Wind power plant equipment anomaly detection and early warning method, system, equipment and medium

Abstract

The invention discloses a wind power plant equipment anomaly detection and early warning method, system, equipment and medium, and belongs to the technical field of wind power generation, and the method comprises the following steps: collecting operation data of wind power plant equipment, including wind speed data, rotating speed data, vibration data and temperature data; working condition state judgment is carried out based on the wind speed data and the output power data; selecting feature extraction parameters according to the working condition state judgment result, and performing feature extraction on the operation data to obtain feature data; performing anomaly detection on the feature data based on a working condition related health state model to obtain an anomaly score; determining early warning levels according to the abnormal score, wherein the early warning levels comprise first-level early warning, second-level early warning, third-level early warning and fourth-level early warning And generating early warning information according to the early warning level and pushing the early warning information. According to the method, the technical problems of accurate extraction and reliable identification of fault features under the variable working conditions of the wind power equipment are solved, and the accuracy of anomaly detection and the timeliness of early warning are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power generation, and in particular relates to a method, system, equipment and medium for detecting and warning abnormalities of wind farm equipment. Background Art

[0002] As an important part of renewable energy power generation, the reliable operation of wind farm equipment is of great significance to ensuring power supply. Various abnormal conditions may occur in wind farm equipment during operation. If these abnormalities are not discovered and handled in time, they may cause equipment damage, reduce power generation efficiency, and even cause safety accidents. Therefore, abnormal detection and early warning of wind farm equipment is a key link to ensure the safe and stable operation of wind farms.

[0003] At present, abnormal detection of wind farm equipment mainly adopts detection methods based on fixed threshold judgment or analysis methods based on a single data source. These methods usually issue an alarm only after the equipment reaches a preset alarm threshold, or only rely on a certain type of sensor data for analysis. At the same time, existing detection methods often ignore the differences in the operating characteristics of wind power equipment under different working conditions and use a unified detection standard for judgment.

[0004] Because wind power equipment has complex and diverse fault characteristics when operating under variable conditions, and early fault characteristics are often relatively weak, it is difficult to accurately extract and reliably identify fault characteristics using existing technologies, resulting in insufficient accuracy and timeliness of anomaly detection, and a high rate of missed reports and false alarms, which affects the safe operation of the equipment. Summary of the invention

[0005] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide a method for abnormal detection and early warning of wind farm equipment to solve the technical problem of how to accurately extract and reliably identify fault characteristics under variable working conditions of wind power equipment.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: The present invention discloses a method for detecting and warning abnormality of wind farm equipment, comprising the following steps: Collecting operating data of wind farm equipment; Based on the output power data and the wind speed data in the operation data, the working condition status is determined to obtain the working condition status determination result; Select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; Perform anomaly detection on feature data based on the working condition-related health status model to obtain anomaly scores; Determine the warning level based on the anomaly score; Generate and push warning information based on the warning level.

[0007] Preferably, the operating data includes wind speed data, rotation speed data, vibration data and temperature data; Wind speed data is collected by an anemometer; the sampling frequency of wind speed data is 10Hz; The speed data is collected by the speed sensor; the sampling frequency of the speed data is 10Hz; The vibration data is collected by a vibration sensor; the sampling frequency of the vibration data is 20kHz; The temperature data is collected by a temperature sensor; the sampling frequency of the temperature data is 1 Hz.

[0008] Preferably, the operating state includes a startup state, a rated operation state, an overspeed protection state and a shutdown state; When determining the working state, record the working state conversion time and eliminate the interference caused by the working state conversion during feature extraction; The calculation formula for output power data in working condition determination is:

[0009] in, Indicates the wind rotor power in kW. It represents the air density in kg / m³, R represents the radius of the wind wheel in m, Indicates wind speed in m / s. Indicates the power factor.

[0010] Preferably, the characteristic data includes time domain characteristics and frequency domain characteristics; the time domain characteristics include root mean square value, peak value, kurtosis value and skewness value; the frequency domain characteristics include characteristic frequency amplitude and spectrum entropy; The RMS value is calculated using the following formula:

[0011] Among them, RMS represents the root mean square value, N represents the number of sampling points, Indicates The vibration amplitude of each sampling point; The spectrum entropy value is calculated by the following formula:

[0012] in, represents the spectrum entropy value, Indicates The normalized power spectral density of the frequency band.

[0013] Preferably, the working condition related health status model is established by the following steps, including: Collect historical data of equipment during normal operation; Establish characteristic distribution models for different working conditions; Build a dynamic threshold model.

[0014] Preferably, the warning levels include level one warning, level two warning, level three warning and level four warning; The criteria for determining the warning level are: When the anomaly score is between 0.6 and 0.7, it is judged as a first-level warning; When the abnormality score is between 0.7 and 0.8, it is judged as a second-level warning; When the anomaly score is between 0.8 and 0.9, it is judged as a level 3 warning; When the anomaly score is greater than 0.9, it is judged as a level 4 warning.

[0015] Preferably, the warning information includes the abnormality type, abnormality location, abnormality degree and handling suggestions; The push methods of early warning information include: Level 1 and level 2 warnings generate system messages and push them to the monitoring platform; Level 3 warning sends system messages, SMS alerts and email alerts; The fourth-level warning triggers an audible and visual alarm, and sends a system message, SMS alarm, and email alarm.

[0016] The present invention also discloses a wind farm equipment abnormality detection and early warning system, comprising: Operation data collection module, used to collect operation data of wind farm equipment; An operating state determination module is used to determine the operating state based on the output power data and the wind speed data in the operation data to obtain an operating state determination result; A feature extraction module is used to select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; An anomaly detection module is used to detect anomalies in feature data based on a working condition-related health status model to obtain an anomaly score; A warning level determination module is used to determine the warning level according to the anomaly score; The warning information generation module is used to generate and push warning information according to the warning level.

[0017] The present invention also discloses an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the wind farm equipment abnormality detection and early warning method when executing the computer program.

[0018] The present invention also discloses a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the wind farm equipment abnormality detection and early warning method are implemented.

[0019] Compared with the prior art, the present invention has the following beneficial effects: The abnormality detection and early warning method for wind farm equipment provided by the present invention realizes accurate extraction and reliable identification of fault features of wind power equipment under variable working conditions by collecting multi-source operating data and combining the working condition status judgment results to perform feature extraction and abnormality detection. The method adopts a feature extraction parameter selection strategy based on working conditions, which effectively overcomes the influence of working condition changes on feature extraction; by establishing a health status model related to working conditions, the accuracy of abnormality detection is improved; and by adopting a graded early warning mechanism and a variety of information push methods, timely early warning of abnormal conditions is realized. The actual application results show that the abnormality detection accuracy of this method reaches more than 95%, the false alarm rate is less than 2%, and the early warning lead time reaches 12 to 48 hours, which provides sufficient time for operation and maintenance personnel to take preventive measures, significantly improves the operating reliability and maintenance efficiency of wind farm equipment, and effectively reduces the equipment failure rate, which has important engineering application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of the abnormality detection and early warning method of wind farm equipment of the present invention; Figure 2 It is a schematic diagram of the operating state conversion of the wind farm equipment of the present invention. DETAILED DESCRIPTION

[0021] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0022] The following detailed description is an exemplary description, which is intended to provide further detailed description of the present invention. Unless otherwise specified, all technical terms used in the present invention have the same meaning as those generally understood by those skilled in the art to which the present application belongs. The terms used in the present invention are only for describing specific embodiments, and are not intended to limit exemplary embodiments according to the present invention.

[0023] The present invention is further described in detail below in conjunction with the accompanying drawings: In one aspect, the present invention discloses a method for detecting and warning abnormality of wind farm equipment, comprising the following steps: Collect the operating data of wind farm equipment, including wind speed data, rotation speed data, vibration data and temperature data; the sampling frequency for collecting vibration data is 20kHz, the sampling frequency for collecting temperature data is 1Hz, and the sampling frequency for collecting wind speed data and rotation speed data is 10Hz; The operating status is determined based on the wind speed data and output power data. The operating status includes the startup status, rated operation status, overspeed protection status and shutdown status; Select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; Based on the pre-established working condition-related health status model, the feature data is detected for anomalies to obtain an anomaly score; The warning level is determined based on the abnormality score, and the warning levels include level 1 warning, level 2 warning, level 3 warning and level 4 warning; Warning information is generated and pushed according to the warning level. The warning information includes the abnormality type, abnormality location, abnormality degree and handling suggestions.

[0024] Furthermore, the feature extraction step includes: Extracting time domain features, which include root mean square value, peak value, kurtosis value and skewness value; Extract frequency domain features, which include characteristic frequency amplitude and spectrum entropy value.

[0025] Furthermore, the RMS value is calculated by the following formula:

[0026] Among them, RMS represents the root mean square value, N represents the number of sampling points, Indicates The vibration amplitude of each sampling point.

[0027] Furthermore, the spectrum entropy value is calculated by the following formula:

[0028] in, represents the spectrum entropy value, Indicates The normalized power spectral density of the frequency band.

[0029] Furthermore, the steps of establishing the working condition related health status model include: Collect historical data of equipment during normal operation; Establish characteristic distribution models for different working conditions; Build a dynamic threshold model.

[0030] Furthermore, the criteria for determining the warning level are: When the anomaly score is between 0.6 and 0.7, it is judged as a first-level warning; When the abnormality score is between 0.7 and 0.8, it is judged as a second-level warning; When the anomaly score is between 0.8 and 0.9, it is judged as a level 3 warning; When the anomaly score is greater than 0.9, it is judged as a level 4 warning.

[0031] Furthermore, the push methods of early warning information include: Level 1 and level 2 warnings generate system messages and push them to the monitoring platform; Level 3 warning sends system messages, SMS alerts and email alerts; The fourth-level warning triggers an audible and visual alarm, and sends a system message, SMS alarm, and email alarm.

[0032] Furthermore, high-precision sensors are used to collect operating data, including: Install vibration sensors to collect vibration data; Install temperature sensors to collect temperature data; Install anemometer to collect wind speed data; Install a speed sensor to collect speed data.

[0033] Furthermore, the calculation formula for the output power data in the working condition determination is:

[0034] in, Indicates the wind rotor power in kW. It represents the air density in kg / m³, R represents the radius of the wind wheel in m, Indicates wind speed in m / s. Indicates the power factor.

[0035] Furthermore, in the operating state determination step, the operating state conversion moment is recorded, and interference caused by the operating state conversion is eliminated during feature extraction.

[0036] The present invention provides a method for detecting and warning abnormalities of wind farm equipment, which realizes abnormality detection and warning of wind farm equipment under variable working conditions through feature extraction and multi-level warning mechanism that are adaptive to working conditions.

[0037] The present invention discloses a method for detecting and warning abnormalities of wind farm equipment, and the steps are summarized as follows: The first step is data collection. Collect the operation data of wind farm equipment, including wind speed data, rotation speed data, vibration data and temperature data; the sampling frequency of vibration data is 20kHz, the sampling frequency of temperature data is 1Hz, and the sampling frequency of wind speed and rotation speed data is 10Hz. All kinds of data are obtained in real time through high-precision sensors.

[0038] The second step is the working condition determination step. Based on the collected wind speed data and output power data, the working condition status is determined to identify the current working condition of the equipment; the working condition status includes the startup state, rated operation state, overspeed protection state and shutdown state; at the same time, the working condition conversion time is recorded to provide a basis for subsequent analysis.

[0039] The third step is feature extraction. According to the working condition judgment results, select the corresponding feature extraction parameters to extract features from the collected operating data; the extracted features include time domain features and frequency domain features; the time domain features include root mean square value, peak value, kurtosis value and skewness value, and the frequency domain features include characteristic frequency amplitude and spectrum entropy.

[0040] The fourth step is the anomaly detection step. Based on the pre-established working condition-related health status model, the extracted feature data is detected for anomalies; the anomaly score is generated by calculating the deviation between the feature data and the health baseline; the anomaly score is compared with the dynamically set threshold to determine the equipment status.

[0041] The fifth step is to determine the warning level. The warning level is determined according to the anomaly score. The warning is divided into four levels: anomaly scores between 0.6-0.7 are judged as level 1 warning; anomaly scores between 0.7-0.8 are judged as level 2 warning; anomaly scores between 0.8-0.9 are judged as level 3 warning; anomaly scores greater than 0.9 are judged as level 4 warning.

[0042] Step 6: Push warning information. Use the corresponding information push method according to the warning level: Level 1 and Level 2 warnings generate system messages and push them to the monitoring platform; Level 3 warnings send SMS and email warnings in addition to system messages; Level 4 warnings trigger sound and light alarms, and send emergency warning information through SMS, email and other methods.

[0043] Another object of the present invention is to provide a wind farm equipment abnormality detection and early warning system, comprising: Operation data collection module, used to collect operation data of wind farm equipment; An operating state determination module is used to determine the operating state based on the output power data and the wind speed data in the operation data to obtain an operating state determination result; A feature extraction module is used to select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; An anomaly detection module is used to detect anomalies in feature data based on a working condition-related health status model to obtain an anomaly score; A warning level determination module is used to determine the warning level according to the anomaly score; The warning information generation module is used to generate and push warning information according to the warning level.

[0044] The third object of the present invention is to provide an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the wind farm equipment abnormality detection and early warning method when executing the computer program.

[0045] The wind farm equipment abnormality detection and early warning method comprises the following steps: Collecting operating data of wind farm equipment; Based on the output power data and the wind speed data in the operation data, the working condition status is determined to obtain the working condition status determination result; Select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; Perform anomaly detection on feature data based on the working condition-related health status model to obtain anomaly scores; Determine the warning level based on the anomaly score; Generate and push warning information based on the warning level.

[0046] A fourth object of the present invention is to provide a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the wind farm equipment abnormality detection and early warning method are implemented.

[0047] The wind farm equipment abnormality detection and early warning method comprises the following steps: Collecting operating data of wind farm equipment; Based on the output power data and the wind speed data in the operation data, the working condition status is determined to obtain the working condition status determination result; Select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; Perform anomaly detection on feature data based on the working condition-related health status model to obtain anomaly scores; Determine the warning level based on the anomaly score; Generate and push warning information based on the warning level.

[0048] Figure 1 Flow chart of the abnormality detection and early warning method of wind farm equipment of the present invention; Figure 1As shown, the method first collects the operating data of the wind power equipment, including wind speed data, rotation speed data, vibration data and temperature data. The data acquisition equipment uses a high-precision sensor, and the sampling frequency is set according to different types of data, where the vibration data sampling frequency is 20kHz, the temperature data sampling frequency is 1Hz, and the wind speed data and rotation speed data sampling frequency are 10Hz.

[0049] In the working condition determination stage, the system divides the working conditions based on the collected wind speed data and output power data. There is the following relationship between the output power data and wind speed data of the wind power equipment:

[0050] in, Indicates the wind rotor power in kW. It represents the air density in kg / m³, R represents the radius of the wind wheel in m, Indicates wind speed in m / s. Represents the power coefficient, which is a dimensionless parameter. This formula can be used to establish a power curve model for operating condition determination.

[0051] Figure 2 Schematic diagram of the state transition of the wind farm equipment in the present invention; Figure 2 As shown in the figure, wind power equipment has multiple operating conditions during operation, such as startup state, rated operation state, overspeed protection state and shutdown state. The system determines the current operating state by real-time monitoring of the correspondence between wind speed data and output power data, and records the operating condition conversion time, providing a basis for subsequent feature extraction.

[0052] The feature extraction stage includes the extraction of time domain features and frequency domain features. The time domain features mainly include root mean square value, peak value, kurtosis value and skewness value. The root mean square value is calculated using the following formula:

[0053] Among them, RMS represents the root mean square value, N represents the number of sampling points, Indicates The root mean square value reflects the energy level of the vibration signal and provides an important basis for fault diagnosis.

[0054] Frequency domain features are obtained by performing spectrum analysis on the signal, mainly including characteristic frequency amplitude and spectrum entropy. The calculation formula of spectrum entropy is:

[0055] in, represents the spectrum entropy value, Indicates The normalized power spectral density of frequency bands, n represents the number of frequency bands; the spectrum entropy reflects the complexity of the spectrum, and its numerical change can indicate the change of device status.

[0056] The anomaly detection module first establishes a baseline model of health status. This model is constructed by collecting historical data of normal operation of the equipment and establishing corresponding feature distribution models for different working conditions. In the actual detection process, the currently extracted features are compared with the baseline model to calculate the anomaly score. The calculation of the anomaly score takes into account the size and duration of the feature deviation, and realizes the distinction between instantaneous fluctuations and persistent anomalies.

[0057] Warning level determination Four warning levels are set based on the anomaly score.

[0058] When the anomaly score is between 0.6 and 0.7, the system issues a level 1 warning, indicating that the device has a minor anomaly; When the anomaly score is between 0.7 and 0.8, a second-level warning is issued, indicating that the device has obvious anomalies; When the anomaly score is between 0.8 and 0.9, a level 3 warning is issued, indicating that the device has a serious anomaly; When the anomaly score exceeds 0.9, the system issues a level 4 warning, indicating that the equipment is in a dangerous state.

[0059] The warning information push adopts a hierarchical push strategy. For level 1 and level 2 warnings, the system generates warning messages and pushes them to the monitoring platform, while recording abnormal information; for level 3 warnings, in addition to pushing system messages, warning information will also be sent to relevant personnel via SMS and email; when a level 4 warning occurs, the system will trigger an audible and visual alarm, and send emergency warning information to relevant personnel via SMS, email, and other methods.

[0060] Application examples: The specific implementation process of the method of the present invention in the actual application of a certain wind power plant is as follows: The wind farm has an installed capacity of 100MW, with a total of 50 wind turbines, model XX-2000, and a rated power of 2MW. One of the wind turbines with a running time of 5 years was selected for application verification of the abnormal detection and early warning system. The generator set is equipped with the following sensors: 8 vibration sensors, installed at key positions of the main bearing, gearbox and generator; 12 temperature sensors, distributed in various key components; 1 anemometer, installed on the top of the nacelle; 2 speed sensors, installed on the main shaft and generator shaft respectively.

[0061] During the 6-month operation verification process, the system detected three equipment abnormal events: The first abnormal event occurred in the middle of a certain month. Two days before the abnormality occurred, the system detected an abnormal change in the spectral entropy value of the gearbox vibration signal, with an abnormal score of 0.65, triggering a level 1 warning. After receiving the warning, the operation and maintenance personnel strengthened monitoring and found that the gearbox lubricant was slightly contaminated. After the lubricant was replaced in time, the vibration characteristics returned to normal, successfully preventing a possible gearbox failure.

[0062] The second abnormal event occurred one month later. That morning, the system detected that the temperature of the generator bearing continued to rise and the vibration characteristics were abnormal. The abnormality score reached 0.82, triggering a third-level warning. The system notified the operation and maintenance personnel in a timely manner through SMS and email. Inspection found that the bearing was severely worn, and the bearing was replaced in time to avoid unplanned equipment downtime.

[0063] The third abnormal event occurred two months later. The system detected a sudden increase in the vibration amplitude of the main bearing on the night before the abnormality occurred. At the same time, the temperature rose rapidly, and the abnormal score exceeded 0.9, triggering a level 4 warning. The system immediately sounded the alarm in various ways. After receiving the alarm, the on-duty personnel immediately stopped the machine for inspection and found that the main bearing was seriously damaged. Timely treatment avoided a major accident of the equipment.

[0064] During the 6-month verification period, the system issued 15 warning messages, including: Level 1 warnings were issued 8 times, and upon inspection, all of them were minor equipment abnormalities; There were 4 Level 2 warnings, 3 of which were actual anomalies and 1 was a false alarm; Level 3 warnings were issued twice, all of which were actual abnormalities; There was one Level 4 warning, which was an actual serious abnormality.

[0065] From the above examples, it can be seen that the method of the present invention can effectively identify various abnormal conditions of wind power equipment and issue early warning information in a timely manner. In particular, for major faults, the system can issue early warnings more than 12 hours in advance, providing sufficient time for operation and maintenance personnel to take preventive measures, significantly improving the reliability of equipment operation.

[0066] The wind farm equipment abnormality detection and early warning method of the present invention realizes accurate identification and timely early warning of abnormal conditions of wind power equipment through multi-source data analysis and working condition adaptive feature extraction method, combined with a multi-level early warning mechanism. In actual wind farm applications, the method has an abnormality detection accuracy rate of more than 95%, a false alarm rate of less than 2%, and an early warning lead time of 12 to 48 hours, effectively improving the operational reliability and maintenance efficiency of wind farm equipment.

[0067] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0068] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0069] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0070] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for detecting and warning abnormalities of wind farm equipment, characterized in that: The following steps are involved: Collecting operating data of wind farm equipment; Based on the output power data and the wind speed data in the operation data, the working condition status is determined to obtain the working condition status determination result; Select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; Perform anomaly detection on feature data based on the working condition-related health status model to obtain anomaly scores; Determine the warning level based on the anomaly score; Generate and push warning information based on the warning level.

2. The wind farm equipment abnormality detection and early warning method according to claim 1 is characterized in that: The operation data includes wind speed data, rotation speed data, vibration data and temperature data; The wind speed data is collected by an anemometer; the sampling frequency of the wind speed data is 10 Hz; The speed data is collected by a speed sensor; the sampling frequency of the speed data is 10 Hz; The vibration data is collected by a vibration sensor; the sampling frequency of the vibration data is 20kHz; The temperature data is collected by a temperature sensor; the sampling frequency of the temperature data is 1 Hz.

3. The wind farm equipment abnormality detection and early warning method according to claim 1 is characterized in that: The operating status includes a startup status, a rated operation status, an overspeed protection status and a shutdown status; When determining the working state, record the working state conversion time and eliminate the interference caused by the working state conversion during feature extraction; The calculation formula for the output power data in the working condition determination is: in, Indicates the wind rotor power in kW. It represents the air density in kg / m³, R represents the radius of the wind wheel in m, Indicates wind speed in m / s. Indicates the power factor.

4. The wind farm equipment abnormality detection and early warning method according to claim 1, characterized in that: The characteristic data includes time domain characteristics and frequency domain characteristics; The time domain features include root mean square value, peak value, kurtosis value and skewness value; the frequency domain features include characteristic frequency amplitude and spectrum entropy; The RMS value is calculated by the following formula: Among them, RMS represents the root mean square value, N represents the number of sampling points, Indicates The vibration amplitude of each sampling point; The spectrum entropy value is calculated by the following formula: in, represents the spectrum entropy value, Indicates The normalized power spectral density of the frequency band.

5. The wind farm equipment abnormality detection and early warning method according to claim 1, characterized in that: The working condition related health state model is established by the following steps, including: Collect historical data of equipment during normal operation; Establish characteristic distribution models for different working conditions; Build a dynamic threshold model.

6. The wind farm equipment abnormality detection and early warning method according to claim 1, characterized in that: The warning levels include level one warning, level two warning, level three warning and level four warning; The criteria for determining the warning level are: When the anomaly score is between 0.6 and 0.7, it is judged as a first-level warning; When the abnormality score is between 0.7 and 0.8, it is judged as a second-level warning; When the anomaly score is between 0.8 and 0.9, it is judged as a level 3 warning; When the anomaly score is greater than 0.9, it is judged as a level 4 warning.

7. The wind farm equipment abnormality detection and early warning method according to claim 1, characterized in that: The warning information includes the abnormality type, abnormality location, abnormality degree and handling suggestions; The push methods of the warning information include: Level 1 and level 2 warnings generate system messages and push them to the monitoring platform; Level 3 warning sends system messages, SMS alerts and email alerts; The fourth-level warning triggers an audible and visual alarm, and sends a system message, SMS alarm, and email alarm.

8. A wind farm equipment abnormality detection and early warning system, characterized in that: include: Operation data collection module, used to collect operation data of wind farm equipment; An operating state determination module is used to determine the operating state based on the output power data and the wind speed data in the operation data to obtain an operating state determination result; A feature extraction module is used to select feature extraction parameters according to the working condition determination result, perform feature extraction on the operating data, and obtain feature data; An anomaly detection module is used to detect anomalies in feature data based on a working condition-related health status model to obtain an anomaly score; A warning level determination module is used to determine the warning level according to the anomaly score; The warning information generation module is used to generate and push warning information according to the warning level.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the abnormality detection and early warning method for wind farm equipment according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the abnormality detection and early warning method for wind farm equipment according to any one of claims 1 to 7.

Images