A method and system for predicting the life of fan parts

Abstract

The present invention discloses a method and system for predicting the life of fan parts. The method comprises the following steps: step 1: real-time collection of operating data of various parts of the fan; step 2: pre-processing the operating data to obtain pre-processed data; step 3: performing abnormality identification on the pre-processed data according to an early warning model to obtain early warning data; step 4: extracting the pre-processed data of abnormal fan parts in the early warning data, and if there are no abnormal fan parts, executing step 1; step 5: performing remaining life prediction on the pre-processed data of abnormal fan parts according to a life prediction model and obtaining life prediction data; the system comprises an operating data acquisition module, a pre-processing module, an abnormality identification module, and a life prediction module.

Description

Technical Field

[0001] The present invention relates to the field of life prediction of fan parts, and in particular to a method and system for predicting the life of fan parts. Background Art

[0002] As the installed capacity of wind turbines increases year by year, new problems arise. The gap between wind turbine operation time and total downtime is huge. Wind turbine failure is an important reason for insufficient wind turbine operation time. The failure rate and repair cost of key components in wind turbine failures account for the vast majority of wind turbine failures. The key components of wind turbines mainly include: gearbox system, hub, transmission system, generator, blades, yaw system, pitch system, sensor, control system, electrical system, etc. Therefore, effective monitoring and prediction of key components of wind turbines have become the main research objects. With the rapid development of information sensing technology, industrial systems have realized real-time monitoring of the operating status of key components of the system, analyzed and mined the data after signal processing, detected the operating status of the system to obtain the remaining life prediction value of the system, thereby identifying the occurrence of wind turbine failures, and rationally planning the wind turbine operation and maintenance based on the prediction results, thereby formulating a reasonable operation and maintenance strategy;

[0003] In the current research on life prediction of key components of wind turbines, the life prediction method based on physical models mainly predicts the remaining life of components based on the principle of component load and fatigue cumulative damage, and obtains the fatigue life of wind turbine components through simulation through finite element analysis method from the perspective of load analysis. This method cannot be directly applied to the life prediction of key components of wind turbines because it does not fully consider the actual operating environment and operating status of wind turbines. Summary of the Invention

[0004] The technical problem to be solved by the present invention is that the current method cannot be directly applied to the life prediction of key components of wind turbines because it does not fully consider the actual operating environment and operating status of wind turbines. The present invention provides a method for predicting the life of wind turbine parts. The present invention also provides a system for predicting the life of wind turbine parts. It uses a data-driven method to bypass the physical analysis of the equipment performance degradation process, and directly uses various statistical methods and intelligent algorithms to perform statistical analysis and information mining on the historical operation monitoring data of similar equipment to obtain certain characteristics in the equipment degradation process for predicting the life of individual equipment, thereby forming a data-driven life prediction method. This method evaluates the current operating status of the parts based on real-time data collected during the operation of the wind turbine, thereby obtaining a prediction result that is more in line with reality, which is used to solve the defects caused by the existing technology.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, a method for predicting the life of a wind turbine component comprises the following steps:

[0007] Step 1: Collect the operating data of each part of the fan in real time;

[0008] Step 2: Preprocessing the operating data to obtain preprocessed data;

[0009] Step 3: Perform anomaly identification on the pre-processed data according to the early warning model to obtain early warning data;

[0010] Step 4: extracting pre-processed data of abnormal fan parts from the warning data. If there are no abnormal fan parts, proceed to step 1.

[0011] Step 5: Based on the life prediction model, the remaining life of the pre-processed data of the abnormal fan parts is predicted and the life prediction data is obtained.

[0012] In the above-mentioned method for predicting the life of wind turbine parts, the specific method for preprocessing the operating data in step 2 is as follows:

[0013] Denoising, cleaning, and classifying the operating data, and then extracting features to obtain feature data;

[0014] The characteristic data is stored.

[0015] In the above-mentioned method for predicting the life of a fan component, the specific method for obtaining the warning data by performing abnormality identification on the pre-processed data according to the warning model in step 3 is as follows:

[0016] Acquiring the characteristic data of each component of the fan;

[0017] Using clustering and linear regression algorithms, the characteristic data is input into the early warning model to obtain the normal range [α1, α2], the high limit range [β1, β2], and the super high limit range [γ1, γ2] of the wind turbine operating parameters;

[0018] The early warning model predicts the characteristic data of the fan parts based on the autoregressive model and the long short-term memory neural network model, inputs the characteristic data into the early warning model and outputs the predicted operation value of the fan parts in the future;

[0019] It is determined whether the predicted value is within the normal range, the high limit range, or the ultra-high limit range, and the range is output as the warning data.

[0020] In the above-mentioned method for predicting the life of wind turbine parts, the method for constructing the early warning model is as follows:

[0021] Collect historical operating data of each part of the fan;

[0022] Normalizing, cleaning, and labeling the historical operation data to generate a training sample data set;

[0023] The training sample data set is trained according to clustering and linear regression algorithms to obtain multi-level interval parameters under the state trend of the equipment operation process;

[0024] The autoregressive model and the long short-term memory network model are used to predict the operation trend based on the operation data of each component of the fan, and the prediction model is obtained by optimization;

[0025] The multi-level interval parameters and the prediction model constitute the early warning model.

[0026] In the above-mentioned method for predicting the life of wind turbine parts, the pre-processing of the operating data in step 2 further comprises the following steps:

[0027] Obtain fault maps of each fan component;

[0028] Extracting the characteristic data according to the fault map to obtain fault operating parameter change characteristic data of the faulty part, wherein the fault operating parameter change characteristic data records the faulty part data, fault occurrence time, and fault change characteristics;

[0029] Performing an operation change trend analysis on the fault operation parameter change characteristic data over the entire life cycle of each wind turbine component according to the degradation linear and nonlinear process characteristics to obtain operation change trend data, wherein the operation change trend data records the component degradation measurement value;

[0030] The fault operation parameter change characteristic data and the operation change trend data are stored.

[0031] In the above-mentioned method for predicting the life of a fan component, the method for constructing the life prediction model is as follows:

[0032] The life of the parts is defined according to the formula T = inf{t: X(t)≥ω|X(0)<ω};

[0033] Where T is the life of the part, ω is the failure threshold preset for the part, X(t) is the degradation process state at time t, and X(0) is the degradation amount when the part is first put into use;

[0034] According to the formula L k =inf{l k >0:X(l k +t k )≥ω} construct the life prediction model and calculate t k Remaining life of parts at the moment;

[0035] Among them, Lk is the remaining life, X is the state quantity of the equipment degradation process at the corresponding moment, inf(x) is the lower limit function, l k is the residual degradation amount at the corresponding moment, t k is the component life at the corresponding moment, and ω is the preset failure threshold for the component.

[0036] Based on the established life prediction model and combined with the errors existing in the data measurement process, a degradation state update prediction model is established, and the position parameters in the established life prediction model are estimated using the maximum likelihood estimation method to obtain the remaining life probability density function, and the remaining life prediction result is obtained based on the measurement value data Y1:k of the measurement error.

[0037] In the above-mentioned method for predicting the life of a fan component, the degradation process state quantity X(t) following the linear Wiener process is expressed as:

[0038] X(t)=λt+σB(t);

[0039] Where λ is the drift coefficient, which reflects the degradation rate of the part when representing the degradation process of the part, σ is the diffusion coefficient, B(t) represents the standard Brownian motion, which is used to describe the uncertainty of the random degradation amount on the time axis, λt is the mean value of X(t), and t is the time point;

[0040] The state variable X(t) of the degradation process following the nonlinear Wiener process is expressed as:

[0041]

[0042] Where X0 is the degradation amount at time t=0, μ(t;θ) represents the drift parameter of the degradation process, θ is the parameter vector; σ B is the diffusion coefficient of the degradation process; the Brownian motion represented by B(t) obeys the normal distribution, B(t) ~ N(0,t), which is used to describe the random dynamics of the degradation process and the uncertainty of the degradation process over time;

[0043] Based on the characteristics of components, component degradation can be divided into linear degradation and nonlinear degradation. According to the two different degradation characteristics, the life of key components is defined by following the linear Wiener process and the nonlinear Wiener process respectively. The concept of the first arrival time of the random degradation process is used to define the component life. That is, when the state quantity X(t) of the random degradation process reaches the failure threshold set by expert experience or component standards for the first time, the component is considered to have failed.

[0044] The data-driven approach bypasses the physical analysis of the equipment degradation process and directly uses various statistical methods and intelligent algorithms to conduct statistical analysis and information mining on the historical operation monitoring data of similar equipment. This method obtains certain characteristics of the equipment degradation process and uses them to predict the life of individual equipment.

[0045] By real-time monitoring of the operating parameters of key wind turbine components, abnormal warnings and remaining life predictions can be carried out to assist in the reasonable arrangement of wind turbine operation and maintenance strategies to ensure the safe and stable operation of the wind turbine. When components are updated or repaired, the degradation state threshold needs to be corrected to obtain a more accurate life prediction.

[0046] In a second aspect, a wind turbine component life prediction system includes an operation data acquisition module, a preprocessing module, an abnormality identification module, and a life prediction module;

[0047] The operation data acquisition module is used to collect the operation data of each component of the fan in real time;

[0048] The preprocessing module is used to preprocess the operating data to obtain preprocessed data;

[0049] The anomaly recognition module is used to perform anomaly recognition on the pre-processed data according to the early warning model to obtain early warning data;

[0050] It is also used to extract pre-processed data of abnormal fan parts from the warning data, and if there are no abnormal fan parts, generate a feedback signal transmitted to the operation data acquisition module;

[0051] The life prediction module is used to perform remaining life prediction on pre-processed data of abnormal fan parts according to the life prediction model and obtain life prediction data.

[0052] In a third aspect, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of the method described in any one of the first aspects are implemented.

[0053] The technical solution provided by the above-mentioned method and system for predicting the life of wind turbine parts of the present invention has the following technical effects:

[0054] The present invention mainly uses a data-driven approach to bypass the physical analysis of the equipment performance degradation process, and directly conducts statistical analysis and information mining on the historical operation monitoring data of similar equipment through various statistical methods and intelligent algorithms, and obtains certain characteristics of the equipment degradation process for life prediction of individual equipment, forming a data-driven life prediction method. This method evaluates the current operating status of parts based on real-time data collected during the operation of the fan, thereby obtaining a more realistic prediction result. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flow chart of a method for predicting the life of a fan component according to the present invention;

[0056] Figure 2 This is a structural diagram of a fan parts life prediction system according to the present invention.

[0057] The accompanying drawings are numerals as follows:

[0058] Operation data acquisition module 100, pre-processing module 200, abnormality identification module 300, life prediction module 400. DETAILED DESCRIPTION

[0059] In order to make the technical means, creative features, objectives and effects of the invention easy to understand, the technical solutions in the embodiments of the present invention are clearly and completely described below in combination with specific illustrations. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0060] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of the present invention.

[0061] It should be noted that the structures, proportions, sizes, etc. illustrated in the drawings in this specification are only used to match the contents disclosed in the specification so that people familiar with this technology can understand and read them. They are not used to limit the conditions under which the present invention can be implemented. Therefore, they have no substantive technical significance. Any modification of the structure, change in the proportion relationship or adjustment of the size should still fall within the scope of the technical content disclosed in the present invention without affecting the efficacy and purpose that can be achieved by the present invention.

[0062] At the same time, the terms such as "upper", "lower", "left", "right", "middle" and "one" quoted in this specification are only for the convenience of description and are not used to limit the scope of implementation of the present invention. Changes or adjustments to their relative relationships should be regarded as the scope of implementation of the present invention without substantially changing the technical content.

[0063] The first embodiment of the present invention is to provide a method for predicting the life of fan parts, and the second embodiment of the present invention is to provide a system for predicting the life of fan parts. The purpose is to use a data-driven method to bypass the physical analysis of the equipment performance degradation process, and directly use various statistical methods and intelligent algorithms to perform statistical analysis and information mining on the historical operation monitoring data of similar equipment, so as to obtain certain characteristics of the equipment degradation process for predicting the life of individual equipment, thereby forming a data-driven life prediction method. This method is to evaluate the current operating status of the parts based on the real-time data collected during the operation of the fan, so as to obtain a prediction result that is more in line with reality.

[0064] like Figure 1 As shown, in a first aspect and a first embodiment, a method for predicting the life of a wind turbine component comprises the following steps:

[0065] Step 1: Collect the operating data of each part of the fan in real time;

[0066] Step 2: Preprocess the running data to obtain preprocessed data;

[0067] Step 3: Identify anomalies in the preprocessed data based on the early warning model to obtain early warning data;

[0068] Step 4: Extract pre-processed data of abnormal fan parts from the warning data. If there are no abnormal fan parts, execute step 1.

[0069] Step 5: Based on the life prediction model, the remaining life of the pre-processed data of the abnormal fan parts is predicted and the life prediction data is obtained.

[0070] In the above-mentioned method for predicting the life of a fan component, the specific method for preprocessing the operating data in step 2 is as follows:

[0071] After denoising, cleaning and classifying the operating data, feature extraction is performed to obtain feature data;

[0072] Store feature data.

[0073] In the above-mentioned method for predicting the life of a fan component, the specific method for obtaining the warning data by performing abnormality identification on the preprocessed data according to the warning model in step 3 is as follows:

[0074] Obtain characteristic data of each part of the fan;

[0075] Using clustering and linear regression algorithms, the characteristic data is input into the early warning model to obtain the normal range [α1, α2], high limit range [β1, β2], and super high limit range [γ1, γ2] of the wind turbine operating parameters;

[0076] The early warning model predicts the characteristic data of wind turbine parts based on the autoregressive model and the long-short-term memory neural network model. The characteristic data is input into the early warning model to output the predicted operation value of the wind turbine parts in the future.

[0077] Determine whether the predicted value is within the normal range, high limit range, or ultra-high limit range, and output the range as warning data.

[0078] In the above-mentioned method for predicting the life of wind turbine parts, the method for constructing the early warning model is as follows:

[0079] Collect historical operating data of each part of the fan;

[0080] Normalize, clean, and label historical operation data to generate a training sample data set;

[0081] The training sample data set is trained according to the clustering and linear regression algorithm to obtain the multi-level interval parameters under the state trend of the equipment operation process;

[0082] The autoregressive model and long short-term memory network model are used to predict the operation trend based on the operation data of each part of the fan, and the prediction model is obtained through optimization;

[0083] Multi-level interval parameters and prediction models constitute the early warning model.

[0084] In the above-mentioned method for predicting the life of wind turbine parts, the pre-processing of the operating data in step 2 further includes the following steps:

[0085] Obtain fault maps of each fan component;

[0086] Extracting characteristic data based on the fault map to obtain fault operating parameter change characteristic data of the faulty part, the fault operating parameter change characteristic data records the faulty part data, fault occurrence time, and fault change characteristics;

[0087] According to the linear and nonlinear process characteristics of degradation, the fault operating parameter change characteristic data is analyzed for the operation change trend of each wind turbine component throughout its life cycle to obtain the operation change trend data, which records the component degradation measurement value;

[0088] The fault operation parameter change characteristic data and operation change trend data are stored.

[0089] In the above-mentioned method for predicting the life of a fan component, the method for constructing the life prediction model is as follows:

[0090] The life of the parts is defined according to the formula T = inf{t: X(t)≥ω|X(0)<ω};

[0091] Where T is the life of the part, ω is the failure threshold preset for the part, X(t) is the degradation process state at time t, and X(0) is the degradation amount when the part is first put into use;

[0092] According to the formula L k =inf{l k >0:X(l k +t k )≥ω} to build a life prediction model and calculate t k Remaining life of parts at the moment;

[0093] Among them, L k is the remaining life, X is the state quantity of the equipment degradation process at the corresponding moment, inf(x) is the lower limit function, l k is the residual degradation amount at the corresponding moment, t k is the component life at the corresponding moment, and ω is the preset failure threshold for the component.

[0094] Based on the established life prediction model and combined with the errors in the data measurement process, a degradation state update prediction model is established. The position parameters in the established life prediction model are estimated using the maximum likelihood estimation method to obtain the remaining life probability density function. The remaining life prediction result is obtained based on the measurement value data Y1:k of the measurement error.

[0095] In the above-mentioned method for predicting the life of a fan component, the degradation process state quantity X(t) following the linear Wiener process is expressed as:

[0096] X(t)=λt+σB(t);

[0097] Where λ is the drift coefficient, which reflects the degradation rate of the part when representing the degradation process of the part, σ is the diffusion coefficient, B(t) represents the standard Brownian motion, which is used to describe the uncertainty of the random degradation amount on the time axis, λt is the mean value of X(t), and t is the time point;

[0098] The state variable X(t) of the degradation process following the nonlinear Wiener process is expressed as:

[0099]

[0100] Where X0 is the degradation amount at time t=0, μ(t;θ) represents the drift parameter of the degradation process, θ is the parameter vector; σ B is the diffusion coefficient of the degradation process; the Brownian motion represented by B(t) obeys the normal distribution, B(t) ~ N(0,t), which is used to describe the random dynamics of the degradation process and the uncertainty of the degradation process over time;

[0101] Based on the characteristics of components, component degradation can be divided into linear degradation and nonlinear degradation. According to the two different degradation characteristics, the life of key components is defined by following the linear Wiener process and the nonlinear Wiener process respectively. The concept of the first arrival time of the random degradation process is used to define the component life. That is, when the state quantity X(t) of the random degradation process reaches the failure threshold set by expert experience or component standards for the first time, the component is considered to have failed.

[0102] The data-driven approach bypasses the physical analysis of the equipment degradation process and directly uses various statistical methods and intelligent algorithms to conduct statistical analysis and information mining on the historical operation monitoring data of similar equipment. This method obtains certain characteristics of the equipment degradation process and uses them to predict the life of individual equipment.

[0103] By real-time monitoring of the operating parameters of key wind turbine components, abnormal warnings and remaining life predictions can be carried out to assist in the reasonable arrangement of wind turbine operation and maintenance strategies to ensure the safe and stable operation of the wind turbine. When components are updated or repaired, the degradation state threshold needs to be corrected to obtain a more accurate life prediction.

[0104] like Figure 2 As shown, in the second aspect, the second embodiment, a wind turbine component life prediction system includes an operation data acquisition module 100, a preprocessing module 200, an abnormality identification module 300, and a life prediction module 400;

[0105] The operation data acquisition module 100 is used to collect the operation data of each component of the wind turbine in real time;

[0106] The preprocessing module 200 is used to preprocess the operating data to obtain preprocessed data;

[0107] The anomaly identification module 300 is used to identify anomalies in the pre-processed data according to the early warning model to obtain early warning data;

[0108] It is also used to extract pre-processed data of abnormal fan parts from the warning data. If there are no abnormal fan parts, it generates a feedback signal that is transmitted to the operation data acquisition module 100;

[0109] The life prediction module 400 is used to perform remaining life prediction on the pre-processed data of abnormal wind turbine parts according to the life prediction model and obtain life prediction data.

[0110] In a third aspect, a computer-readable storage medium is provided, wherein a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0111] For example, the memory may include random access memory, flash memory, read-only memory, programmable read-only memory, non-volatile memory or registers;

[0112] The processor may be a central processing unit (CPU) or a graphics processing unit (GPU). The memory may store executable instructions.

[0113] The processor may execute the execution instructions stored in the memory to implement the various processes described herein.

[0114] It will be appreciated that the memory in this embodiment may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories;

[0115] Among them, the non-volatile memory can be ROM (Read-Only Memory), PROM (Programmable ROM), EPROM (Erasable PROM), EEPROM (Electrically EPROM) or flash memory.

[0116] The volatile memory may be RAM (Random Access Memory), which is used as an external cache;

[0117] By way of example and not limitation, many forms of RAM are available, such as SRAM (Static RAM), DRAM (Dynamic RAM), SDRAM (Synchronous DRAM), DDR SDRAM (Double Data Rate SDRAM), ESDRAM (Enhanced SDRAM), SLDRAM (Synchlink DRAM), and DRRAM (Direct Rambus RAM). The memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0118] In some embodiments, the memory stores the following elements, upgrade packages, executable units or data structures, or a subset thereof, or an extended set thereof: an operating system and applications;

[0119] The operating system includes various system programs, such as the framework layer, core library layer, and driver layer, which are used to implement various basic services and process hardware-based tasks.

[0120] The application program includes various application programs for implementing various application services. The program for implementing the method of the embodiment of the present invention can be included in the application program.

[0121] It will be apparent to those skilled in the art that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of software and electronic hardware;

[0122] Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution;

[0123] Those skilled in the art may 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.

[0124] In the embodiments of the present application, the disclosed systems, devices, and methods may be implemented in other ways;

[0125] For example, the division of units or modules is only a logical function division, and there may be other division methods in actual implementation;

[0126] For example, multiple units or modules or components may be combined or may be integrated into another system;

[0127] In addition, each functional unit or module in the embodiment of the present application may be integrated into one processing unit or module, or may exist as a separate physical entity, etc.

[0128] It should be understood that in the various embodiments of the present application, the size of the serial number of each process does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0129] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a machine-readable storage medium;

[0130] Therefore, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a machine-readable storage medium and can include several instructions for causing the electronic component to execute all or part of the process of the technical solution described in the embodiments of the present application;

[0131] The above-mentioned storage medium may include ROM, RAM, removable disk, hard disk, magnetic disk or optical disk, etc., which can store program codes.

[0132] In summary, the present invention provides a method and system for predicting the life of fan parts. The method uses a data-driven approach to bypass the physical analysis of the equipment performance degradation process, and directly performs statistical analysis and information mining on the historical operation monitoring data of similar equipment through various statistical methods and intelligent algorithms. Certain characteristics of the equipment degradation process are obtained for predicting the life of individual equipment, forming a data-driven life prediction method. This method evaluates the current operating status of the parts based on the real-time data collected during the operation of the fan, thereby obtaining a prediction result that is more in line with reality.

[0133] The above describes specific embodiments of the invention. It should be understood that the invention is not limited to the specific embodiments described above. Parts and structures not described in detail should be understood to be implemented in a common manner in the art. Those skilled in the art may make various modifications, variations, or simple deductions, deformations, or substitutions within the scope of the claims, without affecting the essence of the invention.

Claims

1. A method for predicting the life of a fan component, characterized in that: The following steps are involved: Step 1: Collect the operating data of each part of the fan in real time; Step 2: Preprocessing the operating data to obtain preprocessed data; Step 3: Perform anomaly identification on the pre-processed data according to the early warning model to obtain early warning data; Step 4: extracting pre-processed data of abnormal fan parts from the warning data. If there are no abnormal fan parts, proceed to step 1. Step 5: Predict the remaining life of the pre-processed data of the abnormal fan parts according to the life prediction model and obtain life prediction data; The method for constructing the life prediction model is as follows: The life of the parts is defined according to the formula T = inf{t: X(t)≥ω|X(0)<ω}; Where T is the life of the part, ω is the failure threshold preset for the part, X(t) is the degradation process state at time t, and X(0) is the degradation amount when the part is first put into use; According to the formula L k =inf{l k >0:X(l k +t k )≥ω} construct the life prediction model and calculate t k Remaining life of parts at the moment; Among them, L k is the remaining life, X is the state quantity of the equipment degradation process at the corresponding moment, inf(x) is the lower limit function, l k is the residual degradation amount at the corresponding moment, t k is the component life at the corresponding moment, ω is the failure threshold preset for the component; The degradation process state quantity X(t) following the linear Wiener process is expressed as: X(t)=λt+σB(t); Where λ is the drift coefficient, which reflects the degradation rate of the part when representing the degradation process of the part, σ is the diffusion coefficient, B(t) represents the standard Brownian motion, which is used to describe the uncertainty of the random degradation amount on the time axis, λt is the mean value of X(t), and t is the time point; The state variable X(t) of the degradation process following the nonlinear Wiener process is expressed as: X(t)=x0+∫0 t μ(τ;θ)dτ+σ B B(t); Where X0 is the degradation amount at time t=0, μ(t;θ) represents the drift parameter of the degradation process, θ is the parameter vector; σ B is the diffusion coefficient of the degradation process; the Brownian motion represented by B(t) obeys the normal distribution, B(t) ~ N(0,t), which is used to describe the random dynamics of the degradation process and the uncertainty of the degradation process changing with time.

2. A method for predicting the life of a fan component according to claim 1, characterized in that: The specific method for preprocessing the operating data in step 2 is as follows: Denoising, cleaning, and classifying the operating data, and then extracting features to obtain feature data; The characteristic data is stored.

3. A method for predicting the life of a fan component according to claim 2, characterized in that: The specific method for performing anomaly identification on the pre-processed data according to the early warning model in step 3 to obtain early warning data is as follows: Acquiring the characteristic data of each component of the fan; Using clustering and linear regression algorithms, the characteristic data is input into the early warning model to obtain the normal range [α1, α2], the high limit range [β1, β2], and the super high limit range [γ1, γ2] of the wind turbine operating parameters; The early warning model predicts the characteristic data of the fan parts based on the autoregressive model and the long short-term memory neural network model, inputs the characteristic data into the early warning model and outputs the predicted operation value of the fan parts in the future; It is determined whether the predicted value is within the normal range, the high limit range, or the ultra-high limit range, and the range is output as the warning data.

4. A method for predicting the life of a fan component according to any one of claims 2 to 3, characterized in that: The construction method of the early warning model is as follows: Collect historical operating data of each part of the fan; Normalizing, cleaning, and labeling the historical operation data to generate a training sample data set; The training sample data set is trained according to clustering and linear regression algorithms to obtain multi-level interval parameters under the state trend of the equipment operation process; The autoregressive model and the long short-term memory network model are used to predict the operation trend based on the operation data of each component of the fan, and the prediction model is obtained by optimization; The multi-level interval parameters and the prediction model constitute the early warning model.

5. A method for predicting the life of a fan component according to claim 4, characterized in that: The pre-processing of the operating data in step 2 further includes the following steps: Obtain fault maps of each fan component; Extracting the characteristic data according to the fault map to obtain fault operating parameter change characteristic data of the faulty part, wherein the fault operating parameter change characteristic data records the faulty part data, fault occurrence time, and fault change characteristics; Performing an operation change trend analysis on the fault operation parameter change characteristic data over the entire life cycle of each wind turbine component according to the degradation linear and nonlinear process characteristics to obtain operation change trend data, wherein the operation change trend data records the component degradation measurement value; The fault operation parameter change characteristic data and the operation change trend data are stored.

6. A wind turbine parts life prediction system, characterized in that: Contains operation data acquisition module, preprocessing module, abnormality identification module, and life prediction module; The operation data acquisition module is used to collect the operation data of each component of the fan in real time; The preprocessing module is used to preprocess the operating data to obtain preprocessed data; The anomaly recognition module is used to perform anomaly recognition on the pre-processed data according to the early warning model to obtain early warning data; It is also used to extract pre-processed data of abnormal fan parts from the warning data, and if there are no abnormal fan parts, generate a feedback signal transmitted to the operation data acquisition module; The life prediction module is used to predict the remaining life of the pre-processed data of the abnormal fan parts according to the life prediction model and obtain life prediction data; The method for constructing the life prediction model is as follows: The life of the parts is defined according to the formula T = inf{t: X(t)≥ω|X(0)<ω}; Where T is the life of the part, ω is the failure threshold preset for the part, X(t) is the degradation process state at time t, and X(0) is the degradation amount when the part is first put into use; According to the formula L k =inf{l k >0:X(l k +t k )≥ω} construct the life prediction model and calculate t k Remaining life of parts at the moment; Among them, L k is the remaining life, X is the state quantity of the equipment degradation process at the corresponding moment, inf(x) is the lower limit function, l k is the residual degradation amount at the corresponding moment, t k is the component life at the corresponding moment, ω is the failure threshold preset for the component; The degradation process state quantity X(t) following the linear Wiener process is expressed as: X(t)=λt+σB(t); Where λ is the drift coefficient, which reflects the degradation rate of the part when representing the degradation process of the part, σ is the diffusion coefficient, B(t) represents the standard Brownian motion, which is used to describe the uncertainty of the random degradation amount on the time axis, λt is the mean value of X(t), and t is the time point; The state variable X(t) of the degradation process following the nonlinear Wiener process is expressed as: X(t)=x0+∫0 t μ(τ;θ)dτ+σ B B(t); Where X0 is the degradation amount at time t=0, μ(t;θ) represents the drift parameter of the degradation process, θ is the parameter vector; σ B is the diffusion coefficient of the degradation process; the Brownian motion represented by B(t) obeys the normal distribution, B(t) ~ N(0,t), which is used to describe the random dynamics of the degradation process and the uncertainty of the degradation process changing with time.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which implements the steps of the method according to any one of claims 1 to 5 when executed by a processor.

Images