Wind turbine generator monitoring and early warning system based on Internet of Things sensing technology

Abstract

The invention discloses a wind turbine generator monitoring and early warning system based on the Internet of Things sensing technology, and relates to the technical field of the Internet of Things, and the system comprises a data collection module which collects the operation state information of key parts of a wind turbine generator in real time, and transmits the operation state data to an edge calculation unit; the prediction module is used for identifying an abnormal mode, analyzing time domain and frequency domain characteristics of the vibration signal in combination with fast Fourier transform, detecting a mechanical fault, and obtaining a residual life prediction value according to a residual life prediction model; the preliminary early warning module is used for determining an equipment health state score based on fuzzy reasoning, and performing preliminary early warning locally according to the equipment health state score; and the grading early warning module is used for carrying out grading alarm on the wind turbine generator equipment according to the mechanical fault condition in combination with the equipment health state score, and providing maintenance decision suggestions. By integrating the Internet of Things technology, edge computing and artificial intelligence, intelligent monitoring and early warning of the wind turbine generator are achieved, and the reliability and safety of equipment operation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and more specifically, to a wind turbine monitoring and early warning system based on Internet of Things sensing technology. Background Art

[0002] As an important component of clean energy, wind power generation has developed rapidly in recent years. However, due to the long-term operation of wind turbines at high altitudes, offshore or in remote areas, their equipment maintenance and monitoring face many challenges. The core components of wind turbines (such as gearboxes, main shafts, generators, pitch control systems, etc.) are subjected to complex mechanical loads, environmental influences (such as high wind speeds, temperature changes, humidity corrosion) and electrical stress for a long time, and are prone to fatigue damage, lubrication failure, electrical failure and other problems. These problems not only affect the operating efficiency of wind turbines, but may also cause significant economic losses and safety hazards.

[0003] Deficiencies in the existing technology: Traditional wind turbine operation and maintenance mainly rely on manual inspections and regular maintenance, but due to the large number of wind turbines and their wide distribution, manual operation and maintenance methods are costly, inefficient, and difficult to detect potential faults in a timely manner. Therefore, how to use emerging technologies such as the Internet of Things, artificial intelligence (AI), and edge computing to improve the intelligent monitoring capabilities of wind turbines, reduce unplanned downtime, and optimize maintenance strategies has become a key issue that needs to be urgently addressed in the wind power industry. The present invention provides a wind turbine monitoring and early warning system based on Internet of Things sensing technology, which integrates Internet of Things technology, edge computing, artificial intelligence, and remote expert systems to achieve intelligent monitoring and early warning of wind turbines and improve the reliability and safety of equipment operation.

[0004] In view of the above problems, the present invention proposes a solution. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a wind turbine monitoring and early warning system based on the Internet of Things sensing technology, which solves the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A wind turbine monitoring and early warning system based on IoT sensing technology, comprising:

[0008] Data acquisition module: collects the operating status information of key components of the wind turbine in real time, including vibration signals and temperature data, and uses low-power wireless communication technology to send the operating status data to the edge computing unit;

[0009] Prediction module: Performs real-time analysis of edge computing devices based on neural network algorithms, identifies abnormal patterns, analyzes the time and frequency domain characteristics of vibration signals in combination with fast Fourier transform (FTT), detects mechanical failures, and obtains the remaining life prediction value based on the remaining life prediction (RUL) model;

[0010] Preliminary warning module: through the analysis of remaining life prediction and abnormal mode conditions, the equipment health status score is determined based on fuzzy reasoning, and preliminary warning is made locally based on the equipment health status score;

[0011] Grading warning module: Based on the mechanical failure situation and the equipment health status score, the wind turbine is remotely monitored and analyzed through the remote diagnosis system. According to the monitoring and analysis results, the wind turbine equipment is graded and alarmed, and maintenance decision suggestions and response measures are provided.

[0012] In a preferred embodiment, the process of real-time collection of operating status information of key components of a wind turbine is as follows:

[0013] Select an accelerometer suitable for vibration measurement of key components of wind turbines and install it at appropriate locations of key components to ensure that vibration signals can be accurately captured;

[0014] Amplify and filter the analog signal output by the acceleration sensor to improve the signal quality and meet the input requirements of the analog-to-digital converter (ADC);

[0015] Install the temperature sensor on the surface of key components to monitor temperature changes in real time; linearize and amplify the analog signal output by the temperature sensor and send it to the ADC for digital conversion;

[0016] Use ADC to convert analog signals into digital signals for subsequent processing and transmission.

[0017] In a preferred embodiment, the process of identifying abnormal patterns is as follows:

[0018] After preprocessing the operation status information, a convolutional neural network is used to analyze it and capture the local features of the signal;

[0019] Use the cross entropy loss function as the loss function of the convolutional neural network and update the parameters of the neural network through the SGD optimizer;

[0020] Use the training set to train the neural network model and iteratively update the model parameters to minimize the value of the loss function;

[0021] The preprocessed running status information is input into the trained neural network model, and the model outputs the probability that each sample belongs to the abnormal mode;

[0022] Based on the comparison between the abnormal pattern probability and the set abnormal threshold, it is determined whether the sample is an abnormal pattern.

[0023] In a preferred embodiment, the process of detecting mechanical failure is as follows:

[0024] Combining the time domain and frequency domain features obtained by FFT analysis with the output results of the neural network model, it is possible to determine whether there are mechanical failures in the key components of the wind turbine.

[0025] If the main frequency of the vibration signal changes abnormally and the abnormal probability output by the neural network model exceeds the threshold, it is considered that a mechanical fault exists.

[0026] In a preferred embodiment, the process of obtaining the remaining life prediction value is as follows:

[0027] Select the time domain and frequency domain features of the vibration signal as the input of the RUL model;

[0028] Use historical operation data and corresponding remaining life labels to train the RUL model;

[0029] After the training is completed, the real-time collected operating status information is input into the RUL model to output the predicted value of the remaining life of the equipment.

[0030] In a preferred embodiment, the steps of determining the equipment health status score based on fuzzy reasoning are as follows:

[0031] The abnormal mode probability and the remaining life prediction value are defined as input variables and divided into different fuzzy sets respectively;

[0032] The equipment health status score is defined as the output variable, divided into fuzzy sets, and fuzzy rules are formulated to describe the impact of abnormal mode probability and remaining life prediction value on the equipment health status score;

[0033] Fuzzy reasoning is performed based on fuzzy rules to determine the equipment health status score.

[0034] In a preferred embodiment, the process of making a preliminary warning locally based on the equipment health status score is as follows:

[0035] When the device health score is greater than the preset health threshold, it means that the device is in good health and no preliminary warning is required.

[0036] When the device health score is less than the preset health threshold, it means that the device health is poor and a preliminary warning is required.

[0037] In a preferred embodiment, the process of remote monitoring and analyzing the wind turbine generator set is as follows:

[0038] According to the mechanical failure situation, the number of failures within a certain time period is obtained, and the equipment failure rate is obtained by dividing the number of failures by the time period. Combined with the equipment health status score, the monitoring and evaluation coefficient is calculated based on the preset monitoring and evaluation formula.

[0039] In a preferred embodiment, the process of performing graded alarm on wind turbine equipment according to the monitoring and analysis results is as follows:

[0040] Compare and analyze the monitoring evaluation coefficient with the preset first evaluation threshold and the second evaluation threshold, and the first evaluation threshold is less than the second evaluation threshold;

[0041] If the monitoring evaluation coefficient is greater than the preset second evaluation threshold, the alarm level is the warning level;

[0042] If the monitoring evaluation coefficient is greater than the preset first evaluation threshold and less than the second evaluation threshold, the alarm level is a warning level;

[0043] If the monitoring evaluation coefficient is less than the preset first evaluation threshold, the alarm level is an emergency level.

[0044] In a preferred embodiment, the process of providing maintenance decision suggestions and response measures is as follows:

[0045] If the alarm level is the early warning level, it means that there is a potential risk of failure in the equipment, but it has not yet affected normal operation. The operating data of related components should be continuously monitored to ensure that problems are discovered in time, and regular inspections should be arranged to check the possible sources of the failure;

[0046] If the alarm level is a warning level, it means that the equipment has experienced a serious abnormality and measures need to be taken as soon as possible to conduct a detailed inspection and fault diagnosis of the alarm components, adjust the operating conditions to delay the occurrence of faults and reduce risks;

[0047] If the alarm level is emergency, it means that the equipment has a serious fault. The unit should be stopped immediately and a maintenance team should be quickly organized to conduct on-site fault diagnosis and repair to ensure that the equipment resumes normal operation.

[0048] The technical effects and advantages of the wind turbine monitoring and early warning system based on IoT sensing technology of the present invention are as follows:

[0049] 1. The present invention effectively solves the problems of slow response, high data transmission cost, and insufficient fault prediction ability of traditional wind turbine monitoring systems through edge computing, health assessment, multimodal sensor data fusion and remote intelligent operation and maintenance. Since the traditional SCADA system relies on centralized processing, there is a large delay in data transmission and analysis. The present invention can run the AI ​​health assessment model locally through a high-performance edge computing device to achieve millisecond-level response, greatly improve the real-time performance of wind turbine fault monitoring, and reduce the impact of sudden failures on wind turbine operation. The traditional cloud-based PHM system needs to transmit a large amount of raw data, resulting in excessive bandwidth consumption. The present invention adopts an event-triggered data upload mechanism to upload only key feature data, reduce communication costs, and improve the efficiency and reliability of the system. In addition, most existing monitoring schemes rely on a single sensor type and cannot accurately identify complex fault modes. The present invention improves the accuracy of fault identification and reduces false alarms and missed reports through the fusion of multimodal data such as vibration, temperature, current, acoustics, and vision, combined with an adaptive AI algorithm.

[0050] 2. The present invention supports running health assessment models locally by adopting high-performance embedded chips, reduces dependence on the cloud, and improves prediction accuracy and response speed. Run AI algorithms based on FFT and CNN to perform intelligent prediction and fault warning on the operating status of wind turbines. Local computing + remote optimization, the equipment can independently perform predictive maintenance analysis, and support remote expert system optimization AI model to improve system adaptability. By adopting multimodal sensor data fusion technology, the accuracy of fault identification is improved, and false alarms and missed alarms are reduced. The equipment has a time synchronization mechanism to ensure the timeliness and consistency of various sensor data and improve monitoring accuracy. Edge computing is used to perform data analysis and fault detection locally, achieve millisecond-level response, reduce dependence on remote servers, and improve system reliability and independence. Data preprocessing and noise reduction are performed locally through edge computing devices, only key information is uploaded, bandwidth consumption is reduced, and an event-triggered upload mechanism is adopted to upload detailed data only when an abnormality is detected, optimizing network utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 The present invention is a schematic diagram of the structure of a wind turbine monitoring and early warning system based on Internet of Things sensing technology. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] Embodiment 1, Figure 1 The present invention provides a wind turbine monitoring and early warning system based on Internet of Things sensing technology.

[0054] S10, collecting operating status information of key components of the wind turbine in real time, the operating status information including vibration signals and temperature data, and sending the operating status data to an edge computing unit using low-power wireless communication technology;

[0055] Connect to various IoT sensors in real time to collect the operating status information of wind turbines, and select accelerometers suitable for vibration measurement of key components of wind turbines, such as piezoelectric accelerometers. Install the sensors in appropriate locations of key components (such as gearboxes and generator bearings) to ensure that vibration signals can be accurately captured; the analog signals output by accelerometers usually need to be amplified, filtered, and conditioned to improve signal quality and meet the input requirements of analog-to-digital converters (ADCs);

[0056] Use temperature sensors, such as thermocouples or thermistors, installed on the surface or inside key components to monitor temperature changes in real time; the analog signal output by the temperature sensor needs to be linearized, amplified, and then sent to the ADC for digital conversion;

[0057] Use a suitable ADC to convert the analog signal into a digital signal for subsequent processing and transmission.

[0058] Use a microcontroller (such as a single-chip microcomputer) to control the ADC for data acquisition, and temporarily store the acquired data in an on-chip or external memory. According to the Nyquist sampling theorem, the sampling frequency should be at least twice the highest frequency of the signal;

[0059] Select appropriate technologies, including Bluetooth, ZigBee, and LoRa, according to actual application scenarios and requirements, connect the selected wireless communication module to the microcontroller, and perform corresponding configurations, including setting the communication protocol, frequency band, and baud rate. The microcontroller encapsulates the collected operating status data in the format of the wireless communication protocol and sends it to the edge computing unit through the wireless module;

[0060] The edge computing unit receives data through the wireless communication module, parses it according to the communication protocol, and restores the original operating status data.

[0061] S20, performs real-time analysis on edge computing devices based on a neural network algorithm, identifies abnormal patterns, analyzes time and frequency domain characteristics of vibration signals in combination with a fast Fourier transform (FFT), detects mechanical failures, and obtains a remaining life prediction value based on a remaining life prediction (RUL) model;

[0062] After preprocessing the running status information, the convolutional neural network (CNN) is used to analyze and capture the local features of the signal;

[0063] The cross entropy loss function is used as the loss function of the convolutional neural network to identify abnormal patterns. The calculation formula of the cross entropy loss function is as follows: Where L is the loss function value, N is the number of samples, C is the number of categories, and y is ij is the true label of the i-th sample, p ij is the predicted probability that the i-th sample belongs to the j-th class;

[0064] Update the parameters of the neural network through the SGD optimizer;

[0065] Use the training set to train the neural network model and iteratively update the model parameters to minimize the value of the loss function;

[0066] The preprocessed running status information is input into the trained neural network model, and the model outputs the probability that each sample belongs to an abnormal mode; based on the comparison between the abnormal mode probability and the set abnormal threshold, it is determined whether the sample is an abnormal mode.

[0067] Analyze the time domain characteristics of the vibration signal according to the fast Fourier transform, perform time domain analysis on the vibration signal, and calculate the mean value, root mean square value, and peak value of the vibration signal;

[0068] Use Fast Fourier Transform (FFT) to convert the time domain vibration signal into a frequency domain signal, analyze the frequency components of the vibration signal, and obtain the frequency domain characteristics of the vibration signal;

[0069] The fast Fourier transform formula is as follows: In the formula, x(n) is the time domain signal, X(k) is the frequency domain signal, and N is the number of samples;

[0070] Combining the time domain and frequency domain characteristics obtained by FFT analysis with the output results of the neural network model, it is determined whether there is a mechanical failure in the key components of the wind turbine. If the main frequency of the vibration signal changes abnormally and the abnormal probability output by the neural network model exceeds the threshold, it is considered that a mechanical failure exists.

[0071] Select the time domain and frequency domain features of the vibration signal as the input of the RUL model;

[0072] Use historical operation data and corresponding remaining life labels to train the RUL model;

[0073] After the training is completed, the real-time collected operating status information is input into the RUL model to output the predicted value of the remaining life of the equipment.

[0074] Use the test set to evaluate the abnormal pattern recognition model and RUL model, adjust the learning rate of the neural network model, the number of hidden layer neurons, or select more appropriate features to improve the performance of the model.

[0075] S30, by analyzing the remaining life prediction and abnormal mode conditions, determining the equipment health status score based on fuzzy reasoning, and making a preliminary warning locally according to the equipment health status score;

[0076] The steps of determining the equipment health status score based on fuzzy reasoning are as follows:

[0077] In step C1, the abnormal mode probability and the remaining life prediction value are defined as input variables, and they are divided into different fuzzy sets respectively.

[0078] For example, "Low", "Medium", "High" for abnormal mode probabilities and "Low", "Medium", "High" for remaining life prediction values.

[0079] Step C2, define the equipment health status score as an output variable and divide it into fuzzy sets, for example, "Low", "High", for the equipment health status score.

[0080] Step C3, formulate a set of fuzzy rules to describe the impact of different input variables on output variables. The definition of rules can be based on professional knowledge or obtained through data analysis and experiments. For example:

[0081] Mark the abnormal mode probability as I, the remaining life prediction value as B, and the equipment health status score as W, then we can define

[0082] Rule 1:IF(I is Low)AND(B is High)THEN(W is High)

[0083] Rule 2:IF(I is High)AND(B is Low)THEN(W is Low) ...

[0085] Step C4, performing fuzzy reasoning according to fuzzy rules to determine the equipment health status score.

[0086] It should be noted that the division of fuzzy sets can be adjusted according to actual conditions. For example, although this embodiment takes three fuzzy sets as examples, in fact, the abnormal mode probability and the remaining life prediction value and the equipment health status score can be divided into more than three sets to facilitate better precise adjustment.

[0087] Furthermore, for the judgment of the equipment health status score, a threshold can be set according to the actual situation. For example, when the probability of abnormal mode exceeds 0.6, it is marked as "High", and when the remaining life prediction value is higher than 80 years, it is marked as "High", etc., which will not be elaborated here.

[0088] When the device health score is greater than the preset health threshold, it means that the device is in good health and no preliminary warning is required.

[0089] When the device health score is less than the preset health threshold, it means that the device health is poor and a preliminary warning is required.

[0090] S40, based on the mechanical failure situation and the equipment health status score, remotely monitors and analyzes the wind turbine through the remote diagnosis system, issues graded alarms for the wind turbine equipment based on the monitoring and analysis results, and provides maintenance decision suggestions and response measures;

[0091] According to the mechanical failure situation, the number of failures within a certain time period is obtained, and the number of failures is divided by the time period to obtain the equipment failure rate. Combined with the equipment health status score, the monitoring and evaluation coefficient is calculated based on the preset monitoring and evaluation formula. The specific calculation formula is as follows: P = u 1 *W+u 2 *Y; where P is the monitoring evaluation coefficient, W is the equipment health status score, and u 1 is the equipment health status score weight factor, Y is the equipment failure rate, u 2 is the equipment failure rate weight factor, u 1 greater than 0, u 2 Less than 0.

[0092] Compare and analyze the monitoring evaluation coefficient with the preset first evaluation threshold and the second evaluation threshold, and the first evaluation threshold is less than the second evaluation threshold;

[0093] If the monitoring evaluation coefficient is greater than the preset second evaluation threshold, the alarm level is the warning level;

[0094] If the monitoring evaluation coefficient is greater than the preset first evaluation threshold and less than the second evaluation threshold, the alarm level is a warning level;

[0095] If the monitoring evaluation coefficient is less than the preset first evaluation threshold, the alarm level is an emergency level.

[0096] If the alarm level is the early warning level, it means that the equipment has a potential failure risk, but it has not yet affected normal operation. At this time, the following measures should be taken:

[0097] Monitoring and tracking: Continuously monitor the operating data of relevant components to ensure timely detection of problems.

[0098] Regular inspections: Schedule regular inspections to check for possible sources of failures.

[0099] Predictive maintenance: Based on data analysis, predict when equipment may fail and prepare maintenance plans in advance.

[0100] If the alarm level is a warning level, it means that the equipment has experienced a serious abnormality and measures need to be taken as soon as possible. The maintenance decision suggestions at this time include:

[0101] Instant inspection and diagnosis: Detailed inspection and fault diagnosis of alarm components to locate the source of the problem.

[0102] Arrange maintenance: If the fault affects the efficiency of the equipment operation, arrange to shut down the equipment for maintenance.

[0103] Adjust operating parameters: If possible, delay the occurrence of failure and reduce risks by adjusting operating conditions (such as reducing load, adjusting wind speed adaptation, etc.).

[0104] If the alarm level is emergency, it means that the equipment has a serious fault, which may cause equipment damage or safety accidents. The maintenance decision suggestions at this time include:

[0105] Immediate shutdown: To avoid further damage to the equipment or safety accidents, the unit should be stopped immediately.

[0106] Emergency troubleshooting: Quickly organize a maintenance team to conduct on-site fault diagnosis and repair to ensure that the equipment returns to normal operation.

[0107] Safety assurance: If the failure involves safety risks, safety assurance measures should be taken immediately to ensure the safety of personnel.

[0108] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0109] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0110] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0111] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0112] 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.

[0113] 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.

Claims

1. A wind turbine monitoring and early warning system based on IoT sensing technology, characterized in that: include: Data acquisition module: collects the operating status information of key components of the wind turbine in real time, including vibration signals and temperature data, and uses low-power wireless communication technology to send the operating status data to the edge computing unit; Prediction module: Performs real-time analysis of edge computing devices based on neural network algorithms, identifies abnormal patterns, analyzes the time and frequency domain characteristics of vibration signals in combination with fast Fourier transform (FTT), detects mechanical failures, and obtains the remaining life prediction value based on the remaining life prediction (RUL) model; Preliminary warning module: through the analysis of remaining life prediction and abnormal mode conditions, the equipment health status score is determined based on fuzzy reasoning, and preliminary warning is made locally based on the equipment health status score; Grading warning module: Based on the mechanical failure situation and the equipment health status score, the wind turbine is remotely monitored and analyzed through the remote diagnosis system. According to the monitoring and analysis results, the wind turbine equipment is graded and alarmed, and maintenance decision suggestions and response measures are provided.

2. According to claim 1, a wind turbine monitoring and early warning system based on IoT sensing technology is characterized in that: The process of real-time collection of operating status information of key components of a wind turbine is as follows: Select an accelerometer suitable for vibration measurement of key components of wind turbines and install it at appropriate locations of key components to ensure that vibration signals can be accurately captured; Amplify and filter the analog signal output by the acceleration sensor to improve the signal quality and meet the input requirements of the analog-to-digital converter (ADC); Install the temperature sensor on the surface of key components to monitor temperature changes in real time; linearize and amplify the analog signal output by the temperature sensor and send it to the ADC for digital conversion; Use ADC to convert analog signals into digital signals for subsequent processing and transmission.

3. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 2 is characterized in that: The process of identifying abnormal patterns is as follows: After preprocessing the operation status information, a convolutional neural network is used to analyze it and capture the local features of the signal; The cross entropy loss function is used as the loss function of the convolutional neural network. The calculation formula of the cross entropy loss function is as follows: Where L is the loss function value, N is the number of samples, C is the number of categories, and y is ij is the true label of the i-th sample, p ij is the predicted probability that the i-th sample belongs to the j-th class; Update the parameters of the neural network through the SGD optimizer; Use the training set to train the neural network model and iteratively update the model parameters to minimize the value of the loss function; The preprocessed running status information is input into the trained neural network model, and the model outputs the probability that each sample belongs to the abnormal mode; Based on the comparison between the abnormal pattern probability and the set abnormal threshold, it is determined whether the sample is an abnormal pattern.

4. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 3 is characterized in that: The process of detecting mechanical failure is as follows: Combining the time domain and frequency domain features obtained by FFT analysis with the output results of the neural network model, it is possible to determine whether there are mechanical failures in the key components of the wind turbine. If the main frequency of the vibration signal changes abnormally and the abnormal probability output by the neural network model exceeds the threshold, it is considered that a mechanical fault exists.

5. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 4 is characterized in that: The process of obtaining the remaining life prediction value is as follows: Select the time domain and frequency domain features of the vibration signal as the input of the RUL model; Use historical operation data and corresponding remaining life labels to train the RUL model; After the training is completed, the real-time collected operating status information is input into the RUL model to output the predicted value of the remaining life of the equipment.

6. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 5 is characterized in that: The steps of determining the equipment health status score based on fuzzy reasoning are as follows: The abnormal mode probability and the remaining life prediction value are defined as input variables and divided into different fuzzy sets respectively; The equipment health status score is defined as the output variable and divided into fuzzy sets; Formulate fuzzy rules to describe the impact of abnormal mode probability and remaining life prediction value on equipment health status score; Fuzzy reasoning is performed based on fuzzy rules to determine the equipment health status score.

7. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 6 is characterized in that: The process of making a preliminary warning locally based on the equipment health status score is as follows: When the device health score is greater than the preset health threshold, it means that the device is in good health and no preliminary warning is required. When the device health score is less than the preset health threshold, it means that the device health is poor and a preliminary warning is required.

8. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 7 is characterized in that: The process of remote monitoring and analysis of wind turbines is as follows: According to the mechanical failure situation, the number of failures within a certain time period is obtained, and the number of failures is divided by the time period to obtain the equipment failure rate. Combined with the equipment health status score, the monitoring and evaluation coefficient is calculated based on the preset monitoring and evaluation formula. The specific calculation formula is as follows: P = u1*W+u2*Y; where P is the monitoring and evaluation coefficient, W is the equipment health status score, u1 is the equipment health status score weight factor, Y is the equipment failure rate, u2 is the equipment failure rate weight factor, u1 is greater than 0, and u2 is less than 0.

9. The wind turbine monitoring and early warning system based on IoT sensing technology according to claim 8 is characterized in that: The process of performing graded alarm on wind turbine equipment according to the monitoring and analysis results is as follows: Compare and analyze the monitoring evaluation coefficient with the preset first evaluation threshold and the second evaluation threshold, and the first evaluation threshold is less than the second evaluation threshold; If the monitoring evaluation coefficient is greater than the preset second evaluation threshold, the alarm level is the warning level; If the monitoring evaluation coefficient is greater than the preset first evaluation threshold and less than the second evaluation threshold, the alarm level is a warning level; If the monitoring evaluation coefficient is less than the preset first evaluation threshold, the alarm level is an emergency level.

10. A wind turbine monitoring and early warning system based on IoT sensing technology according to claim 9, characterized in that: The process of providing maintenance decision suggestions and response measures is as follows: If the alarm level is the early warning level, it means that there is a potential risk of failure in the equipment, but it has not yet affected normal operation. The operating data of related components should be continuously monitored to ensure that problems are discovered in time, and regular inspections should be arranged to check the possible sources of the failure; If the alarm level is a warning level, it means that the equipment has experienced a serious abnormality and measures need to be taken as soon as possible to conduct a detailed inspection and fault diagnosis of the alarm components, adjust the operating conditions to delay the occurrence of faults and reduce risks; If the alarm level is emergency, it means that the equipment has a serious fault. The unit should be stopped immediately and a maintenance team should be quickly organized to conduct on-site fault diagnosis and repair to ensure that the equipment resumes normal operation.

Images