A wind turbine blade icing blur early warning method and system
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- LINYI UNIVERSITY
- Filing Date
- 2026-04-29
- Publication Date
- 2026-06-05
Smart Images
Figure CN122148512A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wind turbine technology, and in particular to a method and system for early warning of icing on wind turbine blades. Background Technology
[0002] Wind turbines are core equipment for clean energy power generation, and their blades are prone to icing in low-temperature and high-humidity environments. As wind turbines expand into colder regions, the requirements for their operational safety are further increasing. Icing on the blade surface can lead to decreased aerodynamic performance and load imbalance, potentially causing increased vibration or shutdown during operation. This not only affects power generation efficiency but also shortens equipment lifespan and may even cause mechanical failures.
[0003] Existing icing detection technologies mostly rely on temperature threshold judgment or visual recognition, which suffers from problems such as response lag and high false alarm rates. Furthermore, because the risk of icing changes dynamically under different environmental conditions, traditional methods are difficult to provide timely and accurate early warnings. Therefore, it is necessary to develop an intelligent early warning system that can assess the risk of icing in real time through multi-parameter fusion and implement tiered prevention and control measures to ensure the safe operation of wind turbines. Summary of the Invention
[0004] The technical problem to be solved by the present invention is to provide a fuzzy early warning method and system for wind turbine blade icing, which solves the problems of slow response and high false alarm rate in the existing wind turbine blade icing detection technology, and ensures the safe operation of wind turbines.
[0005] This invention is achieved through the following technical solution: A method for fuzzy early warning of icing on wind turbine blades, comprising the following steps: S1: Identify the key factors affecting the risk of icing on wind turbine blades and monitor these key factors in real time; S2: Establish a blade icing risk assessment model based on fuzzy logic according to the key factors affecting the icing risk of wind turbine blades; S3: Establish a fuzzy inference rule table for the blade icing risk assessment model based on the historical operating status data of wind turbines, input the key factor data of real-time monitoring into the blade icing risk assessment model, and obtain the abstract risk level based on the fuzzy inference rule table; S4: The abstract risk level is transformed into the membership degree of the output result through the blade icing risk assessment model. Then, the abstract risk level is defuzzified by the weighted average method based on the membership degree of the output result to obtain the accurate blade icing risk value and the corresponding accurate blade icing risk level. S5: Issues early warnings based on the blade icing risk level, enabling multi-level icing early warning and protection for wind turbine blades.
[0006] The optimized key factors affecting the risk of icing on wind turbine blades, as described in step S1, include ambient temperature, blade vibration frequency, and ambient humidity.
[0007] In the optimized version, step S1 involves real-time monitoring of ambient temperature using a temperature sensor, real-time monitoring of blade vibration frequency using an accelerometer, and real-time monitoring of ambient humidity using a humidity sensor.
[0008] Furthermore, the method for establishing a fuzzy logic-based blade icing risk assessment model in step S2 based on the key factors affecting the icing risk of wind turbine blades is as follows: S211: Using real-time monitoring data of key factors as the precise input variable of the leaf icing risk assessment model, and the estimated degree of leaf icing risk as the precise output variable of the leaf icing risk assessment model, the basic domain of the precise input and output variables of the leaf icing risk assessment model is established. S212: Discretize the basic universe of discourse of the precise input variables and precise output variables of each blade icing risk assessment model, and perform targeted fuzzification processing on the precise input variables and precise output variables of each blade icing risk assessment model to obtain the fuzzy set of fuzzy output variables and the fuzzy set of input variables of the blade icing risk assessment model. S213: Calculate the corresponding quantification factor and proportional factor based on the basic domain of the precise input variables and precise output variables of the leaf icing risk assessment model; S214: Based on the fuzzy set of input variables and the corresponding quantization factor and scaling factor of the blade icing risk assessment model, determine the membership function between the corresponding input variables and the fuzzy set, and establish a blade icing risk assessment model based on fuzzy logic.
[0009] In the optimized step S212, the fuzzy set of the fuzzy output variable of the blade icing risk assessment model is {L, M, H}, where L represents low risk, M represents medium risk, and H represents high risk, and the value of the fuzzy set of the fuzzy output variable is [0, 2].
[0010] Furthermore, in step S213, the quantization factor is calculated according to equation (1), and the scaling factor is calculated according to equation (2): (1); (2); in: Indicates the quantification factor. Indicates the scaling factor. This represents the maximum absolute value of the discretized universe of discourse. Denotes the maximum value of the fundamental universe of discourse. It represents the minimum value of the fundamental universe of discourse.
[0011] Furthermore, the membership function of the low-temperature environment temperature and the fuzzy set determined in step S214 is Equation (3), the membership function of the medium-temperature environment temperature and the fuzzy set is Equation (4), the membership function of the high-temperature environment temperature and the fuzzy set is Equation (5), the membership function of the normal leaf vibration frequency and the fuzzy set is Equation (6), the membership function of the slightly abnormal leaf vibration frequency and the fuzzy set is Equation (7), the membership function of the severely abnormal leaf vibration frequency and the fuzzy set is Equation (8), the membership function of the low-humidity environment humidity and the fuzzy set is Equation (9), the membership function of the medium-humidity environment humidity and the fuzzy set is Equation (10), and the membership function of the high-humidity environment humidity and the fuzzy set is Equation (11). (3); (4); (5); (6); (7); (8); (9); (10); (11); in: The membership function representing the temperature of a low-temperature environment and the fuzzy set. The membership function representing the temperature of a medium-temperature environment and the fuzzy set. The membership function representing the temperature of a high-temperature environment and the fuzzy set. Indicates ambient temperature. The membership function representing the normal blade vibration frequency and the fuzzy set. The membership function of the fuzzy set represents the frequency of slight abnormal blade vibration. The membership function of the fuzzy set represents the frequency of severely abnormal blade vibration. Indicates the blade vibration frequency. The membership function representing the humidity of a low-humidity environment and the fuzzy set. The membership function of a fuzzy set represents the humidity in a moderately humid environment. The membership function representing the humidity of a fuzzy set in a high-humidity environment. Indicates ambient humidity.
[0012] Furthermore, in step S4, the weighted average method is used to defuzzify the abstract risk level output value according to equation (12): (12); in: This indicates the estimated risk level of leaf icing. The first one in the fuzzy inference rule table represents the... The membership degree of the rule output results. Indicates the weighting coefficient. Indicates the total number of rules.
[0013] A fuzzy early warning system for wind turbine blade icing, used to execute a fuzzy early warning method for wind turbine blade icing as described in any of the above, comprising a key factor determination and acquisition module, a blade icing risk assessment model construction module, an abstract risk level acquisition module, a precise blade icing risk level acquisition module, and an early warning module. The key factor identification and acquisition module is used to identify the key factors affecting the risk of icing on wind turbine blades and to monitor these key factors in real time. The blade icing risk assessment model construction module is used to establish a blade icing risk assessment model based on fuzzy logic according to the key factors affecting the icing risk of wind turbine blades. The abstract risk level acquisition module is used to establish a fuzzy inference rule table for the blade icing risk assessment model based on the historical operating status data of the wind turbine, input the real-time monitored key factor data into the blade icing risk assessment model, and obtain the abstract risk level based on the fuzzy inference rule table. The precise blade icing risk level acquisition module is used to convert the abstract risk level into the membership degree of the output result through the blade icing risk assessment model. Then, based on the membership degree of the output result, the weighted average method is used to defuzzify the abstract risk level to obtain the precise blade icing risk value and the corresponding precise blade icing risk level. The early warning module is used to issue early warnings based on the blade icing risk level, thereby achieving multi-level icing early warning and protection for wind turbine blades.
[0014] Beneficial effects of the invention: This invention provides a fuzzy early warning method and system for wind turbine blade icing. It uses ambient temperature, humidity, and blade vibration frequency as comprehensive reference parameters, avoiding the lag issues associated with relying solely on temperature sensors or visual detection. By establishing an icing risk assessment model based on fuzzy logic, it performs fuzzy inference and defuzzification to quickly obtain icing risk information, which is then transmitted to the early warning controller for alerting. This invention only requires the integration of conventional sensors and controllers into the wind turbine, eliminating the need for additional high-cost equipment. It offers advantages such as fast response and low cost, solving the problems of lag and high false alarm rates in existing wind turbine blade icing detection technologies, thus ensuring the safe operation of wind turbines. Attached Figure Description
[0015] Figure 1 This is a schematic diagram of the process of this invention.
[0016] Figure 2 This is a schematic diagram of the relationship between ambient temperature and the membership function of a fuzzy set in this invention.
[0017] Figure 3 This is a schematic diagram of the blade vibration frequency and the membership function of the fuzzy set in this invention.
[0018] Figure 4 This is a schematic diagram of the relationship between environmental humidity and the membership function of fuzzy sets in this invention. Detailed Implementation
[0019] A fuzzy early warning method for icing on wind turbine blades, the flowchart of which is shown below. Figure 1 As shown, it includes the following steps: S1: Identify the key factors affecting the risk of icing on wind turbine blades and monitor these key factors in real time; The optimized key factors affecting the risk of icing on wind turbine blades include ambient temperature, blade vibration frequency, and ambient humidity.
[0020] Among them, the ambient temperature can be monitored in real time by a temperature sensor, the blade vibration frequency can be monitored in real time by an acceleration sensor, and the ambient humidity can be monitored in real time by a humidity sensor. The temperature sensor can be installed in the nacelle of the wind turbine, the acceleration sensor can be installed at the root of the wind turbine blade, preferably a triaxial acceleration sensor, and the humidity sensor can be installed in the middle of the wind turbine tower.
[0021] S2: Establish a blade icing risk assessment model based on fuzzy logic according to the key factors affecting the icing risk of wind turbine blades; Specifically, the method for establishing a fuzzy logic-based blade icing risk assessment model based on the key factors affecting the icing risk of wind turbine blades is as follows: S211: Using real-time monitoring data of key factors as the precise input variable of the leaf icing risk assessment model, and the estimated degree of leaf icing risk as the precise output variable of the leaf icing risk assessment model, the basic domain of the precise input and output variables of the leaf icing risk assessment model is established. Specifically, the precise input variables for the blade icing risk assessment model include ambient temperature, blade vibration frequency, and ambient humidity.
[0022] Preferably, the basic domain of the precise input variables for the blade icing risk assessment model is: ambient temperature T∈[-20,10], blade vibration frequency F∈[0.5,3], ambient humidity S∈[0,100], and blade icing risk level R∈[0,1].
[0023] S212: Discretize the basic universe of discourse of the precise input variables and precise output variables of each blade icing risk assessment model, and perform targeted fuzzification processing on the precise input variables and precise output variables of each blade icing risk assessment model to obtain the fuzzy set of fuzzy output variables and the fuzzy set of input variables of the blade icing risk assessment model. Preferably, the fuzzy set of the fuzzy output variables of the blade icing risk assessment model is {L, M, H}, where L represents low risk, M represents medium risk, and H represents high risk, and the value of the fuzzy set of the fuzzy output variables is [0,2].
[0024] S213: Calculate the corresponding quantification factor and proportional factor based on the basic domain of the precise input variables and precise output variables of the leaf icing risk assessment model; Specifically, the quantification factor can be calculated according to equation (1), and the scaling factor can be calculated according to equation (2): (1); (2); in: Indicates the quantification factor. Indicates the scaling factor. This represents the maximum absolute value of the discretized universe of discourse. Denotes the maximum value of the fundamental universe of discourse. It represents the minimum value of the fundamental universe of discourse.
[0025] The function of quantization factor and proportional factor is to transform the ambient temperature, blade vibration frequency, ambient humidity, and blade icing risk level according to a certain proportion, so as to meet the fuzzification and defuzzification requirements of fuzzy control and improve the accuracy of icing risk assessment.
[0026] Assuming ambient temperature T ∈ [-20, 10], blade vibration frequency F ∈ [0.5, 3], ambient humidity S ∈ [0, 100], and blade icing risk R ∈ [0, 1]; and assuming the fuzzy set of the fuzzy output variable takes the value [0, 2], then the maximum value of the absolute value of the discretized universe of discourse is... Given 2, according to equations (1) and (2), we can obtain: ; ; 4; ;in: A quantification factor representing ambient temperature. The quantization factor representing the blade vibration frequency. A quantitative factor representing ambient humidity. A proportional factor representing the degree of risk of icing.
[0027] S214: Based on the fuzzy set of input variables and the corresponding quantization factor and scaling factor of the blade icing risk assessment model, determine the membership function between the corresponding input variables and the fuzzy set, and establish a blade icing risk assessment model based on fuzzy logic.
[0028] Since the icing protection system needs to respond to environmental changes in real time, the triangular membership function, which has high computational efficiency, is chosen as the membership function between the input variable and the fuzzy set.
[0029] Specifically, the membership function of the low-temperature environment temperature and the fuzzy set is given by equation (3), the membership function of the medium-temperature environment temperature and the fuzzy set is given by equation (4), the membership function of the high-temperature environment temperature and the fuzzy set is given by equation (5), the membership function of the normal leaf vibration frequency and the fuzzy set is given by equation (6), the membership function of the slightly abnormal leaf vibration frequency and the fuzzy set is given by equation (7), the membership function of the severely abnormal leaf vibration frequency and the fuzzy set is given by equation (8), the membership function of the low-humidity environment humidity and the fuzzy set is given by equation (9), the membership function of the medium-humidity environment humidity and the fuzzy set is given by equation (10), and the membership function of the high-humidity environment humidity and the fuzzy set is given by equation (11). (3); (4); (5); (6); (7); (8); (9); (10); (11); in: The membership function representing the temperature of a low-temperature environment and the fuzzy set. The membership function representing the temperature of a medium-temperature environment and the fuzzy set. The membership function representing the temperature of a high-temperature environment and the fuzzy set. Indicates ambient temperature. The membership function representing the normal blade vibration frequency and the fuzzy set. The membership function of the fuzzy set represents the frequency of slight abnormal blade vibration. The membership function of the fuzzy set represents the frequency of severely abnormal blade vibration. Indicates the blade vibration frequency. The membership function representing the humidity of a low-humidity environment and the fuzzy set. The membership function of a fuzzy set represents the humidity in a moderately humid environment. The membership function representing the humidity of a fuzzy set in a high-humidity environment. Indicates ambient humidity.
[0030] A schematic diagram of the relationship between ambient temperature and the membership function of a fuzzy set is shown below. Figure 2 As shown in the diagram, the blade vibration frequency is related to the membership function of the fuzzy set. Figure 3 As shown in the diagram, the relationship between ambient humidity and the membership function of a fuzzy set is illustrated below. Figure 4 As shown.
[0031] S3: Establish a fuzzy inference rule table for the blade icing risk assessment model based on the historical operating status data of wind turbines, input the key factor data of real-time monitoring into the blade icing risk assessment model, and obtain the abstract risk level based on the fuzzy inference rule table; Specifically, the fuzzy inference rule table for the established blade icing risk assessment model is shown in Table 1: Table 1
[0032] Taking rule 1 as an example: If T is in set L, S is in set L, and F is in set L, then the output result is that the icing risk level is in set L. Since T's membership degree in set L is... S is the membership degree of set L. F is the membership degree of set L. Therefore, the membership degree of the output result is min( , , .
[0033] S4: The abstract risk level is transformed into the membership degree of the output result through the blade icing risk assessment model. Then, the abstract risk level is defuzzified by the weighted average method based on the membership degree of the output result to obtain the accurate blade icing risk value and the corresponding accurate blade icing risk level. Specifically, the abstract risk level output value can be defuzzified using the weighted average method according to equation (12): (12); in: This indicates the estimated risk level of leaf icing. The first one in the fuzzy inference rule table represents the... The membership degree of the rule output results. Indicates the weighting coefficient. The size can be 0.5. Indicates the total number of rules.
[0034] S5: Issues early warnings based on the blade icing risk level, enabling multi-level icing early warning and protection for wind turbine blades.
[0035] Specifically, the precise blade icing risk value corresponds to the precise blade icing risk level, and the warning is issued based on the blade icing risk level to achieve multi-level icing early warning and protection for wind turbine blades, as shown in Table 2: Table 2
[0036] A fuzzy early warning system for wind turbine blade icing, used to execute a fuzzy early warning method for wind turbine blade icing as described in any of the above, comprising a key factor determination and acquisition module, a blade icing risk assessment model construction module, an abstract risk level acquisition module, a precise blade icing risk level acquisition module, and an early warning module. The key factor identification and acquisition module is used to identify the key factors affecting the risk of icing on wind turbine blades and to monitor these key factors in real time. The blade icing risk assessment model construction module is used to establish a blade icing risk assessment model based on fuzzy logic according to the key factors affecting the icing risk of wind turbine blades. The abstract risk level acquisition module is used to establish a fuzzy inference rule table for the blade icing risk assessment model based on the historical operating status data of the wind turbine, input the real-time monitored key factor data into the blade icing risk assessment model, and obtain the abstract risk level based on the fuzzy inference rule table. The precise blade icing risk level acquisition module is used to convert the abstract risk level into the membership degree of the output result through the blade icing risk assessment model. Then, based on the membership degree of the output result, the weighted average method is used to defuzzify the abstract risk level to obtain the precise blade icing risk value and the corresponding precise blade icing risk level. The early warning module is used to issue early warnings based on the blade icing risk level, thereby achieving multi-level icing early warning and protection for wind turbine blades.
[0037] In summary, the fuzzy early warning method and system for wind turbine blade icing provided by this invention uses ambient temperature, humidity, and blade vibration frequency as comprehensive reference parameters. By establishing an icing risk assessment model based on fuzzy logic, fuzzy reasoning and defuzzification are performed to quickly obtain icing risk information. Then, an early warning is issued based on the icing risk information. This method has the advantages of fast response and low cost, and solves the problems of slow response and high false alarm rate in existing wind turbine blade icing detection technologies, thus ensuring the safe operation of wind turbines.
[0038] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A fuzzy early warning method for icing on wind turbine blades, characterized in that: Includes the following steps: S1: Identify the key factors affecting the risk of icing on wind turbine blades and monitor these key factors in real time; S2: Establish a blade icing risk assessment model based on fuzzy logic according to the key factors affecting the icing risk of wind turbine blades; S3: Establish a fuzzy inference rule table for the blade icing risk assessment model based on the historical operating status data of wind turbines, input the key factor data of real-time monitoring into the blade icing risk assessment model, and obtain the abstract risk level based on the fuzzy inference rule table; S4: The abstract risk level is transformed into the membership degree of the output result through the blade icing risk assessment model. Then, the abstract risk level is defuzzified by the weighted average method based on the membership degree of the output result to obtain the accurate blade icing risk value and the corresponding accurate blade icing risk level. S5: Issues early warnings based on the blade icing risk level, enabling multi-level icing early warning and protection for wind turbine blades.
2. The method for fuzzy early warning of icing on wind turbine blades according to claim 1, characterized in that: The key factors affecting the risk of icing on wind turbine blades mentioned in step S1 include ambient temperature, blade vibration frequency, and ambient humidity.
3. The method for fuzzy early warning of icing on wind turbine blades according to claim 2, characterized in that: In step S1, the ambient temperature is monitored in real time by a temperature sensor, the blade vibration frequency is monitored in real time by an acceleration sensor, and the ambient humidity is monitored in real time by a humidity sensor.
4. The method for fuzzy early warning of icing on wind turbine blades according to claim 1, characterized in that: The method for establishing a fuzzy logic-based blade icing risk assessment model in step S2, based on the key factors affecting the icing risk of wind turbine blades, is as follows: S211: Using real-time monitoring data of key factors as the precise input variable of the leaf icing risk assessment model, and the estimated degree of leaf icing risk as the precise output variable of the leaf icing risk assessment model, the basic domain of the precise input and output variables of the leaf icing risk assessment model is established. S212: Discretize the basic universe of discourse of the precise input variables and precise output variables of each blade icing risk assessment model, and perform targeted fuzzification processing on the precise input variables and precise output variables of each blade icing risk assessment model to obtain the fuzzy set of fuzzy output variables and the fuzzy set of input variables of the blade icing risk assessment model. S213: Calculate the corresponding quantification factor and proportional factor based on the basic domain of the precise input variables and precise output variables of the leaf icing risk assessment model; S214: Based on the fuzzy set of input variables and the corresponding quantization factor and scaling factor of the blade icing risk assessment model, determine the membership function between the corresponding input variables and the fuzzy set, and establish a blade icing risk assessment model based on fuzzy logic.
5. The method for fuzzy early warning of icing on wind turbine blades according to claim 4, characterized in that: The fuzzy set of the fuzzy output variables of the blade icing risk assessment model in step S212 is {L, M, H}, where L represents low risk, M represents medium risk, and H represents high risk, and the value of the fuzzy set of the fuzzy output variables is [0,2].
6. The method for fuzzy early warning of icing on wind turbine blades according to claim 4, characterized in that: In step S213, the quantization factor is calculated according to equation (1), and the scaling factor is calculated according to equation (2): （1）； （2）； in: Indicates the quantification factor. Indicates the scaling factor. This represents the maximum absolute value of the discretized universe of discourse. Denotes the maximum value of the fundamental universe of discourse. It represents the minimum value of the fundamental universe of discourse.
7. The method for fuzzy early warning of icing on wind turbine blades according to claim 4, characterized in that: The membership function of the low-temperature environment temperature and the fuzzy set determined in step S214 is Equation (3), the membership function of the medium-temperature environment temperature and the fuzzy set is Equation (4), the membership function of the high-temperature environment temperature and the fuzzy set is Equation (5), the membership function of the normal leaf vibration frequency and the fuzzy set is Equation (6), the membership function of the slightly abnormal leaf vibration frequency and the fuzzy set is Equation (7), the membership function of the severely abnormal leaf vibration frequency and the fuzzy set is Equation (8), the membership function of the low-humidity environment humidity and the fuzzy set is Equation (9), the membership function of the medium-humidity environment humidity and the fuzzy set is Equation (10), and the membership function of the high-humidity environment humidity and the fuzzy set is Equation (11). （3）； （4）； （5）； （6）； （7）； （8）； （9）； （10）； （11）； in: The membership function representing the temperature of a low-temperature environment and the fuzzy set. The membership function representing the temperature of a medium-temperature environment and the fuzzy set. The membership function representing the temperature of a high-temperature environment and the fuzzy set. Indicates ambient temperature. The membership function representing the normal blade vibration frequency and the fuzzy set. The membership function of the fuzzy set represents the frequency of slight abnormal blade vibration. The membership function of the fuzzy set represents the frequency of severely abnormal blade vibration. Indicates the blade vibration frequency. The membership function representing the humidity of a low-humidity environment and the fuzzy set. The membership function of a fuzzy set represents the humidity in a moderately humid environment. The membership function representing the humidity of a fuzzy set in a high-humidity environment. Indicates ambient humidity.
8. The method for fuzzy early warning of icing on wind turbine blades according to claim 1, characterized in that: In step S4, the weighted average method is used to defuzzify the abstract risk level output value according to equation (12): （12）； in: This indicates the estimated risk level of leaf icing. The first one in the fuzzy inference rule table represents the... The membership degree of the rule output results. Indicates the weighting coefficient. Indicates the total number of rules.
9. A fuzzy early warning system for wind turbine blade icing, used to execute a fuzzy early warning method for wind turbine blade icing as described in any one of claims 1 to 8, characterized in that, It includes a key factor identification and data collection module, a blade icing risk assessment model construction module, an abstract risk level acquisition module, a precise blade icing risk level acquisition module, and an early warning module; The key factor identification and acquisition module is used to identify the key factors affecting the risk of icing on wind turbine blades and to monitor these key factors in real time. The blade icing risk assessment model construction module is used to establish a blade icing risk assessment model based on fuzzy logic according to the key factors affecting the icing risk of wind turbine blades. The abstract risk level acquisition module is used to establish a fuzzy inference rule table for the blade icing risk assessment model based on the historical operating status data of the wind turbine, input the real-time monitored key factor data into the blade icing risk assessment model, and obtain the abstract risk level based on the fuzzy inference rule table. The precise blade icing risk level acquisition module is used to convert the abstract risk level into the membership degree of the output result through the blade icing risk assessment model. Then, based on the membership degree of the output result, the weighted average method is used to defuzzify the abstract risk level to obtain the precise blade icing risk value and the corresponding precise blade icing risk level. The early warning module is used to issue early warnings based on the blade icing risk level, thereby achieving multi-level icing early warning and protection for wind turbine blades.