fusion calculation on the data analysis results in the preliminary analysis module, the risk assessment coefficient of abnormal equipment condition is obtained; Abnormal detection module: Based on the result of data fusion, the abnormal risk assessment coefficient of the equipment condition is compared with the abnormal risk threshold of the equipment condition, so as to detect abnormalities of the wind power equipment in the inspection sub-area, and transmit the abnormal detection results to the fault prediction module; Fault prediction module: Based on the detection results of the anomaly detection module, the historical power generation, voltage and current are combined with the real-time power generation, voltage and current for analysis, and the fault prediction coefficient is calculated to predict the faults of abnormal wind power equipment in the inspection sub-area; Fault diagnosis module: performs fault diagnosis on abnormal wind power equipment in the inspection sub-area based on the results of fault prediction; Remote monitoring module: By setting up a remote monitoring platform, the wind power equipment inspection and monitoring report is automatically generated, and the fault diagnosis results of the wind power equipment are transmitted to the remote monitoring platform through the cloud server to provide early warning feedback to the management personnel.

2. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The inspection area division module divides the wind farm station into several relatively independent inspection sub-areas according to the type of equipment. Each inspection sub-area should contain a certain number of wind power equipment of the same type, and each divided inspection sub-area is numbered, starting from 1 and increasing in sequence until n.

3. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The inspection route planning module selects drones equipped with high-definition cameras, infrared thermal imagers and sensors as intelligent inspection equipment for new energy stations, and uses the path optimization algorithm in artificial intelligence, combined with the equipment distribution and road conditions in the inspection sub-area, to intelligently plan and adjust the inspection routes of drones in real time, and automatically generate the optimal inspection route, and then enable the drones to inspect the inspection sub-areas in the new energy stations according to the planned inspection routes.

4. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The inspection data acquisition module starts the drone and collects inspection data for each inspection sub-area according to the preset inspection route, and then uploads the collected images, temperature data and vibration data to the cloud server in real time through the wireless transmission module of the drone. The cloud server performs denoising, image enhancement and data cleaning on the images, temperature data and vibration data. The specific process of data acquisition includes: By using drones equipped with high-resolution cameras to fly over wind farms, detailed visual inspections of wind power equipment can be carried out, capturing images of the equipment's appearance in real time. The thermal imagers and vibration sensors installed on the drones can collect the equipment's temperature and vibration data in real time.

5. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The preliminary analysis module monitors the characteristic change trend of the wind power equipment in each inspection sub-area in real time by collaboratively analyzing the image, temperature data and vibration data of the wind power equipment in each inspection sub-area; The image analysis unit uses a deep learning algorithm to perform feature analysis on a single device in the device image when the wind power device is running, and calculates the blade appearance condition index C ij and tower appearance change index I ij , thereby automatically identifying blade appearance damage and tower appearance defects; The temperature analysis unit analyzes the temperature change trend of a single device in the wind power equipment in the inspection sub-area and calculates the temperature fluctuation index T ij ′; The vibration analysis unit analyzes the vibration data of individual wind turbines in the inspection sub-area, including wind speed, wind direction, vibration frequency and external wind pressure, and calculates the equipment wear impact index D ij .

6. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The data fusion module first performs data standardization processing on the blade appearance condition index, tower appearance change index, temperature fluctuation index, and equipment wear impact index, and then performs fusion calculation to obtain the specific calculation formula of the equipment condition abnormal risk assessment coefficient: Q ij =ln(C ij +a)×ln(I ij +a)×ln(T′ ij +a)×ln(D ij +a), C ij Indicates the leaf appearance index, I ij Represents the tower appearance change index, T′ ij Indicates the temperature fluctuation index, D ij It represents the equipment wear impact index, and a represents a constant.

7. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The abnormality detection module sets an abnormal risk threshold value θ for each wind power equipment in the inspection sub-area. i , the equipment condition abnormal risk assessment coefficient Q ij and the abnormal risk threshold of equipment condition θ i Comparison is performed to detect abnormal wind power equipment in the inspection sub-area; When the equipment condition is abnormal, the risk assessment coefficient Q ij ≤ Equipment condition abnormal risk threshold θ i When the equipment status abnormality risk assessment coefficient Q ij > Equipment condition abnormal risk threshold θ i When , it indicates that there is a single wind power equipment with abnormal problems in the inspection sub-area, and all wind power equipment with abnormal problems in the inspection sub-area are screened out, and the screening results are transmitted to the fault prediction module.

8. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The fault prediction module collects the historical average power generation, average voltage and average current of each abnormal wind power equipment within a period of time and the real-time power generation, voltage and current for difference analysis, and calculates the fault prediction coefficient as G im，实时 represents the real-time power generation currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, R im，实时 represents the real-time current currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, V im，实时 represents the real-time voltage currently measured by the mth abnormal wind power equipment in the i-th inspection sub-area, G im，历史 represents the historical average power generation of the mth abnormal wind power equipment in the i-th inspection sub-area, R im，历史 represents the historical average current of the mth abnormal wind power equipment in the ith inspection sub-area, V im，历史 represents the historical average voltage of the mth abnormal wind power equipment in the i-th inspection sub-area, represents the standard deviation of the historical power generation of the mth abnormal wind power equipment in the i-th inspection sub-area, represents the standard deviation of the historical current of the mth abnormal wind power equipment in the i-th inspection sub-area, It represents the standard deviation of the historical voltage of the mth abnormal wind power equipment in the i-th inspection sub-area.

9. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The fault diagnosis module calculates the fault prediction coefficient H im With the preset fault threshold λ i Compare the results to determine whether there is a fault in the abnormal wind power equipment in the inspection sub-area. If the fault prediction coefficient H im >Preset fault threshold λ i , then there is no fault problem with the abnormal wind power equipment in the inspection sub-area, and it is necessary to focus on monitoring the abnormal wind power equipment. If the fault prediction coefficient H im >Preset fault threshold λ i , it is determined that a single abnormal wind power equipment in the inspection sub-area has a fault problem, and the fault diagnosis result is immediately fed back to the result feedback module.

10. The wind farm station remote monitoring system based on artificial intelligence according to claim 1 is characterized by: The remote monitoring module is used to receive the automatically generated wind power equipment inspection and monitoring reports and wind power equipment fault diagnosis results transmitted by the cloud server. Management personnel can view the operating status and inspection results of the new energy station anytime and anywhere through the remote monitoring platform, and receive early warning results, thereby realizing remote monitoring and management.

Images