CN115659612A - Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration - Google Patents
Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration Download PDFInfo
- Publication number
- CN115659612A CN115659612A CN202211244944.4A CN202211244944A CN115659612A CN 115659612 A CN115659612 A CN 115659612A CN 202211244944 A CN202211244944 A CN 202211244944A CN 115659612 A CN115659612 A CN 115659612A
- Authority
- CN
- China
- Prior art keywords
- typhoon
- impact
- component
- model
- degradation
- 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
- 238000000034 method Methods 0.000 title claims abstract description 76
- 238000006731 degradation reaction Methods 0.000 claims abstract description 76
- 230000015556 catabolic process Effects 0.000 claims abstract description 65
- 238000012544 monitoring process Methods 0.000 claims abstract description 31
- 239000002245 particle Substances 0.000 claims abstract description 9
- 238000005457 optimization Methods 0.000 claims abstract description 8
- 230000007246 mechanism Effects 0.000 claims abstract description 4
- 230000006870 function Effects 0.000 claims description 31
- 230000008569 process Effects 0.000 claims description 26
- 238000012937 correction Methods 0.000 claims description 16
- 230000001186 cumulative effect Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000007704 transition Effects 0.000 claims description 9
- 238000005315 distribution function Methods 0.000 claims description 7
- 230000036541 health Effects 0.000 claims description 7
- 238000012546 transfer Methods 0.000 claims description 6
- 239000013598 vector Substances 0.000 claims description 6
- 230000001419 dependent effect Effects 0.000 claims description 5
- 238000012614 Monte-Carlo sampling Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 4
- 238000007476 Maximum Likelihood Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000000513 principal component analysis Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000017105 transposition Effects 0.000 claims description 2
- 238000010977 unit operation Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000005314 correlation function Methods 0.000 description 3
- 230000007423 decrease Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 241000242759 Actiniaria Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
Images
Landscapes
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a method for predicting the residual life of an offshore wind turbine generator set by considering typhoon influence, which comprises the following steps: constructing a failure rate increment model of the components under typhoon impact based on an influence mechanism of typhoon on degradation of the components of the unit; the method comprises the steps of constructing a model of the degradation and impact correlation of the running state of a component by combining the running state monitoring data of a unit, and quantifying the capacity of bearing external impact; based on the time-varying randomness characteristic of unit performance degradation caused by typhoon random impact, constructing a time-varying failure rate model of the offshore wind turbine unit component considering degradation and impact dependence, constructing a unit component residual life prediction model under a typhoon impact scene based on the model, and obtaining an optimized residual life prediction model by predicting the minimum optimization parameter of errors by adopting a particle swarm optimization algorithm; and modifying and optimizing the residual life prediction model based on the real-time data and realizing residual life prediction. Compared with the prior art, the method has the advantages of high service life prediction accuracy and the like.
Description
Technical Field
The invention relates to the field of life prediction of an offshore wind turbine, in particular to a method for predicting the residual life of an offshore wind turbine by considering typhoon influence.
Background
Offshore wind power is important force for accelerating the promotion of China to realize the '3060' double-carbon strategy, and the high-reliability operation of an offshore wind power generation unit is a key factor of large-scale offshore wind power sustainable development. The method for accurately predicting the residual life of the offshore wind turbine generator has important theoretical significance and engineering application value for reducing the sudden failure risk, operating reliably and maintaining decisions with high efficiency. Coastal areas with large-scale and clustered offshore wind power development in China are seriously affected by severe weather such as typhoon, and the storm and billow caused by the typhoon bring huge challenges to the accurate prediction of the residual life of offshore wind turbines.
The existing research on the aspect of life prediction of an offshore wind turbine mainly has the following two problems: 1) In the past, in the research of prediction of the residual life of the wind turbine generator, the performance of the wind turbine generator is supposed to be gradually deteriorated along with time based on the natural degradation process of the wind turbine generator, the objective random impact influence of the offshore typhoon weather is ignored, and the evolution law of the actual degradation of the wind turbine generator under the severe offshore operation environment is difficult to accurately describe. 2) Under the background of marine multi-source multi-dimensional operation and maintenance big data, the incidence relation between random impact and degradation states of the unit is accurately described, and the problems of accurate modeling of the unit failure process and uncertainty quantification of residual life prediction still need to be solved.
Disclosure of Invention
The invention aims to provide a method for predicting the residual life of an offshore wind turbine generator set by considering typhoon influence, which guarantees the minimum prediction mean square error by considering typhoon weather, realizes the interactive linkage of the unit component state and a random degradation model and obtains an accurate life prediction value of the offshore wind turbine generator set.
The purpose of the invention can be realized by the following technical scheme:
a method for predicting the residual life of an offshore wind turbine generator considering typhoon influence comprises the following steps:
analyzing the influence of direct impact and indirect impact on the component fault rate based on the influence mechanism of the typhoon on the degradation of the unit components, and constructing a component fault rate increment model under the typhoon impact;
aiming at the quantitative problem of the relevance between the degradation process and the impact of the component, a degradation and impact relevance model of the component operation state is constructed by combining the unit operation state monitoring data, and the capacity of bearing the external impact is quantized;
based on the time-varying randomness characteristic of unit performance degradation caused by typhoon random impact, a component failure rate increment model under typhoon impact and a degradation and impact correlation model are combined to construct an offshore wind turbine component time-varying failure rate model considering degradation and impact dependence;
on the basis of a time-varying fault rate model of an offshore wind turbine unit component, under a typhoon impact scene, a residual life prediction model of the unit component is constructed, and parameters of a component fault rate increment model and a degradation and impact correlation model under typhoon impact are optimized by adopting a particle swarm optimization algorithm according to the minimum prediction error, so that an optimized residual life prediction model is obtained;
and based on the optimized residual life prediction model, correcting the optimized residual life prediction model by combining typhoon impact monitoring data and unit real-time running state monitoring data, and realizing residual life prediction.
The construction of the component failure rate incremental model under the typhoon impact comprises the following steps:
acquiring typhoon data in a typhoon multi-occurrence area;
constructing a Batts typhoon wind field model;
assuming that the number of times of typhoon impact of an offshore wind turbine unit in the operation period obeys a Poission process with a parameter of lambda, a typhoon wind field model is kept unchanged in the processes before and after the typhoon passes through a wind power plant, a Monte Carlo sampling method is adopted to obtain the impact probability of each typical typhoon, and a typhoon random impact model is established;
and analyzing the influence of direct impact and indirect impact on the failure rate of the component based on a typhoon random impact model, and constructing a component failure rate increment model under typhoon impact.
The typhoon random impact model is as follows:
where P (N (t) = N) represents the probability that a typhoon has N impacts within (0, t), λ is the arrival rate of the typhoon, f G (t si ) Time t required for the first arrival of the ith impact si Gamma function value is shown as gamma (i); v Rmax The gradient wind speed at the maximum wind radius, K is an empirical parameter, the value range is 6.93-6.97, delta p is the central pressure difference of the typhoon, R max F is a scientific parameter; v denotes the instantaneous wind speed of the typhoon, V s The typhoon moving speed; r is the distance from any position of the typhoon to the center of the typhoon; x is an empirical parameter and has a value range of 0.5-0.7。
The component failure rate increment model under typhoon impact is as follows:
wherein,
representing the component k in an operating state as a function of the dependency of component degradation on impact
Time-dependent impact on the increment of the failure rate, t si And t ei Respectively showing the first arrival moment and the removal moment of the impact influence of the typhoon on the unit, wherein alpha and epsilon are direct or indirect impact correction coefficients of the typhoon of the component, and v is the instantaneous wind speed of the typhoon.
And distinguishing the influence of direct impact and indirect impact, wherein the incremental model of the failure rate of the part under the direct impact is expressed as follows:
the incremental model of component failure rate under indirect impact is expressed as:
wherein, Δ λ sk1i Failure rate of external units, alpha, caused by direct impact of typhoon k1 、ε k1 Respectively, the direct impact correction coefficient of typhoon, w is the wind load perpendicular to the surface of the fan, rho is the air density, C s Is the form factor, S α Is the projection area on the plane vertical to the wind direction; delta lambda sk2i Indicating the failure rate, alpha, of the internal components of the unit due to typhoon indirect impact k2 、ε k2 Respectively, the indirect impact correction coefficient of typhoon, p is the power of the fan, C p Is the coefficient of wind energy conversion efficiency, λ a Is the tip speed ratio of the fan, beta a Pitch angle, r, of fan blades a Is the fan blade radius.
The construction of the degradation and impact correlation model based on the component operating state comprises the following steps:
acquiring historical operation monitoring data of an offshore wind turbine;
extracting the health state characteristics of the offshore wind turbine component under different operating conditions through principal component clustering analysis, and dividing different component states according to the difference value between other states and the health state;
and establishing a historical state transfer process of a random state model description component according to the component state division result and the historical state evolution process under different working conditions and based on the Markov chain state transfer process to obtain a degradation and impact correlation model.
The degradation and impact correlation model is represented as:
R＝X'X/n q
Y＝XU＝[Y 1 ,Y 2 ,…,Y q ]
wherein,
in order to be able to set the operating state of the component k,
for a component degradation and impact correlation function, mu is a state quantity correction coefficient, an input variable matrix X after SCADA data standardization processing related to a certain component of the fan contains q-dimensional initial parameters, and each-dimensional parameter contains n q One sample, X' represents the transpose of X, the eigen equation of the correlation coefficient matrix R has q eigenvalues, λ 1 ≥λ 2 ≥…≥λ q ，U＝(U 1 ,U 2 ,…,U q ) The feature vectors corresponding to the feature values are obtained, Y is a principal component, sorting is carried out according to the sequence of the variance of the principal component from large to small, and Y is q Is the qth principal component, λ ji Selecting omega for the characteristic value corresponding to the ith principal component under the working condition j, wherein delta is the cumulative variance contribution rate, and the principal component analysis result of each type of working condition j is selected according to the condition that the cumulative variance contribution rate is greater than delta j A main component, pi (t) g ) Is a monitoring point t g Probability distribution of each state of the fan component at the moment, A is a state transition matrix, delta t is any time interval, t 0 For long time of state transition, S k A vector matrix of values is quantized for each state.
Suppose failure rate λ of system component k k Failure rate lambda from natural degradation 0k And component cumulative failure rate increase Δ λ due to multiple typhoon impacts ski (t), the service life of the unit component is a non-negative continuous random variable, the natural degradation process is described by Weibull distribution, and a component failure rate increment model and a degradation and impact correlation model under typhoon impact are combined to obtain an offshore wind turbine unit component time-varying failure rate model considering degradation and impact dependence:
wherein i is the typhoon impact frequency; n (t) is the cumulative number of typhoon impacts at the moment t; d i (t) is the impact amplitude at the time t during the ith typhoon;
representing the component k in an operating state as a function of the dependency of component degradation on impact
The time is affected by impact on the increment of the failure rate; t is t si And t ei Respectively representing the first arrival moment and the removal moment of the impact influence of the typhoon on the unit; beta is a shape parameter and eta is a scale parameter.
The model for predicting the residual life of the unit component considering the dependence of the degradation and the impact is as follows:
wherein the component reliability R k (t) indicates that the component uptime is greater than the ttrobit, P indicates the probability of an event,
representing the component failure Rate expectation function considering the impact of random typhoons, n z The total times of typhoon impact scenes are represented, Z is the typhoon impact scene, the single typhoon impact scene is the set of the maximum wind speed radius, the moving speed, the central air pressure difference, the wind speed at the maximum wind speed radius and the moving direction of the typhoon, Z is a typhoon impact scene variable, T is the time when the component fails, and the residual service life of the component is equal to the residual service life of the component when the component operates to T
A remaining life expectancy value, x a remaining life time variable,
indicating the expected value of remaining life
The distribution function of (a) is determined,
in order to obtain a function of expected value of reliability of the component under historical typhoon impact,
representing reliability function under random typhoon impact scenario
Is composed of
Derivative function of t s To obtain probability density function of residual life expectancy value when t =0
The time at which the maximum value corresponds to,
is a reliability failure threshold.
Carrying out natural degradation fault rate model parameter estimation by counting the service life frequency of a certain part of a fan in an offshore wind farm and utilizing a maximum likelihood estimation method, and extracting n on the basis of natural degradation w The machine set carries out parameter estimation on the component fault rate increment model under typhoon impact:
wherein beta is a shape parameter, eta is a scale parameter, f 0k (t) is a Weibull distribution probability density distribution function, L (β, η) isLikelihood function, t z For component life sample values, lnL (β, η) is the log-likelihood function, n c Is the sample volume, U j Is the root mean square error of the set j,
and y j (t gi ) Are respectively set j at monitoring point t gi Estimated value of remaining life at time and actual value of remaining life, n g N is the number of monitoring points w For the number of experimental fan samples drawn,
is n w Mean root mean square error of the unit.
Compared with the prior art, the invention has the following beneficial effects:
(1) The method considers the influence of typhoon random impact on the prediction result, establishes a degradation and impact dependent offshore wind turbine component time-varying fault rate model and a component fault rate increment model, and can accurately describe the evolution law of the actual degradation of the turbine under severe offshore operating environment.
(2) According to the method, the model of the relevance between the degradation and the impact of the component is established, the incidence relation between the random impact and the degradation state is accurately described, and the residual life prediction model is corrected by utilizing the actual monitoring data of the component, so that the component prediction degradation process is more consistent with the actual degradation process, and the life prediction precision is improved.
Drawings
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a schematic diagram of the evolution of the failure rate of offshore wind turbine components under typhoon impact;
FIG. 3 is a schematic diagram of a Batts typhoon wind field model and the influence on an offshore wind field;
FIG. 4 is a schematic diagram of the state division of the components;
FIG. 5 is a diagram illustrating the effect of condition partitioning on principal component cluster analysis results;
FIG. 6 is a schematic diagram of remaining life of a unit component in a random typhoon impact scenario;
FIG. 7 is a flow chart of parameter estimation by particle swarm optimization algorithm;
FIG. 8 shows the results of the evaluation of the current change and the state of the generator;
FIG. 9 is a graph of generator stochastic state quantized expected values versus run time;
FIG. 10 is a probability density distribution of expected values of remaining life of a generator;
FIG. 11 is a process of reliability change for three life prediction methods;
FIG. 12 is a process of reliability change corresponding to the correction process of method 3;
FIG. 13 is a comparison of the predicted residual life results for three methods;
FIG. 14 is a probability density curve of the expected value of the remaining life at different arrival rates of typhoon λ;
FIG. 15 is a graph of the effect of random typhoon impact number on prediction error.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
A method for predicting the residual life of an offshore wind turbine generator considering typhoon influence, as shown in FIG. 1, comprises the following steps:
1) Based on the influence mechanism of the typhoon on the degradation of the unit components, the influence of direct impact and indirect impact on the component failure rate is analyzed, and a component failure rate increment model under the typhoon impact is constructed.
1-1) acquiring typhoon data in a typhoon-prone area;
1-2) constructing a Batts typhoon wind field model; FIG. 3 shows a schematic diagram of a typhoon wind farm model and an influence on an offshore wind farm;
1-3) assuming that the frequency of typhoon impact of an offshore wind turbine generator set in the operation period obeys a Poission process with a parameter of lambda, and a typhoon wind field model is kept unchanged in the processes before and after the typhoon passes through a wind power plant, obtaining each typical typhoon impact probability by adopting a Monte Carlo sampling method, and establishing a typhoon random impact model;
the typhoon random impact model comprises the following steps:
where P (N (t) = N) represents the probability that a typhoon has N impacts within (0, t), λ is the arrival rate of the typhoon, f G (t si ) Time t required for the first arrival of the ith impact si Gamma function value, gamma (i); v Rmax The gradient wind speed at the maximum wind radius, K is an empirical parameter, the value range is 6.93-6.97, delta p is the central pressure difference of the typhoon, R max F is a scientific parameter; v denotes the instantaneous wind speed of the typhoon, V s The typhoon moving speed; r is the distance from any position of the typhoon to the center of the typhoon; x is an empirical parameter and has a value range of 0.5-0.7.
1-4) analyzing the influence of direct impact and indirect impact on the component failure rate based on a typhoon random impact model, and constructing a component failure rate increment model under typhoon impact:
wherein,
as a function of the impact dependence of component degradationIndicating that component k is in an operating state
Time-dependent impact on the increment of the failure rate, t si And t ei Respectively showing the first arrival moment and the removal moment of the impact influence of the typhoon on the unit, wherein alpha and epsilon are direct or indirect impact correction coefficients of the typhoon of the component, and v is the instantaneous wind speed of the typhoon.
And distinguishing the influence of direct impact and indirect impact, wherein the incremental model of the failure rate of the part under the direct impact is expressed as follows:
the incremental model of component failure rate under indirect impact is expressed as:
wherein, Δ λ sk1i Failure rate of external units, alpha, caused by direct impact of typhoon k1 、ε k1 Respectively, the direct impact correction coefficient of typhoon, w is the wind load perpendicular to the surface of the fan, rho is the air density, C s Is the form factor, S α Is the projection area on the plane vertical to the wind direction; delta lambda sk2i Indicating the failure rate, alpha, of the internal components of the unit due to typhoon indirect impact k2 、ε k2 Respectively, the indirect impact correction coefficient of typhoon, p is the power of the fan, C p Is the coefficient of wind energy conversion efficiency, λ a Is the tip speed ratio of the fan, beta a Pitch angle, r, of fan blades a Is the fan blade radius.
In the embodiment, a certain offshore wind farm in China is taken as an example, the offshore wind farm comprises 36 3MW wind turbines, the layout is planned according to 4 rows and 9 rows, the distance between the south and the north of the wind turbines is 0.5km, and the distance between the east and the west of the wind turbines is 1km.
Based on historical typhoon yearbook and statistics of historical typhoon process monitoring data of the wind power plant, the parameter estimation value of the typhoon random impact model is obtained and is shown in table 1.
TABLE 1 typhoon model parameter estimation
2) Aiming at the problem of quantifying the relevance between the degradation process of the component and the impact, a model of the relevance between the degradation of the component operating state and the impact is constructed by combining the monitoring data of the unit operating state, and the capacity of bearing the external impact is quantified.
2-1) obtaining historical operation monitoring data of the offshore wind turbine;
2-2) extracting the health state characteristics of the offshore wind turbine component under different operating conditions through principal component clustering analysis, and dividing different component states according to the difference values between other states and the health states;
2-3) according to the component state division results and the historical state evolution process under different working conditions, establishing the historical state transition process of the random state model description component based on the Markov chain state transition process, and obtaining a degradation and impact correlation model.
Describing component stochastic state models using Markov chain state transition processes
Under the random disturbance of typhoon weather, along with the aggravation of the random failure state of the unit components, the components are more sensitive to external impact, and the impact bearing capacity is gradually reduced. Component degradation and impact correlation function
Is a monotonically increasing function.
The degradation and impact correlation model is represented as:
R＝X'X/n q
Y＝XU＝[Y 1 ,Y 2 ,…,Y q ]
wherein,
in order to be able to set the operating state of the component k,
for a component degradation and impact correlation function, mu is a state quantity correction coefficient, an input variable matrix X after SCADA data standardization processing related to a certain component of the fan contains q-dimensional initial parameters, and each-dimensional parameter contains n q One sample, X' represents the transposition of X, the eigen equation of the correlation coefficient matrix R has q eigenvalues, lambda 1 ≥λ 2 ≥…≥λ q ，U＝(U 1 ,U 2 ,…,U q ) The feature vectors corresponding to the feature values are obtained, Y is a principal component, sorting is carried out according to the sequence of the variance of the principal component from large to small, and Y is q Is the qth principal component, Y 1 Is the first principal component, has the largest variance, and can interpret most of the information of the data, lambda ji Is the ith principal component under the working condition jDividing corresponding characteristic values, wherein delta is the cumulative variance contribution rate, and selecting omega according to the principle component analysis result of each type of working condition j when the cumulative variance contribution rate is larger than delta j A main component, pi (t) g ) Is a monitoring point t g Probability distribution of each state of the fan component at the moment, A is a state transition matrix, delta t is any time interval, t 0 For the state to be transferred long, S k A vector matrix of values is quantized for each state.
In this embodiment, taking a generator with a high failure rate and a serious failure consequence as an example, the states of the generator are divided into four states, i.e., a healthy state, a slightly abnormal state, an abnormal state, and a failure state, as shown in fig. 4, the initial probability distribution is pi (t) g ＝0)＝[1000]。
And dividing the operation conditions into three types under low, medium and high wind speeds according to the initial principal component clustering analysis result by combining the SCADA monitoring parameters related to the generator. And respectively carrying out principal component clustering analysis on the three types of working condition data, selecting principal components according to the accumulated variance contribution rate of more than 90%, wherein clustering results of the health characteristics of the generator under different working conditions after the principal component analysis are shown in FIG. 5.
3) Based on the time-varying randomness characteristic of unit performance degradation caused by typhoon random impact, a time-varying fault rate model of the offshore wind turbine unit component considering dependence of degradation and impact is constructed by combining a component fault rate increment model under typhoon impact and a degradation and impact correlation model.
Suppose failure rate λ of system component k k Failure rate lambda from natural degradation 0k And component cumulative failure rate increase Δ λ due to multiple typhoon impacts ski (t), the service life of the unit component is a non-negative continuous random variable, the natural degradation process of the unit component is described by Weibull distribution, and a component failure rate increment model and a degradation and impact correlation model under typhoon impact are combined to obtain an offshore wind turbine unit component time-varying failure rate model considering degradation and impact dependence:
wherein i is the typhoon impact frequency; n (t) is the cumulative number of typhoon impacts at the moment t; d i (t) is the impact amplitude at the t moment during the ith typhoon;
representing the component k in an operating state as a function of the dependency of the degradation of the component on the impact
The time is affected by impact on the increment of the failure rate; t is t si And t ei Respectively representing the first arrival moment and the removal moment of the impact influence of the typhoon on the unit; beta is a shape parameter and eta is a scale parameter.
The schematic diagram of the fault rate evolution of the offshore wind turbine component under the impact of typhoon is shown in fig. 2.
4) Based on an offshore wind turbine unit component time-varying fault rate model, under a typhoon impact scene, a unit component residual life prediction model is constructed, and parameters of a component fault rate increment model under typhoon impact and a degradation and impact correlation model are optimized by adopting a particle swarm optimization algorithm according to the minimum prediction error, so that an optimized residual life prediction model is obtained.
And according to the historical typhoon distribution probability, adopting a Monte Carlo sampling method to randomly sample and establish N (t) times of random typhoon impact scenes in (0, t), wherein the single typhoon impact scenes are a set of the maximum wind speed radius of the typhoon, the moving speed, the central air pressure difference, the wind speed at the maximum wind speed radius and the moving direction of the typhoon. And according to the historical state monitoring data of the unit, constructing a unit component random state model based on a Markov transfer process to obtain a component historical state transfer process. And combined with any operating time t g Previous actual typhoon weather data and component healthThe state monitoring data corrects typhoon impact scenes and component state transition processes, and therefore a unit component residual life prediction model considering degradation and impact dependence is obtained:
wherein the component reliability R k (t) indicates that the component uptime is greater than the probability of ttp, P indicates the probability of an event,
representing the expected value function of failure rate of a component taking into account the impact of random typhoons, n z The total times of typhoon impact scenes are represented, Z is the typhoon impact scene, Z is a typhoon impact scene variable, T is the time when the component fails, and the residual service life of the component is equal to
The expected value of the remaining life, x is a variable of the remaining life time,
indicating the expected value of remaining life
The distribution function of (a) is determined,
in order to obtain a function of expected value of reliability of the component under historical typhoon impact,
representing reliability function in view of random typhoon impact scenarios
Is composed of
Derivative function of t s To obtain probability density function of residual life expectancy value when t =0
The time instant corresponding to the maximum value is,
is a reliability failure threshold.
Carrying out natural degradation fault rate model parameter estimation by counting the service life frequency of a certain part of a fan in an offshore wind farm and utilizing a maximum likelihood estimation method, and extracting n on the basis of natural degradation w The machine set carries out parameter estimation on the component fault rate increment model under typhoon impact:
wherein beta is a shape parameter, eta is a scale parameter, f 0k (t) is a Weibull distribution probability density distribution function, L (beta, eta) is a likelihood function, t z For component life sample values, lnL (β, η) is the log-likelihood function, n c Is the sample volume, U j Is the root mean square error of the unit j,
and y j (t gi ) Are respectively set j at monitoring point t gi Estimated value of remaining life at time and actual value of remaining life, n g N is the number of monitoring points w For the number of extracted experimental fan samples, U is n w Mean root mean square error of the unit.
A schematic of the remaining life of the crew components in a random typhoon impact scenario is shown in FIG. 6. FIG. 7 is a flow chart of a particle swarm optimization algorithm. In this embodiment, the initial particle number is set to 40, the learning factor is also set to 1.494, and the maximum particle velocity is set to 0.8. Parameter optimization is performed through a particle swarm algorithm, and the parameter estimation result is shown in a table 2:
TABLE 2 residual Life prediction model parameter estimation results
5) Based on the optimized residual life prediction model, the optimized residual life prediction model is corrected and optimized by combining typhoon impact monitoring data and unit real-time running state monitoring data, and residual life prediction is realized.
Fig. 8 shows the evaluation result of the generator state for a period of time before the unit is stopped due to the abnormal generator current. FIG. 9 shows the change of the expected value of the generator random state quantization in different initial states with the running time at any monitoring point according to the Markov state transition process.
Fig. 10 is a distribution diagram of a probability density distribution function of expected values of remaining life of the generator in consideration of the random impact influence of typhoon.
And (3) taking the prediction of the residual life of the generator of a certain unit as an example, and verifying the effectiveness of the model by combining historical monitoring data. Three methods were compared, method 1: a natural degradation prediction model without considering typhoon impact influence; the method 2 comprises the following steps: in a random typhoon impact scene, a prediction model with degradation and typhoon impact dependent is considered; the method 3 comprises the following steps: on the basis of the method 2, a prediction model for correcting the actual typhoon impact is combined. And performing model correction at intervals of 50 days according to the interval time of the historical typhoon and the duration of each state.
FIG. 11 shows that method 1 considers only the natural degradation process of the unit components without considering the typhoon influence, the prediction result is most optimistic, and the reliability and the residual life of the components are overestimated. Compared with the method 1, the method 2 and the method 3 consider the influence of typhoon on the degradation process of the unit components, the reliability evolution process is closer to the reality, and the Root Mean Square Error (RMSE) of the residual life prediction is respectively reduced by 17.4% and 25.1%. Compared with the method 2, the method 3 is combined with the actual typhoon impact correction model, so that the prediction error is further reduced, and is reduced by 7.7% compared with the method 2.
Fig. 12 shows the modification process for the model part in method 3 after the unit has been subjected to the impacts of typhoon "sea anemone" and typhoon "bravay" in sequence during the operation.
Fig. 13 shows that as the monitoring data increases, the prediction result of the method 3 is closer to the actual life value, and the prediction accuracy is improved.
And respectively analyzing the influence of the typhoon arrival rate lambda on the residual life prediction and the influence of the random typhoon impact times on the prediction error. Fig. 14 shows that as the typhoon impact frequency increases, the peak value of the probability density curve moves to the upper left, and the curve shape becomes narrower and the probability distribution becomes more concentrated, which indicates that the increase of the typhoon impact frequency accelerates the speed of the reliability reduction of the unit components, the expected value of the remaining life prediction decreases, the probability distribution becomes more concentrated and the uncertainty of the remaining life prediction decreases. Fig. 15 shows that when the number of typhoon impacts is the same as the actual number of occurrences, the remaining life prediction error RMSE is minimized, and monitoring data correction is more advantageous for reduction of RMSE. When the typhoon impact frequency deviates from the actual typhoon occurrence frequency, the residual life prediction error caused by the typhoon impact frequency deviation is reduced.
The method provided by the invention is effective and feasible, and can provide reference for predicting the residual life of the offshore wind turbine generator considering typhoon influence.
The foregoing detailed description of the preferred embodiments of the invention has been presented. It should be understood that numerous modifications and variations could be devised by those skilled in the art in light of the present teachings without departing from the inventive concepts. Therefore, the technical solutions that can be obtained by a person skilled in the art through logic analysis, reasoning or limited experiments based on the prior art according to the concept of the present invention should be within the scope of protection defined by the claims.
Claims (10)
1. A method for predicting the residual life of an offshore wind turbine generator considering typhoon influence is characterized by comprising the following steps:
analyzing the influence of direct impact and indirect impact on the failure rate of the component based on the influence mechanism of the typhoon on the degradation of the unit component, and constructing a component failure rate increment model under the typhoon impact;
aiming at the quantitative problem of the relevance between the degradation process and the impact of the component, a degradation and impact relevance model of the component operation state is constructed by combining the unit operation state monitoring data, and the capacity of bearing the external impact is quantized;
based on the time-varying randomness characteristic of unit performance degradation caused by typhoon random impact, a time-varying fault rate model of the offshore wind turbine unit component considering dependence of degradation and impact is constructed by combining a component fault rate increment model under typhoon impact and a degradation and impact correlation model;
on the basis of an offshore wind turbine unit component time-varying fault rate model, under a typhoon impact scene, a unit component residual life prediction model is constructed, and a particle swarm optimization algorithm is adopted to optimize parameters of a component fault rate increment model under typhoon impact and a degradation and impact correlation model with the minimum prediction error, so that an optimized residual life prediction model is obtained;
based on the optimized residual life prediction model, the optimized residual life prediction model is corrected and optimized by combining typhoon impact monitoring data and unit real-time running state monitoring data, and residual life prediction is realized.
2. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence according to claim 1, wherein the construction of the incremental model of the failure rate of the components under the typhoon impact comprises the following steps:
acquiring typhoon data in a typhoon multi-occurrence area;
constructing a Batts typhoon wind field model;
supposing that the number of times of typhoon impact of an offshore wind turbine unit in the operation period obeys a Possion process with a parameter of lambda, a typhoon field model is kept unchanged before and after the typhoon passes through a wind power plant, a Monte Carlo sampling method is adopted to obtain each typical typhoon impact probability, and a typhoon random impact model is established;
and analyzing the influence of direct impact and indirect impact on the failure rate of the component based on a typhoon random impact model, and constructing a component failure rate increment model under typhoon impact.
3. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence as recited in claim 2, wherein the typhoon random impact model is as follows:
where P (N (t) = N) represents the probability that a typhoon has N impacts within (0, t), λ is the arrival rate of the typhoon, f G (t si ) Time t required for the first arrival of the ith impact si Gamma function value, gamma (i); v Rmax The gradient wind speed at the maximum wind radius, K is an empirical parameter, the value range is 6.93-6.97, delta p is the central pressure difference of the typhoon, R max F is a scientific parameter; v denotes the typhoon instantaneous wind speed, V s The typhoon moving speed; r is the distance from any position of the typhoon to the center of the typhoon; x is an empirical parameter and has a value range of 0.5-0.7.
4. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence as recited in claim 1 or 2, wherein the incremental model of the failure rate of the component under the typhoon impact is as follows:
wherein,
representing the component k in an operating state as a function of the dependency of the degradation of the component on the impact
Time-dependent impact on the increment of the failure rate, t si And t ei Respectively showing the first arrival moment and the removal moment of the impact influence of the typhoon on the unit, wherein alpha and epsilon are direct or indirect impact correction coefficients of the typhoon of the component, and v is the instantaneous wind speed of the typhoon.
5. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence as claimed in claim 4, wherein the influence of direct impact and indirect impact is distinguished, and the incremental model of the failure rate of the component under the direct impact is represented as follows:
the incremental model of component failure rate under indirect impact is expressed as:
wherein, Δ λ sk1i Failure rate of external parts of the unit, alpha, due to direct typhoon impact k1 、ε k1 Respectively, the direct impact correction coefficient of typhoon, w is the wind load perpendicular to the surface of the fan, rho is the air density, C s Is the form factor, S α Is the projection area on the plane vertical to the wind direction; delta lambda sk2i Indicating the failure rate of the internal components of the unit, alpha, due to the indirect impact of typhoons k2 、ε k2 Respectively, the indirect impact correction coefficient of typhoon, p is the power of the fan, C p Is the coefficient of wind energy conversion efficiency, λ a Is the tip speed ratio of the fan, beta a Is the pitch angle of the fan blade, r a Is the fan blade radius.
6. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence as claimed in claim 1, wherein the construction of the model based on the correlation between the degradation and the impact of the component running state comprises the following steps:
acquiring historical operation monitoring data of an offshore wind turbine;
extracting the health state characteristics of the offshore wind turbine component under different operating conditions through principal component clustering analysis, and dividing different component states according to the difference values between other states and the health states;
and establishing a historical state transfer process of a random state model description component according to the component state division result and the historical state evolution process under different working conditions and based on the Markov chain state transfer process to obtain a degradation and impact correlation model.
7. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence according to claim 6, wherein the model of the correlation between the degradation and the impact is represented as:
R＝X'X/n q
Y＝XU＝[Y 1 ,Y 2 ,…,Y q ]
wherein,
in order to be able to operate the component k,
mu is a state quantity correction coefficient, an input variable matrix X after SCADA data standardization processing related to a certain component of the fan contains q-dimensional initial parameters, and each-dimensional parameter contains n q One sample, X' represents the transposition of X, the eigen equation of the correlation coefficient matrix R has q eigenvalues, lambda 1 ≥λ 2 ≥…≥λ q ，U＝(U 1 ,U 2 ,…,U q ) The feature vectors corresponding to the feature values are obtained, Y is a principal component, sorting is carried out according to the sequence of the variance of the principal component from large to small, and Y is q Is the qth principal component, λ ji Selecting omega for the characteristic value corresponding to the ith principal component under the working condition j, wherein delta is the cumulative variance contribution rate, and the principal component analysis result of each type of working condition j is selected according to the condition that the cumulative variance contribution rate is greater than delta j A main component, pi (t) g ) Is a monitoring point t g Probability distribution of each state of the fan component at the moment, A is a state transition matrix, delta t is any time interval, t 0 For the state to be transferred long, S k A vector matrix of values is quantized for each state.
8. The method for predicting the residual life of an offshore wind turbine generator considering typhoon influence according to claim 1, wherein the failure rate λ of a system component k is assumed k Failure rate lambda from natural degradation 0k And component cumulative failure rate increase Δ λ due to multiple typhoon impacts ski (t) the service life of the unit component is a non-negative continuous random variable, the natural degradation process of the unit component is described by adopting Weibull distribution, and a component fault rate increment model and degradation and impact under typhoon impact are combinedAnd (3) hitting a correlation model to obtain an offshore wind turbine component time-varying fault rate model considering the dependence of degradation and impact:
wherein i is the typhoon impact frequency; n (t) is the cumulative number of typhoon impacts at the moment t; d i (t) is the impact amplitude at the time t during the ith typhoon;
representing the component k in an operating state as a function of the dependency of the degradation of the component on the impact
The time is affected by impact on the increment of the failure rate; t is t si And t ei Respectively representing the first arrival moment and the removal moment of the impact influence of the typhoon on the unit; beta is a shape parameter and eta is a scale parameter.
9. The method for predicting the residual life of the offshore wind turbine generator considering the typhoon influence as recited in claim 8, wherein the model for predicting the residual life of the wind turbine generator components considering the dependence of the degradation and the impact is as follows:
wherein the component reliability R k (t) indicates that the component uptime is greater than the probability of ttp, P indicates the probability of an event,
representing the expected value function of failure rate of a component taking into account the impact of random typhoons, n z The total times of typhoon impact scenes are represented, Z is the typhoon impact scene, the single typhoon impact scene is the set of the maximum wind speed radius, the moving speed, the central air pressure difference, the wind speed at the maximum wind speed radius and the moving direction of the typhoon, Z is a typhoon impact scene variable, T is the time when the component fails, and the residual service life of the component is T when the component operates to T t ＝T-t,
The expected value of the remaining life, x is a variable of the remaining life time,
indicating the expected value of remaining life
The distribution function of (a) is determined,
in order to obtain a function of expected value of reliability of the component under historical typhoon impact,
representing reliability function in view of random typhoon impact scenarios
Is composed of
Derivative function of (a), t s To obtain probability density function of residual life expectancy value when t =0
The time instant corresponding to the maximum value is,
is a reliability failure threshold.
10. The method for predicting the remaining life of the offshore wind turbine generator considering typhoon influence according to claim 1, wherein the method comprises the steps of carrying out natural degradation fault rate model parameter estimation by carrying out statistics on the life frequency of a certain component of a wind turbine in an offshore wind farm and utilizing a maximum likelihood estimation method, and extracting n on the basis of natural degradation w The machine set carries out parameter estimation on the failure rate increment model of the component under typhoon impact:
wherein beta is a shape parameter, eta is a scale parameter, f 0k (t) is a Weibull distribution probability density distribution function, L (beta, eta) is a likelihood function, t z For component life sample values, lnL (β, η) is the log-likelihood function, n c Is the sample volume, U j Is the root mean square error of the set j,
and y j (t gi ) Are respectively set j at monitoring point t gi Estimated value of remaining life at time and actual value of remaining life, n g Is the number of monitoring points, n w For the number of experimental fan samples drawn,
is n w Mean root mean square error of the unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211244944.4A CN115659612A (en) | 2022-10-12 | 2022-10-12 | Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211244944.4A CN115659612A (en) | 2022-10-12 | 2022-10-12 | Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115659612A true CN115659612A (en) | 2023-01-31 |
Family
ID=84986798
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211244944.4A Pending CN115659612A (en) | 2022-10-12 | 2022-10-12 | Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115659612A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117686935A (en) * | 2024-01-30 | 2024-03-12 | 湖北工业大学 | Battery RUL prediction method based on voltage probability density |
| CN118053046A (en) * | 2024-04-11 | 2024-05-17 | 中国海洋大学 | Fusion characterization and judgment method for health state of offshore wind power generation connection structure |
| CN118211082A (en) * | 2024-05-21 | 2024-06-18 | 山东海纳智能装备科技股份有限公司 | Oil filter element residual life prediction method and system based on data analysis |
-
2022
- 2022-10-12 CN CN202211244944.4A patent/CN115659612A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117686935A (en) * | 2024-01-30 | 2024-03-12 | 湖北工业大学 | Battery RUL prediction method based on voltage probability density |
| CN117686935B (en) * | 2024-01-30 | 2024-05-14 | 湖北工业大学 | Battery RUL prediction method based on voltage probability density |
| CN118053046A (en) * | 2024-04-11 | 2024-05-17 | 中国海洋大学 | Fusion characterization and judgment method for health state of offshore wind power generation connection structure |
| CN118211082A (en) * | 2024-05-21 | 2024-06-18 | 山东海纳智能装备科技股份有限公司 | Oil filter element residual life prediction method and system based on data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115659612A (en) | Method for predicting residual life of offshore wind turbine generator with typhoon influence taken into consideration | |
| CN104299044A (en) | Clustering-analysis-based wind power short-term prediction system and prediction method | |
| CN113205210B (en) | Wind power plant wind speed and power prediction method, system and equipment for complex terrain and storage medium | |
| CN111260126A (en) | Short-term photovoltaic power generation prediction method considering correlation degree of weather and meteorological factors | |
| CN113822418A (en) | Wind power plant power prediction method, system, device and storage medium | |
| CN111191832A (en) | Typhoon disaster power distribution network tower fault prediction method and system | |
| CN115977874B (en) | Wind turbine generator yaw self-adaptive calibration method and system based on laser wind-finding radar | |
| CN112832960A (en) | Fan blade icing detection method based on deep learning and storage medium | |
| CN110852492A (en) | Photovoltaic power ultra-short-term prediction method for finding similarity based on Mahalanobis distance | |
| CN116702957A (en) | New energy power prediction method, equipment and storage medium for extreme weather | |
| CN108471137B (en) | Wind power prediction wind speed power probabilistic mapping method | |
| CN105741192B (en) | Short-term wind speed combined forecasting method for wind turbine engine room of wind power plant | |
| CN113610285A (en) | Power prediction method for distributed wind power | |
| CN112052996B (en) | Small hydropower station cluster power prediction method based on convolutional neural network technology | |
| CN110296055B (en) | Wind direction prediction associated seed unit screening method | |
| CN116307139A (en) | Wind power ultra-short-term prediction method for optimizing and improving extreme learning machine | |
| Peng et al. | Wind turbine blades icing failure prognosis based on balanced data and improved entropy | |
| CN103473461B (en) | Based on the wind power prediction error estimation that data characteristics is extracted | |
| Deshmukh et al. | Application of genetic algorithm to neural network model for estimation of wind power potential | |
| CN117893030B (en) | Power system risk early warning method based on big data | |
| CN111794921A (en) | Onshore wind turbine generator blade icing diagnosis method based on migration component analysis | |
| Peng et al. | Short-Term Prediction of Generator Blade Ice Fault Based on Multi-AN | |
| Fang et al. | Research on Short-Term and Medium-Term Power Load Forecasting Based on STL-LightGBM | |
| CN111539577A (en) | Short-term wind power prediction method based on wind speed change rate and Gaussian process regression | |
| CN118035875B (en) | Wind power generation power prediction method and system based on extreme weather influence factors |
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 |