CN117028169A - Dynamic adjustment method for wind turbine performance degradation and life extension - Google Patents
Dynamic adjustment method for wind turbine performance degradation and life extension Download PDFInfo
- Publication number
- CN117028169A CN117028169A CN202311073378.XA CN202311073378A CN117028169A CN 117028169 A CN117028169 A CN 117028169A CN 202311073378 A CN202311073378 A CN 202311073378A CN 117028169 A CN117028169 A CN 117028169A
- Authority
- CN
- China
- Prior art keywords
- wind turbine
- life
- data
- performance
- dynamic adjustment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
- F03D—WIND MOTORS
- F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/70—Wind energy
- Y02E10/72—Wind turbines with rotation axis in wind direction
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Wind Motors (AREA)
Abstract
本发明提供了风电机组性能退化与延寿的动态调整方法，提高风电机组的性能和寿命。其包括以下步骤：S101、在风电机组上安装传感器、摄像头和麦克风，以收集运行数据、视觉数据和声音数据；S102、将收集到的数据通过无线网络传输到云端服务器，并进行本地处理和存储；S103、利用数据分析和机器学习方法，建立风电机组的性能模型和寿命预测模型，对风电机组进行状态诊断和健康评估；S104、根据风电机组的状态诊断和健康评估结果，利用模糊控制和优化算法，自动调整风电机组的工作参数；S105、通过图形化界面和语音交互技术，向风电机组的运维人员展示风电机组的运行状态、性能曲线、寿命预测、故障报警信息，并提供智能化的维护建议和故障诊断方法。
The present invention provides a dynamic adjustment method for wind turbine performance degradation and life extension, thereby improving the performance and life of the wind turbine. It includes the following steps: S101. Install sensors, cameras and microphones on the wind turbine to collect operating data, visual data and sound data; S102. Transmit the collected data to the cloud server through the wireless network and process and store it locally. ; S103. Use data analysis and machine learning methods to establish the performance model and life prediction model of the wind turbine, and conduct status diagnosis and health assessment of the wind turbine; S104. Use fuzzy control and optimization based on the status diagnosis and health assessment results of the wind turbine. algorithm to automatically adjust the working parameters of the wind turbine; S105, through the graphical interface and voice interaction technology, display the operating status, performance curve, life prediction, and fault alarm information of the wind turbine to the operation and maintenance personnel of the wind turbine, and provide intelligent Maintenance recommendations and troubleshooting methods.
Description
Technical Field
The invention relates to the technical field of wind power generation, in particular to a dynamic adjustment method for performance degradation and life extension of a wind turbine generator.
Background
During operation of the wind turbine, the wind turbine may generate a large amount of operational data, visual data, and sound data. These data have an important impact on the performance and lifetime of the wind turbine. However, conventional data collection and transmission techniques may result in loss or delay of data, thereby affecting the accuracy and timeliness of data analysis and machine learning.
Furthermore, conventional wind turbines often lack efficient performance and life management mechanisms. This may lead to a gradual degradation of the performance of the wind turbine during operation, and a reduction of the lifetime due to variations in various factors, such as wind speed, wind direction, operating parameters, etc.
In addition, conventional wind turbines typically rely on manual status diagnostics and health assessment, which is not only inefficient, but also may present a risk of misand missed decisions. Meanwhile, conventional wind turbines generally lack self-learning and optimizing capabilities, which makes it impossible to timely and effectively adjust their own working strategies when facing complex and varying operating environments.
Therefore, how to improve the transmission speed and stability of the running data, visual data and sound data of the wind turbine, and avoid data loss or delay, so as to perform data analysis and machine learning in time, thereby improving the performance and service life of the wind turbine, is a great challenge faced by the current wind power technology.
Disclosure of Invention
Aiming at the problems, the invention provides a dynamic adjustment method for performance degradation and life prolongation of a wind turbine, which can improve the transmission speed and stability of running data, visual data and sound data of the wind turbine, avoid data loss or delay, and further improve the performance and life of the wind turbine.
The dynamic adjustment method for the performance degradation and the life prolongation of the wind turbine generator is characterized by comprising the following steps:
s101, installing a sensor, a camera and a microphone on a wind turbine to collect operation data, visual data and sound data of the wind turbine;
s102, transmitting the collected data to a cloud server through a wireless network, and carrying out local processing and storage through edge computing equipment;
s103, establishing a performance model and a life prediction model of the wind turbine by using a data analysis and machine learning method, and performing state diagnosis and health assessment on the wind turbine according to set threshold values and standards;
s104, according to the state diagnosis and health evaluation results of the wind turbine, working parameters of the wind turbine, such as blade angle, pitch speed and frequency converter frequency, are automatically adjusted by using a fuzzy control and optimization algorithm so as to achieve optimal performance output and service life extension; meanwhile, the wind turbine generator system also utilizes a reinforcement learning method to autonomously learn and optimize own working strategy through interaction with the environment;
s105, displaying the running state, the performance curve, the life prediction and the fault alarm information of the wind turbine to operation and maintenance personnel of the wind turbine by a graphical interface and a voice interaction technology, and providing an intelligent maintenance suggestion and a fault diagnosis method.
It is further characterized by:
in step S101, the sensors include a wind speed sensor, a wind direction sensor, a rotation speed sensor, a power sensor, a temperature sensor, and a vibration sensor;
in step S103, the data analysis and machine learning methods include neural networks, support vector machines, random forests, and the like;
in step S103, the state diagnosis includes determining whether the wind turbine generator is in a normal, abnormal or fault state, and a specific abnormal or fault type and cause;
in step S103, the health evaluation includes predicting a current performance level and a future performance trend of the wind turbine, and a current life remaining and a future life consumption, and giving a health index and a life index of the wind turbine, and giving a health grade and a life grade of the wind turbine according to a set threshold and a set standard;
in step S104, the fuzzy control includes converting the state diagnosis and health evaluation results of the wind turbine into a fuzzy set, and performing fuzzy reasoning on the working parameters of the wind turbine according to a set fuzzy rule base, and obtaining a specific control value by a defuzzification method;
in step S104, the optimization algorithm includes genetic algorithm, particle swarm optimization, ant colony algorithm, and under a given constraint condition, searching for an optimal or near optimal solution;
in step S104, the reinforcement learning includes selecting an action according to the current state and executing the action in each time step, then observing feedback of the environment, obtaining a reward according to the feedback, and updating the own strategy according to the reward;
in step S105, the graphical interface includes a web page, a mobile phone application, and a tablet computer device, and intuitively displays various data and information of the wind turbine in the form of a chart, a curve, and an instrument panel, prompts the state diagnosis and health evaluation results of the wind turbine in the form of colors, icons, and characters, and gives corresponding control adjustment and maintenance suggestions;
in step S105, the voice interaction technology includes voice recognition and voice synthesis technology, so as to implement natural language interaction between the wind turbine generator and the operation and maintenance personnel.
After the method is adopted, the transmission speed and stability of the running data, visual data and sound data of the wind turbine can be improved, and the data loss or delay is avoided, so that the performance and service life of the wind turbine are improved;
the quality and the accuracy of wind speed and wind direction data can be improved, so that the accuracy and the adaptability of a performance model and a life prediction model of the wind turbine generator are improved;
the accuracy and timeliness of the state diagnosis and health evaluation results of the wind turbine can be improved, so that the adjustment effect and efficiency of the working parameters of the wind turbine are improved;
the learning capacity of the wind turbine generator can be improved, so that the self-adaptability and the intelligent degree of the wind turbine generator are improved.
Drawings
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a system block diagram of a dynamic adjustment system of an embodiment of the present invention;
FIG. 3 is a schematic diagram of fuzzy control in an embodiment of the present invention;
FIG. 4 is a schematic diagram of an optimization algorithm in an embodiment of the invention;
FIG. 5 is a schematic diagram of reinforcement learning in an embodiment of the invention;
FIG. 6 is a graphical interface schematic in an embodiment of the invention;
FIG. 7 is a schematic illustration of voice interactions in an embodiment of the invention.
Detailed Description
The dynamic adjustment method for the performance degradation and the life prolongation of the wind turbine generator, see fig. 1, comprises the following steps:
s101: installing various sensors, cameras and microphones on the wind turbine to collect operation data, visual data and sound data of the wind turbine;
s102: transmitting the collected data to a cloud server through a wireless network, and carrying out local processing and storage through edge computing equipment;
s103: establishing a performance model and a life prediction model of the wind turbine by using a data analysis and machine learning method, and performing state diagnosis and health assessment on the wind turbine according to set threshold values and standards;
s104: according to the state diagnosis and health evaluation results of the wind turbine, working parameters of the wind turbine, such as blade angle, pitch speed, frequency converter frequency and the like, are automatically adjusted by utilizing a fuzzy control and optimization algorithm so as to achieve optimal performance output and service life extension; meanwhile, the wind turbine generator system also utilizes a reinforcement learning method to autonomously learn and optimize own working strategy through interaction with the environment;
s105: through a graphical interface and a voice interaction technology, information such as the running state, the performance curve, the life prediction, the fault alarm and the like of the wind turbine generator set is displayed to operation and maintenance personnel of the wind turbine generator set, and an intelligent maintenance suggestion and fault diagnosis method is provided.
The dynamic adjustment system for performance degradation and life prolongation of the wind turbine generator, as shown in fig. 2, comprises the following components:
wind turbine generator system 1: the invention provides a device for generating electric energy by utilizing a wind driven generator, which comprises a wind wheel, a variable pitch system, a frequency converter, a generator and other parts;
sensor 2: the data acquisition component comprises a wind speed sensor, a wind direction sensor, a rotating speed sensor, a power sensor, a temperature sensor, a vibration sensor and the like, and is used for measuring the operation data of the wind turbine generator 1;
camera 3: the data acquisition component is used for shooting images of the appearance and the surrounding environment of the wind turbine 1;
microphone 4: the data acquisition component is used for recording the running sound of the wind turbine generator 1 and the sound of the surrounding environment;
wireless network 5: the data transmission component is used for transmitting the collected data to the cloud server 6 through a wireless signal;
cloud server 6: the data processing component is positioned at the cloud end and is used for carrying out data analysis and machine learning on the collected data, establishing a performance model and a life prediction model of the wind turbine 1 and carrying out state diagnosis and health evaluation on the wind turbine 1;
edge computing device 7: the data processing component is arranged on the wind turbine generator system 1 and is used for locally processing and storing partial data;
fuzzy controller 8: the control and adjustment component is positioned on the cloud server 6 or the edge computing equipment 7 and is used for automatically adjusting the working parameters of the wind turbine generator 1 by utilizing a fuzzy control method according to the state diagnosis and health evaluation results of the wind turbine generator 1;
optimization algorithm 9: the control and adjustment component is positioned on the cloud server 6 or the edge computing equipment 7 and is used for automatically adjusting the working parameters of the wind turbine generator 1 by utilizing an optimization algorithm method according to the state diagnosis and health evaluation results of the wind turbine generator 1;
reinforcement learning 10: the control and adjustment component is positioned on the cloud server 6 or the edge computing equipment 7 and is used for enabling the wind turbine generator 1 to autonomously learn and optimize own working strategies through interaction with the environment;
graphical interface 11: the maintenance management component is positioned on equipment such as a webpage, a mobile phone application, a tablet personal computer and the like, and is used for displaying information such as the running state, the performance curve, the service life prediction, the fault alarm and the like of the wind turbine 1 to operation and maintenance personnel of the wind turbine 1 and providing an intelligent maintenance suggestion and fault diagnosis method;
voice interaction technology 12: the maintenance management component is positioned on the graphical interface 11 and is used for realizing natural language interaction between the wind turbine generator 1 and operation and maintenance personnel.
The steps and components of the present invention will be described in detail with reference to specific examples.
The first step: a plurality of sensors 2, cameras 3 and microphones 4 are arranged on the wind turbine 1 to collect operation data, visual data and sound data of the wind turbine 1; the specific contents are as follows:
the wind speed sensor is arranged in front of the wind wheel and is used for measuring the wind speed and the change trend of the position where the wind turbine generator 1 is located so as to adjust the angle of the blade and the pitch speed and improve the wind energy utilization rate. For example, when the wind speed sensor measures that the current wind speed is 5m/s, the data is sent to the cloud server 6 or the edge computing device 7;
the wind direction sensor is arranged at the top of the tower and is used for measuring the direction and the change trend of the wind direction of the position where the wind turbine generator 1 is positioned so as to adjust the yaw angle of the wind wheel and enable the wind wheel to face the optimal wind direction. For example, when the wind direction sensor measures that the current wind direction is northeast, the data is sent to the cloud server 6 or the edge computing device 7;
the rotating speed sensor is arranged on the generator and used for measuring the rotating speed and the change trend of the wind turbine generator system 1 so as to adjust the frequency of the frequency converter and keep the optimal rotating speed range. For example, when the rotation speed sensor measures that the current rotation speed is 1200rpm, the data is sent to the cloud server 6 or the edge computing device 7;
the power sensor 2d is installed on the output line and is used for measuring the output power and the change trend of the wind turbine generator 1 so as to evaluate the performance level and the service life condition of the wind turbine generator 1. For example, when the power sensor measures that the current output power is 200kW, the data is transmitted to the cloud server 6 or the edge computing device 7;
the temperature sensor is arranged on each component of the wind turbine generator 1, such as a blade, a bearing, a generator and the like, and is used for measuring the temperature and the change trend of each component of the wind turbine generator 1 so as to find out abnormal temperature rise or overheat phenomenon and timely perform heat dissipation or shutdown treatment. For example, when the temperature sensor measures that the temperature of the generator is 80 ℃, the data is sent to the cloud server 6 or the edge computing device 7;
the vibration sensor 2f is installed on each component of the wind turbine generator 1, such as a blade, a bearing, a generator, and the like, and is used for measuring the vibration amplitude and frequency of each component of the wind turbine generator 1 so as to find abnormal vibration or unbalance phenomenon and perform balancing or replacement processing in time. For example, when the vibration sensor measures that the vibration amplitude of the blade is 0.5mm, the data is transmitted to the cloud server 6 or the edge computing device 7;
the camera 3 is installed at the top of the tower or other proper positions and is used for shooting images of the appearance and the surrounding environment of the wind turbine generator 1 so as to observe the surface conditions of the components such as the blades, the tower, the foundation and the like, discover the problems such as cracks, corrosion, pollution and the like and timely carry out cleaning or repairing treatment. For example, when the camera 3 shoots that there is a significant crack in the blade, the image is sent to the cloud server 6 or the edge computing device 7;
the microphone 4 is installed at the top of the tower or other suitable positions, and is used for recording the running sound of the wind turbine generator system 1 and the sound of the surrounding environment, so as to analyze the sound characteristics, find problems such as abnormal sound, looseness, friction and the like, and timely perform fastening or lubricating treatment. For example, when microphone 4 records a sound of a metal friction of wind turbine 1, the sound is sent to cloud server 6 or edge computing device 7.
And a second step of: the collected data are transmitted to a cloud server 6 through a wireless network 5 and are locally processed and stored through an edge computing device 7; the specific contents are as follows:
the wireless network 5 transmits the collected data to the cloud server 6 through wireless signals by using wireless communication technology such as Wi-Fi, bluetooth, 4G, 5G and the like; for example, when the wireless network 5 detects that new data is generated, the data is packaged and encrypted, and sent to the cloud server 6 through the optimal channel and protocol;
the edge computing device 7 utilizes edge computing technology, such as raspberry pie, jetson Nano and the like, to locally process and store part of data, so that the safety and instantaneity of the data are improved, and the network delay and bandwidth consumption are reduced; for example, when the edge computing device 7 receives new data, the data is compressed, encrypted, buffered, etc., and according to the set rule, it is determined whether to send the data to the cloud server 6 or to the local storage.
And a third step of: establishing a performance model and a life prediction model of the wind turbine 1 by using a data analysis and machine learning method, and performing state diagnosis and health assessment on the wind turbine 1 according to set thresholds and standards; the specific contents are as follows:
data analysis
Preprocessing, cleaning, normalizing, dimension reducing and other operations are carried out on the collected data by utilizing a data analysis technology such as Python, R, matlab, and information such as characteristic values, statistical values, correlation and the like of the data is extracted to provide a foundation for the establishment of a subsequent model; for example, when the cloud server 6 or the edge computing device 7 receives new data, the data is subjected to operations such as denoising, filling up missing values, standardization, principal component analysis and the like, and information such as mean, variance and covariance of the data is calculated;
machine learning
By means of machine learning technology, such as neural network, support vector machine, random forest and other methods, according to historical data and label data, a model capable of accurately predicting the performance output and life remaining of the wind turbine 1 and a model capable of identifying the running state and health condition of the wind turbine 1 are trained. For example, when the cloud server 6 or the edge computing device 7 acquires enough training data, the data is input into a multi-layer perceptron neural network, and training is performed through a back propagation algorithm, so as to obtain a model capable of outputting corresponding performance values and life values according to the input operation data; meanwhile, the data is input into a support vector machine model, and training is carried out through a kernel function and a soft interval method, so that a model which can output corresponding state labels and health labels according to the input operation data is obtained;
state diagnostics
And analyzing and judging real-time data by using the trained model, judging whether the wind turbine generator system 1 is in a normal, abnormal or fault state, and specific abnormal or fault type and cause, and sending out alarm signals and prompt information in time. For example, when the cloud server 6 or the edge computing device 7 receives new real-time data, the data is input into a support vector machine model, and corresponding state labels and health labels are obtained; if the state label is normal, the wind turbine generator 1 is indicated to run normally; if the state label is abnormal, indicating that the wind turbine generator 1 has a certain abnormal condition; if the state label is a fault, the state label indicates that a certain fault condition occurs in the wind turbine generator 1; meanwhile, specific abnormality or fault type and cause can be judged according to the health label; for example, if the health label is damaged, it indicates that the blade of the wind turbine generator 1 has a problem such as crack or fracture; if the health label is bearing abrasion, the abrasion or clamping stagnation and other problems of the bearing of the wind turbine generator 1 are indicated; if the health label is that the generator is overheated, the generator of the wind turbine generator system 1 has the problems of overheating or poor heat dissipation and the like; according to different conditions, corresponding alarm signals and prompt information such as red light, beeping sound, voice prompt and the like are sent out so as to remind operation and maintenance personnel to check and process in time;
health assessment
And analyzing and predicting real-time data by using the trained model, predicting the current performance level and future performance trend, the current life remaining and future life consumption of the wind turbine generator 1, giving the health index and the life index of the wind turbine generator 1, and giving the health grade and the life grade of the wind turbine generator 1 according to the set threshold and standard. For example, when the cloud server 6 or the edge computing device 7 receives new real-time data, the data is input into the multi-layer perceptron neural network, and corresponding performance values and life values are obtained; then, calculating the health index and the life index of the wind turbine generator 1 according to the performance value and the life value; for example, if the performance value is 200kW and the lifetime value is 10 years, the health index is 200/250=0.8 and the lifetime index is 10/20=0.5; then, according to the set threshold value and standard, giving the health grade and life grade of the wind turbine 1; for example, if the health index is greater than 0.8, the health grade is excellent; if the health index is between 0.6 and 0.8, the health rating is good; health grade is general if the health index is between 0.4 and 0.6; if the health index is less than 0.4, the health grade is poor; similarly, if the life index is greater than 0.8, the life class is long; if the life index is between 0.6 and 0.8, the life class is medium; if the life index is between 0.4 and 0.6, the life class is short; if the life index is less than 0.4, the life class is extremely short.
Fourth step: according to the state diagnosis and health evaluation results of the wind turbine generator system 1, working parameters of the wind turbine generator system 1, such as blade angle, pitch speed, frequency converter frequency and the like, are automatically adjusted by utilizing a fuzzy control and optimization algorithm so as to achieve optimal performance output and service life extension; meanwhile, the wind turbine generator system 1 also utilizes a reinforcement learning method to autonomously learn and optimize own working strategy through interaction with the environment; the specific contents are as follows:
fuzzy control
The state diagnosis and health evaluation results of the wind turbine generator 1 are converted into fuzzy sets by using a fuzzy control technology, such as Matlab Fuzzy Logic Toolbox, and the working parameters of the wind turbine generator 1 are subjected to fuzzy reasoning according to a set fuzzy rule base, and a specific control magnitude is obtained by a defuzzification method. For example, as shown in fig. 3, when the condition diagnosis result is abnormal and the abnormal type is blade damage, the result is converted into a fuzzy set, for example, the degree of abnormality is 0.7 and the degree of damage is 0.5; then, according to a set fuzzy rule base, if the degree of abnormality is high and the degree of damage is medium, the blade angle is large, the pitch speed is high, the frequency of a frequency converter is low, and the like, fuzzy reasoning is carried out on the working parameters of the wind turbine generator set 1, so as to obtain a fuzzy control value, if the blade angle is 0.8, the pitch speed is 0.9, and the frequency of the frequency converter is 0.2; then, specific control values are obtained through a defuzzification method, such as a gravity center method, an average method and the like, wherein the blade angle is 25 degrees, the pitch speed is 15rpm, and the frequency of a frequency converter is 40Hz; then, the control values are sent to the wind turbine generator 1, and corresponding control adjustment is executed;
optimization algorithm
And (3) searching an optimal or near-optimal solution under a given constraint condition by utilizing an optimization algorithm technology, such as a genetic algorithm, a particle swarm algorithm, an ant swarm algorithm and the like. For example, as shown in fig. 4, when the health evaluation result is general and the life class is short, the result is taken as an objective function such as max (health index+life index); then, according to given constraint conditions, such as safe operation of the wind turbine generator 1, meeting of load requirements and the like, working parameters of the wind turbine generator 1 are used as decision variables, such as blade angle, pitch speed, frequency converter frequency and the like; then, by utilizing a genetic algorithm, under a certain iteration number and population scale, searching a solution for maximizing or minimizing an objective function through operations such as selection, crossing, variation and the like, wherein the blade angle is 20 degrees, the pitch speed is 10rpm, and the frequency of a frequency converter is 50Hz; these solutions are then sent to the wind turbine 1 and corresponding control adjustments are performed;
reinforcement learning
By means of reinforcement learning technology, such as Q-learning, SARSA, actor-Critic and other methods, the wind turbine generator system 1 can autonomously learn and optimize own working strategies through interaction with the environment. For example, as shown in fig. 5, in each time step, an action is selected according to the current state and performed, and then feedback of the environment is observed, and a reward is obtained according to the feedback, and its policy is updated according to the reward. For example, at time t, according to the current state s t (e.g., wind speed, rotational speed, power, etc.), an action a is selected t (e.g., adjusting blade angle, pitch speed, frequency converter frequency, etc.), and performing the action; feedback s of the observed environment t +1 (e.g. wind speed, rotational speed, power, etc.), and obtaining a prize r based on feedback t +1 (e.g., performance improvement or life extension, etc.), and updates its own policy Q(s) t ,a t ) (Q(s) as in the Q-learning method t ,a t )＝Q(s t ,a t )+α[r t +1+γmax a Q(s t +1,a)-Q(s t ,a t )]Wherein Q(s) t ,a t ) Is the current estimated state s t Lower selection action a t Is of value (c). Alpha is the learning rate, which determines the rate at which new information will overwrite old information. r is (r) t +1 is in state s t Lower selection action a t The immediate rewards obtained. Gamma is a discounting factor that determines the importance of future rewards. max (max) a Q(s t +1, a) is in the next state s t The maximum expected return for all possible actions at +1); the above process is then repeated at time t+1. By using the reinforcement learning method, the wind turbine generator system 1 can adapt to different wind power and load conditions, and self-optimization and self-learning are realized.
Fifth step: through a graphical interface 11 and a voice interaction technology 12, information such as the running state, the performance curve, the life prediction, the fault alarm and the like of the wind turbine 1 is displayed to operation and maintenance personnel of the wind turbine 1, and an intelligent maintenance suggestion and fault diagnosis method is provided; the specific contents are as follows:
the graphical interface 11 utilizes graphical interface technology, such as HTML, CSS, javaScript, to display various data and information of the wind turbine generator 1, such as wind speed, rotation speed, power, temperature, vibration, vision, sound, etc., and intuitively displays the performance output and life remaining of the wind turbine generator 1 in the form of charts, curves, instrument panels, etc., prompts the state diagnosis and health evaluation results of the wind turbine generator 1 in the form of colors, icons, characters, etc., and gives corresponding control adjustment and maintenance suggestions. For example, as shown in fig. 6, an operator can view various data and information of the wind turbine generator 1 through a web page, a mobile phone application, a tablet computer and other devices, for example, the current wind speed is 5m/s, the rotation speed is 1200rpm, the power is 200kW, the temperature is 80 ℃, the vibration amplitude is 0.5mm, a crack is formed on a visual image display blade, and a sound characteristic is displayed with a metal friction sound and the like; and the change trend of the performance output and the life remaining of the wind turbine generator system 1 is displayed by a graph, such as the performance curve is in a descending trend, the life curve is in a descending trend and the like; the red light and the characters are used for prompting the state diagnosis and health evaluation results of the wind turbine generator system 1, if the state is abnormal, the abnormal type is blade damage, the health grade is general, the service life grade is short, and the like; and corresponding control adjustment and maintenance suggestions are given, such as increasing the blade angle and the pitch speed, reducing the frequency of the frequency converter, repairing the blade cracks in time and the like.
The voice interaction technology 12 implements natural language interaction between the wind turbine generator 1 and the operation and maintenance personnel by using voice interaction technologies, such as voice recognition and voice synthesis technologies. For example, as shown in fig. 7, the operator may query various data and information of the wind turbine 1, such as "what is the current wind speed? "," what is the blade angle? "," there is an abnormality? "etc. and listens to the answers of the wind turbines 1 by voice. Meanwhile, the wind turbine generator system 1 can also report the running condition of the wind turbine generator system to operation and maintenance personnel through voice, such as 'low wind speed', increased blade angle ', high rotating speed, reduced frequency of the frequency converter', 'found blade crack, timely repair', and the like, and receive instructions or feedback of the operation and maintenance personnel through voice.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present disclosure describes embodiments, not every embodiment is provided with a separate embodiment, and that this description is provided for clarity only, and that the disclosure is not limited to the embodiments described in detail below, and that the embodiments described in the examples may be combined as appropriate to form other embodiments that will be apparent to those skilled in the art.
Claims (10)
1. The dynamic adjustment method for the performance degradation and the life prolongation of the wind turbine generator is characterized by comprising the following steps:
s101, installing a sensor, a camera and a microphone on a wind turbine to collect operation data, visual data and sound data of the wind turbine;
s102, transmitting the collected data to a cloud server through a wireless network, and carrying out local processing and storage through edge computing equipment;
s103, establishing a performance model and a life prediction model of the wind turbine by using a data analysis and machine learning method, and performing state diagnosis and health assessment on the wind turbine according to set threshold values and standards;
s104, according to the state diagnosis and health evaluation results of the wind turbine, working parameters of the wind turbine, such as blade angle, pitch speed and frequency converter frequency, are automatically adjusted by using a fuzzy control and optimization algorithm so as to achieve optimal performance output and service life extension; meanwhile, the wind turbine generator system also utilizes a reinforcement learning method to autonomously learn and optimize own working strategy through interaction with the environment;
s105, displaying the running state, the performance curve, the life prediction and the fault alarm information of the wind turbine to operation and maintenance personnel of the wind turbine by a graphical interface and a voice interaction technology, and providing an intelligent maintenance suggestion and a fault diagnosis method.
2. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S101, the sensors include a wind speed sensor, a wind direction sensor, a rotation speed sensor, a power sensor, a temperature sensor, and a vibration sensor.
3. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S103, the data analysis and machine learning methods include neural networks, support vector machines, random forests, and the like.
4. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S103, the state diagnosis includes determining whether the wind turbine generator is in a normal, abnormal or fault state, and a specific abnormal or fault type and cause.
5. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S103, the health evaluation includes predicting the current performance level and the future performance trend, and the current life remaining and the future life consumption of the wind turbine, and giving the health index and the life index of the wind turbine, and giving the health grade and the life grade of the wind turbine according to the set threshold and standard.
6. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S104, the fuzzy control includes converting the state diagnosis and health evaluation results of the wind turbine into a fuzzy set, and performing fuzzy reasoning on the working parameters of the wind turbine according to a set fuzzy rule base, and obtaining a specific control value by a defuzzification method.
7. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S104, the optimization algorithm includes genetic algorithm, particle swarm algorithm, ant colony algorithm, and under a given constraint condition, an optimal or near optimal solution is found.
8. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S104, the reinforcement learning includes selecting an action according to the current state and executing the action, and then observing feedback of the environment, obtaining a reward according to the feedback, and updating its own policy according to the reward, in each time step.
9. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S105, the graphical interface includes a web page, a mobile phone application, and a tablet computer device, and intuitively displays various data and information of the wind turbine in the form of a chart, a curve, and an instrument panel, prompts the state diagnosis and health evaluation results of the wind turbine in the form of colors, icons, and characters, and gives corresponding control adjustment and maintenance suggestions.
10. The dynamic adjustment method for performance degradation and life extension of a wind turbine according to claim 1, wherein the method comprises the following steps: in step S105, the voice interaction technology includes voice recognition and voice synthesis technology, so as to implement natural language interaction between the wind turbine generator and the operation and maintenance personnel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311073378.XA CN117028169A (en) | 2023-08-24 | 2023-08-24 | Dynamic adjustment method for wind turbine performance degradation and life extension |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311073378.XA CN117028169A (en) | 2023-08-24 | 2023-08-24 | Dynamic adjustment method for wind turbine performance degradation and life extension |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117028169A true CN117028169A (en) | 2023-11-10 |
Family
ID=88622603
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311073378.XA Pending CN117028169A (en) | 2023-08-24 | 2023-08-24 | Dynamic adjustment method for wind turbine performance degradation and life extension |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117028169A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118327909A (en) * | 2024-04-12 | 2024-07-12 | 明阳智慧能源集团股份公司 | A method for vibration monitoring and fault diagnosis of wind turbine generator set |
| CN118982348A (en) * | 2024-10-22 | 2024-11-19 | 南通市万帝来机电有限公司 | A method and system for intelligent operation and maintenance analysis of ventilators based on big data |
| CN119027101A (en) * | 2024-10-23 | 2024-11-26 | 山东华方智联科技股份有限公司 | Directed maintenance method and platform for electromechanical equipment combined with dynamic prediction of component life |
| CN119146017A (en) * | 2024-11-14 | 2024-12-17 | 浙江华科同安监控技术有限公司 | Wind generating set fault diagnosis system |
| CN119476511A (en) * | 2024-12-26 | 2025-02-18 | 广州施杰节能科技有限公司 | An intelligent algorithm optimization combination and control method for parallel water pumps |
| CN119493402A (en) * | 2024-10-29 | 2025-02-21 | 浙江百灵智能科技有限公司 | A loom data monitoring method, system, computer equipment and storage medium |
| CN120449723A (en) * | 2025-07-11 | 2025-08-08 | 兰州理工大学 | A high-fidelity simulation platform, storage medium, and electronic device for wind turbine degradation-temperature coupling operation status |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102108936A (en) * | 2009-12-25 | 2011-06-29 | 通用电气公司 | System and method for monitoring and controlling a wind park |
| US20130214534A1 (en) * | 2012-02-21 | 2013-08-22 | Mitsubishi Heavy Industries, Ltd. | Method for operating wind farm and operation control system for wind farm |
| CN105179164A (en) * | 2015-06-25 | 2015-12-23 | 江苏科技大学 | Wind energy converting system sliding mode control method and device based on T-S fuzzy model |
| CN111322206A (en) * | 2020-02-28 | 2020-06-23 | 唐智科技湖南发展有限公司 | Intelligent operation and maintenance system and method for large mechanical part of wind turbine generator |
| CN112483334A (en) * | 2020-12-11 | 2021-03-12 | 重庆科凯前卫风电设备有限责任公司 | Intelligent control method of wind turbine generator based on edge calculation |
| CN112727702A (en) * | 2020-12-11 | 2021-04-30 | 中国大唐集团科学技术研究院有限公司火力发电技术研究院 | Health management and fault early warning method for wind turbine generator |
| CN116146436A (en) * | 2023-02-21 | 2023-05-23 | 南昌华梦达航空科技发展有限公司 | A multi-mode monitoring system and method for fan operation |
-
2023
- 2023-08-24 CN CN202311073378.XA patent/CN117028169A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102108936A (en) * | 2009-12-25 | 2011-06-29 | 通用电气公司 | System and method for monitoring and controlling a wind park |
| US20130214534A1 (en) * | 2012-02-21 | 2013-08-22 | Mitsubishi Heavy Industries, Ltd. | Method for operating wind farm and operation control system for wind farm |
| CN105179164A (en) * | 2015-06-25 | 2015-12-23 | 江苏科技大学 | Wind energy converting system sliding mode control method and device based on T-S fuzzy model |
| CN111322206A (en) * | 2020-02-28 | 2020-06-23 | 唐智科技湖南发展有限公司 | Intelligent operation and maintenance system and method for large mechanical part of wind turbine generator |
| CN112483334A (en) * | 2020-12-11 | 2021-03-12 | 重庆科凯前卫风电设备有限责任公司 | Intelligent control method of wind turbine generator based on edge calculation |
| CN112727702A (en) * | 2020-12-11 | 2021-04-30 | 中国大唐集团科学技术研究院有限公司火力发电技术研究院 | Health management and fault early warning method for wind turbine generator |
| CN116146436A (en) * | 2023-02-21 | 2023-05-23 | 南昌华梦达航空科技发展有限公司 | A multi-mode monitoring system and method for fan operation |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118327909A (en) * | 2024-04-12 | 2024-07-12 | 明阳智慧能源集团股份公司 | A method for vibration monitoring and fault diagnosis of wind turbine generator set |
| CN118982348A (en) * | 2024-10-22 | 2024-11-19 | 南通市万帝来机电有限公司 | A method and system for intelligent operation and maintenance analysis of ventilators based on big data |
| CN119027101A (en) * | 2024-10-23 | 2024-11-26 | 山东华方智联科技股份有限公司 | Directed maintenance method and platform for electromechanical equipment combined with dynamic prediction of component life |
| CN119027101B (en) * | 2024-10-23 | 2025-01-17 | 山东华方智联科技股份有限公司 | Electromechanical equipment directional maintenance method and platform combined with dynamic prediction of service life of parts |
| CN119493402A (en) * | 2024-10-29 | 2025-02-21 | 浙江百灵智能科技有限公司 | A loom data monitoring method, system, computer equipment and storage medium |
| CN119146017A (en) * | 2024-11-14 | 2024-12-17 | 浙江华科同安监控技术有限公司 | Wind generating set fault diagnosis system |
| CN119476511A (en) * | 2024-12-26 | 2025-02-18 | 广州施杰节能科技有限公司 | An intelligent algorithm optimization combination and control method for parallel water pumps |
| CN120449723A (en) * | 2025-07-11 | 2025-08-08 | 兰州理工大学 | A high-fidelity simulation platform, storage medium, and electronic device for wind turbine degradation-temperature coupling operation status |
| CN120449723B (en) * | 2025-07-11 | 2025-09-23 | 兰州理工大学 | Running state high-fidelity simulation platform for degradation-temperature coupling of wind turbine generator, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117028169A (en) | Dynamic adjustment method for wind turbine performance degradation and life extension | |
| CN116937575A (en) | An energy monitoring and management system for grid systems | |
| CN118915566A (en) | Heating ventilation equipment abnormity on-line monitoring system based on Internet of things | |
| CN118462501A (en) | A preventive maintenance and fault diagnosis system and method for offshore wind turbines | |
| CN117028147B (en) | Wind power variable pitch control system and wind power variable pitch system | |
| CN103758696A (en) | SCADA (supervisory control and data acquisition) temperature parameter based wind turbine set security evaluation method | |
| CN112253404A (en) | Intelligent operation and maintenance system of fan drive chain based on multi-level health assessment | |
| CN117390944A (en) | Substation operation condition simulation system | |
| CN103321839A (en) | Fan vibration monitoring method and system as well as fan monitor | |
| CN115456041A (en) | Equipment failure early warning method and device, computing equipment and storage medium | |
| JP2021161932A (en) | Power generation amount prediction device for wind power generator | |
| CN117521961A (en) | Power distribution Internet of things monitoring system based on deep learning | |
| CN115249972A (en) | A system and method for evaluating the performance of wind turbines in wind farms under big data | |
| CN114117887A (en) | Real-time evaluation method, system and medium for online power generation performance of wind turbines | |
| CN108959498A (en) | A kind of big data processing platform and its design method for health monitoring | |
| CN118167563A (en) | SCADA-based fan control parameter abnormality detection method and system | |
| CN117390396A (en) | A method for establishing a power communication model based on digital twins | |
| CN114218690A (en) | Blade breakage early warning method and device for wind turbine generator | |
| CN118242304A (en) | Variable frequency control method and system for wind power installation ship cabin fan | |
| CN120222959A (en) | Intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things | |
| CN119579140B (en) | A wind turbine nacelle equipment quality assessment and maintenance decision-making method based on reinforcement learning | |
| CN120277580A (en) | Wind turbine generator early warning method and device based on IPC hardware platform | |
| CN118214153A (en) | A wind farm intelligent operation and maintenance management system | |
| CN119047683A (en) | Intelligent industrial park energy consumption management system and method based on big data | |
| CN118820926A (en) | Frequent alarm monitoring method and system for equipment operation status based on hydropower units |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |