CN107238765A - LED integrated driving power supply reliability analysis methods based on acceleration degradation parameter - Google Patents
LED integrated driving power supply reliability analysis methods based on acceleration degradation parameter Download PDFInfo
- Publication number
- CN107238765A CN107238765A CN201611235962.0A CN201611235962A CN107238765A CN 107238765 A CN107238765 A CN 107238765A CN 201611235962 A CN201611235962 A CN 201611235962A CN 107238765 A CN107238765 A CN 107238765A
- Authority
- CN
- China
- Prior art keywords
- power supply
- driving power
- distribution
- parameters
- reliability
- 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
- 238000006731 degradation reaction Methods 0.000 title claims abstract description 68
- 230000015556 catabolic process Effects 0.000 title claims abstract description 67
- 238000004458 analytical method Methods 0.000 title claims abstract description 21
- 230000001133 acceleration Effects 0.000 title description 7
- 238000000034 method Methods 0.000 claims abstract description 67
- 238000012360 testing method Methods 0.000 claims abstract description 27
- 238000007476 Maximum Likelihood Methods 0.000 claims abstract description 14
- 230000005653 Brownian motion process Effects 0.000 claims abstract description 10
- 230000004907 flux Effects 0.000 claims abstract description 8
- 238000013179 statistical model Methods 0.000 claims abstract description 8
- 238000009826 distribution Methods 0.000 claims description 56
- 230000035882 stress Effects 0.000 claims description 55
- 230000008569 process Effects 0.000 claims description 9
- 238000004422 calculation algorithm Methods 0.000 claims description 5
- 230000007774 longterm Effects 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000010998 test method Methods 0.000 claims description 4
- 238000001276 Kolmogorov–Smirnov test Methods 0.000 claims description 2
- 230000032683 aging Effects 0.000 claims description 2
- 238000013401 experimental design Methods 0.000 claims description 2
- 238000004088 simulation Methods 0.000 claims description 2
- 238000003745 diagnosis Methods 0.000 claims 1
- 238000007689 inspection Methods 0.000 claims 1
- 238000013507 mapping Methods 0.000 claims 1
- 238000012216 screening Methods 0.000 claims 1
- 238000010187 selection method Methods 0.000 claims 1
- 230000007850 degeneration Effects 0.000 abstract 2
- 238000012850 discrimination method Methods 0.000 abstract 2
- 230000001186 cumulative effect Effects 0.000 description 8
- 238000011160 research Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 239000003990 capacitor Substances 0.000 description 2
- 238000012938 design process Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 238000003646 Spearman's rank correlation coefficient Methods 0.000 description 1
- 241001125929 Trisopterus luscus Species 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 238000005562 fading Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000000528 statistical test Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/003—Environmental or reliability tests
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/40—Testing power supplies
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- Environmental & Geological Engineering (AREA)
- Testing Resistance To Weather, Investigating Materials By Mechanical Methods (AREA)
Abstract
The present invention relates to a kind of analysis method for reliability in accelerated degradation test field, and in particular to a kind of to carry out the method based on the fail-safe analysis for accelerating degeneration electrical parameter to LED integrated drivings power supply in the whole lamp life cycles of LED based on luminous flux.Its object is to carry out fail-safe analysis to LED integrated drivings power supply using electrical parameter, to draw influence result of the driving power supply to the whole lamp life-span.For the reliability consideration of driving power supply, also mostly is from design angle.The present invention sets up the statistical model based on Wiener processes, and combination Maximum Likelihood Estimation Method and MCMC methods progress identification of Model Parameters to solve the problems, such as fail-safe analysis of the integrated driving power supply in the whole lamp life cycles of LED based on degeneration electrical parameter is accelerated.Drawn by improving discrimination method, improved discrimination method reduces sample requirement on the premise of preferably identification precision is ensured.Finally under the whole lamp life-span constraints of LED based on luminous flux, reliability index of the LED integrated drivings power supply in whole lamp life cycle is provided.
Description
Technical Field
The invention relates to a reliability analysis method in the field of accelerated degradation tests, in particular to a reliability analysis method of an LED integrated driving power supply based on accelerated performance degradation parameters
Background
The LED is widely applied, and the reliability analysis of the LED is very important in the application and popularization of the LED. The integrated driving power supply is used as a part of an LED whole lamp system, and the reliability of the integrated driving power supply is directly related to the LED performance. Therefore, the research on the reliability of the integrated driving power has become an irrevocable issue in the field.
The integrated driving power supply is a highly integrated system and consists of an integrated sub-module and a few discrete components. At present, most of the existing research is based on reliability analysis of a single component, and does not relate to a specific engineering application background. On one hand, the driving power supply is a whole, and a designer must consider certain fault tolerance in the design process, so that the failure of any component cannot represent the whole failure of the driving power supply. On the other hand, for the driving power supplies serving different objects, the standards for judging the reliability and failure mechanisms are different. Compared with a sudden failure research method, the LED integrated driving power supply is a long-life and high-reliability product, and the prediction of the life and the reliability based on the performance degradation information becomes an effective way. At present, the performance degradation mainly has two ideas, one is to call each sample function of a random process of the performance degradation quantity changing along with time as a degradation track, and prediction is carried out based on the degradation track. The method can describe the degradation track of a single sample accurately, but lacks of macroscopic statistical description of the overall degradation rule of the sample. And secondly, parameters of the distribution obeyed by the performance degradation amount at different moments are regarded as random variables, and prediction is carried out based on the degradation amount distribution. The method can be used for macroscopically describing the statistical law of the degradation of all samples.
In summary, the existing method for analyzing the reliability of the driving power supply lacks the background of practical engineering application, and the selected parameters do not have the problem of integral representativeness, so that the influence result of the LED integrated driving power supply in the whole lamp life cannot be obtained.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides an integrated driving power reliability analysis method based on accelerated performance degradation parameters. And identifying model parameters by using a statistical model based on a Winner process and an average service life model under Bayes distribution and combining a maximum likelihood estimation method and an MCMC method to obtain a reliability index of the LED integrated driving power supply and an influence result on the service life of the whole lamp.
The technical scheme adopted by the invention for solving the technical problem is as follows:
the method for analyzing the reliability of the LED integrated driving power supply based on the accelerated performance degradation parameters comprises the following steps:
determining a stress application mode, a stress application type, stress application intensity, single stress or multiple stresses according to an accelerated life test scheme of the LED integrated driving power supply; adding an LED luminous flux aging test as an auxiliary test;
step two, recording input and output parameters of the test sample, including: output current, output voltage, output power, input current and input voltage; in the auxiliary test, recording the light flux attenuation conditions of samples of the same type and the same batch;
analyzing the change trend of the long-term recorded data by using a Daniel test method based on a Sperman correlation coefficient, and selecting performance degradation parameters by combining the short-term recorded data under the condition of applying each strength single stress;
fourthly, in order to obtain the reliability index of the LED integrated driving power supply, establishing a Wiener process statistical model based on the accelerated degradation electric parameter; selecting reasonable auxiliary models according to specific experimental design, wherein the reasonable auxiliary models comprise an Arrhenius model corresponding to temperature stress and a Bayes life distribution model; giving basic assumption, and then checking the assumption; estimating undetermined coefficients in the model by using a maximum likelihood estimation method and a Markov Chain Monte Carlo (MCMC) method according to experimental data and the established model;
and fifthly, utilizing a Wiener process statistical model based on the accelerated degradation electric parameters to perform reliability analysis, obtaining characteristic indexes of the life cycle of the whole LED lamp, and calculating the failure probability of the LED integrated driving power supply.
The invention has the beneficial effects that:
combining with practical engineering application: the invention combines the market practical application with the LED integrated driving power supply for the first time to analyze the reliability of the driving power supply and the influence on the service life of the whole lamp. And (3) selecting degradation parameters according to actual data and theoretical calculation: the traditional research method selects more capacitors as performance degradation parameters according to market feedback. However, as the design process improves, it is found that reliability analysis using sudden failure of the capacitor is not feasible in the actual test process. For the LED driving power supply adopting the constant current driving mode, the output current is not lost as the optimal parameter no matter the actual experimental record or the theoretical calculation result. Compared with the method of selecting parameters according to simulation results, the method is more persuasive in theory. The improved model parameter identification method comprises the following steps: the maximum likelihood estimation method assumes that the parameters are unknown, but a certain value, and the method is simple and easy to implement. However, in engineering practice, the degradation amount distribution parameter is varied. The MCMC method is generated by considering the randomness and periodicity of the variation of the degradation quantity distribution parameters. However, the MCMC method needs to give expert distribution before parameter prediction, and the accuracy of the expert distribution directly determines the reliability of the final identification result. The method combines two identification methods, takes the result of the maximum likelihood estimation method as the distribution parameter of the degradation amount, and then carries out MCMC estimation. The practical evidence proves that the method not only gives the values of the model parameters, but also gives the parameter distribution, thereby improving the reliability of the result. Establishing a relationship between the model parameters and the stress level: in an accelerated degradation test, the relationship between different parameters in degradation amount distribution and the stress level is different, and the invention combines the practical situation of engineering and establishes the relationship between different parameters (mean value and variance) and the stress level according to the existing Arrhenius acceleration model.
Drawings
FIG. 1 is a workflow block diagram of reliability analysis
FIG. 2 is an iterative trajectory diagram of a Markov chain
FIG. 3 is a graph of nuclear density of estimated parameters
FIG. 4 is a graph of reliability under stress at three temperatures
FIG. 5 is a graph of stress, failure threshold versus Bayes distribution mean life
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
The invention provides an LED integrated driving power reliability analysis method based on accelerated performance degradation parameters, which combines an engineering actual improvement model and a parameter identification method to realize reliability analysis of the LED integrated driving power and influence analysis on the service life of the whole lamp. The method comprises the following specific steps:
1) accelerated performance degradation parameter selection
The selection of the acceleration performance degradation parameters is based on the long-term recorded data of the acceleration test and the short-term recorded data under the single-temperature stress of each intensity, the short-term recorded data is used for obtaining the relation between the intensity of the applied stress and the plate-level output parameters and verifying the feasibility of the applied temperature stress, the long-term data is used for giving the influence degree of different types of plate-level output parameters under the same temperature stress, the Daniel test method based on the Sperman correlation coefficient is used for quantitatively giving the change trend of three groups of electrical parameters, and for the significant level α, the time sequence X is used for obtaining the time sequence XtCalculating (t, R)t) Spearman rank correlation coefficient q of 1,2, L, nsIf | T | > Tα/2(n-2), then H is rejected0The sequence was considered to be non-stationary. And when q issWhen > 0, the sequence is consideredThere is an increasing trend; q. q.ssWhen < 0, the sequence is considered to have a downward trend, | qsThe greater the | value, the more pronounced the trend. When | T | ≦ Tα/2When (n-2), H is accepted0Can be regarded as XtIs a stationary sequence.
TABLE 1 examination of the parameters
| Output current (Iout) | Output voltage (Vout) | Output power (Pout) | |
| T is the total value of T | -18.4096 | -14.3079 | -10.9165 |
| Upper a/2 quantile value | 1.9833 | 1.9833 | 1.9833 |
| qs | -0.8757 | -0.8156 | -0.7324 |
2) Selection and verification of distributions
The method is based on the reliability analysis of performance degradation parameters, most samples do not completely fail in a truncation experiment, and the degradation process of a single sample shows certain randomness. Therefore, a current degradation increment normal distribution statistical model is established through research on the degradation increment of the output current of the driving power supply.
Accelerating stress levels S for a set of temperatures1,S2,...,SKLet the jth measurement of the ith sample be time tkijThe amount of degradation is xkij(K1, 1.. K, K; i 1, 1.. m; j 1.. n), let Δ dkij＝xkij-xkij-1,Δtkij＝tkij-tkij-1. First for Δ dkijNormal distribution verification is carried out, a single-sample K-S detection method is utilized, and the detection result is shown in table 2.
Results of K-S test of current degradation increment at 2120 DEG C
a. The distribution was checked as normal.
b. And calculating according to the data.
3) Model building
And on the premise that the failure data meet normal distribution, performing statistical modeling on all sample data, and establishing a model based on a Wiener process from the whole sample. (the distribution of wiener process increments is only time difference dependent, so it is a homogeneous independent incremental process, it is also a normal process
The performance degradation parameter of the sample is gradually degraded under the action of constant stress, and the degradation increment of the sample in each extremely short time interval is a random variable, and the value is influenced by the sample manufacturing process. Assume that the value is yp(P ═ 1, 2.., P), and let ypIs independently distributed with mean value of μyVariance ofThe variance is generally considered independent of stress, and the mean is dependent on stress. The cumulative amount of degradation of the output current is considered herein to be a linear superposition of the increments of degradation per unit time. Namely, the damage model function is 1, and the cumulative damage function is a unit proportional function.
According to the central limit theorem, when n is large, Δ dkijThe distribution of (a) converges to a normal distribution (which is consistent with the actual test result), it is obvious that the degradation increment Δ d at the time tkijThe mean function and variance function of (t) are respectively:
μd(t)＝E[Δdkij(t)]＝μygτ(t) (2)
another α ═ muy，α is temperature stress dependent (the magnitude of the degradation increase μ caused per unit time for different stress levels under ideal single stress conditions) when not accounting for other factorsyInfluenced by stress S), β2It is a stress independent parameter. When temperature is taken as the acceleration stress, an Arrhenius acceleration model is used (the accelerated degradation equation describes the relationship between the amount of degradation and the stress level):
α＝Aexp[B/(273+T)](4)
as can be seen from the results of the normal distribution test, the increment of degradation follows a normal distribution, i.e.
Where C (g) is the cumulative damage function. When the failure threshold is ω, the cumulative damage function c (x) is x, and the initial cumulative degradation amount is 0, the cumulative failure probability function is:
when the performance degradation parameter is measured at equal intervals, τ (t) ═ t is a constant. Substituting the formula (4) into the formula (6).
For Bayes distributed mean life, when it is assumed that the sampling periods of different test samples are fixed and all at Δ t at different stress levels when product degradation occurs in a certain random process regardless of the product, at a certain stress level SKAmount of degradation in j-th measurement cycle of i-th sampleThen stress level SKThe Bayes distribution function below is:
wherein,is the failure threshold.
4) Model parameter identification
Commonly used parameter estimation methods include a maximum likelihood estimation method, a maximum posterior estimation method and a Bayesian estimation method. The maximum likelihood estimation method is the simplest form, and the maximum likelihood estimation method assumes that parameters are unknown but determines values, and only needs to find the parameter with the maximum sample log likelihood distribution.
For a set of acceleration stress levels S1,S2,...,SKLet the jth measurement of the ith sample be time tkijThe amount of degradation is xkij(K1, 1.. K, K; i 1, 1.. m; j 1.. n), let Δ dkij＝xkij-xkij-1,Δtkij＝tkij-tkij-1The independent incremental property according to the Wiener process is Δ dkij:N(αkΔtkij,β2Δtkij)。
Taking the natural logarithm of both sides of the likelihood function shown in the formula (8) to obtain a formula (9), obtaining a formula (10) by the maximum likelihood estimation principle, obtaining A, B and β by solving the equation system according to experimental data, and obtaining the failure distribution of the product degradation with time and stress level as variables
The above is the failure distribution in the accelerated life test derived based on the Wiener process of product degradation.
5) Improvements in models
In the above model derivation process, it is considered that only the mean value of the performance degradation increment is limited by the stress, and the dispersion of the performance degradation increment is independent of the stress. The traditional model established according to the thought can only be suitable for the condition that the stress changes in a small range, and the accelerated life test finally aims to deduce the product reliability condition at normal temperature according to the degradation data under the accelerated stress, so that a model suitable for the condition that the stress changes in a large range needs to be established.
Since the stress of the invention is temperature, a model of the mean and variance of the performance degradation increment is established according to an Arrhenius model. As follows
And solving the undetermined coefficient of the formula according to the existing data by using a least square method.
Substituting equation (15) into equation (13) yields equation (16), which is a functional relationship between B-S distribution mean life and failure threshold and temperature stress.
Substituting equation (15) into equation (17) yields a cumulative probability of failure function associated with temperature stress and failure threshold as equation (18) under a Bayes distribution.
6) Improved model parameter identification method
On the premise of considering the randomness and periodicity of the parameter change of the degradation quantity distribution, the method improves the existing parameter identification method. Firstly, giving out MCMC distribution before test by a rough maximum likelihood estimation method based on a small subsample, then carrying out parameter estimation by using the characteristic that the MCMC method is suitable for the small subsample, and finally giving out the distribution condition of the parameters to be estimated.
Typical methods for MCMC are the Metropolis-Hastings algorithm, which is a special case of the former, and the Gibbs algorithm, which replaces the distribution in the sample with a fully conditional distribution and accepts the proposed value x with a probability of 1*. Assuming x is an n-dimensional vector, the Gibbs sampling algorithm is described as follows:
initialization: selecting an initial parameter vector x0
Iteration: in the kth iteration, do:
(1) selecting index p from {1,2, …, n } completely randomly;
(2) from conditional distributionExtracting a random variable yp；
(3) Definition of
Based on the experimental data of the first 40 small samples under the temperature stress of 120 ℃, the improved parameter identification method is applied to carry out model parameter identification, and the result is as follows. The prior distribution of the parameters given by the invention is as follows: a to N (0, 0.001)2)、B～N(0,0.0012) And τ to G (0.001),0.012) And 3 Markov chains are constructed simultaneously, 11000 iterations are carried out by utilizing Gibbs sampling, and the first 1000 iterations are taken as a combustion period.
TABLE 3 parameter identification results
| node | mean | sd | MCerror | 2.50％ | 97.50％ |
| mu | 0.2009 | 0.463 | 0.006274 | -0.6942 | 1.164 |
| sigma | 3.302 | 0.3691 | 0.004147 | 2.675 | 4.117 |
The iteration tracks of the three Markov chains are shown in figure 2, and the three chains are almost overlapped and can be approximately judged to reach a steady state. The kernel density of the estimated parameters satisfies a unimodal distribution as shown in fig. 3.
7) Obtaining the service life index of the whole lamp and calculating the failure probability of the integrated driving power supply
The reliability curves at 80 ℃, 100 ℃ and 120 ℃ temperature stress are finally obtained according to the above method and are shown in fig. 4.
The accelerated life of the individual sample is calculated according to the fading coefficient calculated according to the luminous flux attenuation data recorded in the auxiliary experiment on the premise of defining the failure criterion as the luminous flux attenuation to 70 percent of the initial value. Then constructing a Weibull life distribution model, and carrying out distribution statistical test through model conversion. Finally, the model parameters are solved, and the characteristic service life is given. Since the invention is not focused here, a detailed calculation process is not given, and only the final whole lamp life cycle index is given. The integrated driving power supply is independently placed in a normal-temperature test environment, the whole lamp is placed in a test environment at the temperature of 80 ℃, and finally the calculated characteristic service life is 6556 hours. For convenience of calculation, the index of the life characteristic of the whole lamp is taken as 6396 h-24 h × 17 × 17.
Table 4 reliability of the driving power supply over the entire lamp life cycle and the average life of the product under the Bayes distribution.
| Temperature stress (. degree.C.) | 80 | 100 | 120 |
| Mean time to failure greater than 6396h probability | 99.96％ | 99.03％ | 90.54％ |
| Mean time to failure greater than 9600h probability | 99.56％ | 95.97％ | 80.94％ |
| Average life (h) of product under B-S distribution | 46858 | 33843 | 25680 |
As can be seen from the table data, the cumulative failure probability of the driving power supply in the whole lamp life cycle is extremely low, and the reliability is extremely high.
By establishing the relationship between the model parameters and the stress levels, it can be finally obtained that the relationship between the stress, the failure threshold value and the Bayes distribution mean life is as shown in fig. 5.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (7)
1. The method for analyzing the reliability of the LED integrated driving power supply based on the accelerated performance degradation parameters is characterized by comprising the following steps of:
determining a stress application mode, a stress application type, stress application intensity, single stress or multiple stresses according to an accelerated life test scheme of the LED integrated driving power supply; adding an LED luminous flux aging test as an auxiliary test;
step two, recording input and output parameters of the test sample, including: output current, output voltage, output power, input current and input voltage; in the auxiliary test, recording the light flux attenuation conditions of samples of the same type and the same batch;
analyzing the change trend of the long-term recorded data by using a Daniel test method based on a Sperman correlation coefficient, and selecting performance degradation parameters by combining the short-term recorded data under the condition of applying each strength single stress;
fourthly, in order to obtain the reliability index of the LED integrated driving power supply, establishing a Wiener process statistical model based on the accelerated degradation electric parameter; selecting reasonable auxiliary models according to specific experimental design, wherein the reasonable auxiliary models comprise an Arrhenius model corresponding to temperature stress and a Bayes life distribution model; giving basic assumption, and then checking the assumption; estimating undetermined coefficients in the model by using a maximum likelihood estimation method and a Markov Chain Monte Carlo (MCMC) method according to experimental data and the established model;
and fifthly, utilizing a Wiener process statistical model based on the accelerated degradation electric parameters to perform reliability analysis, obtaining characteristic indexes of the life cycle of the whole LED lamp, and calculating the failure probability of the LED integrated driving power supply.
2. The method for analyzing the reliability of the LED integrated driving power supply based on the accelerated performance degradation parameter as claimed in claim 1, wherein the algorithm for modeling and analyzing the reliability of the LED integrated driving power supply based on the accelerated degradation parameter in the fourth step comprises the following specific steps:
step A, selecting a distribution model and carrying out distribution inspection according to distribution prediction;
b, establishing a statistical model based on a Wiener process;
step C, identifying model parameters by using a maximum likelihood estimation method;
step D, improving an identification method, and combining a maximum likelihood estimation method and an MCMC method;
and E, improving the model, and establishing the relation between the distribution parameters and the stress.
3. The method for analyzing the reliability of the LED integrated driving power supply according to claim 1, wherein in the second step, the current working status of the entire LED integrated driving power supply is not represented by recording the performance parameters of a discrete component, but the board-level input/output parameters directly related to the entire lamp are classified into the category of the screening parameters.
4. The method for analyzing the reliability of the LED integrated driving power supply based on the accelerated performance degradation parameters of claim 1, wherein the parameters in the step three are selected from two angles of short-term multiple-intensity stress and long-term single-intensity stress, and the general trend of actual test data is verified by using a Daniel test method based on a Sperman correlation coefficient; for time series sample X1,X2,…,XnRecord XtIs Rt＝R(Xt) Consider the variable pair (t, R)t) Sperman correlation coefficient for t ═ 1,2, …, nConstruct statisticsHypothesis testing is performed. The data selection method has theoretical support and is more convincing than a simulation method.
5. The method for analyzing the reliability of the LED integrated driving power supply based on the accelerated performance degradation parameter as claimed in claim 2, wherein in the step B distribution test, a single-sample Kolmogorov-Smirnov test is used to observe Dmax＝max|Sn(x)-F0(x) L, wherein Sn(x) As a function of empirical distribution, F0(x) Determining a critical value D according to a given significance level α and the number n of sample dataαThen by comparison of DmaxAnd DαThe size determines whether to reject the original hypothesis.
6. The base of claim 2The method for analyzing the reliability of the LED integrated driving power supply for accelerating performance degradation parameters is characterized in that in the step D, model parameter identification is carried out by combining an MCMC method with a maximum likelihood estimation method from actual requirements, and a Markov chain is a random process { theta }(1),θ(2),…,θ(T)H, and f (θ)t+1|θ(t),…,θ(1))＝f(θt+1|θ(t)) (ii) a θ when t → ∞ time(t)Converges to an equilibrium distribution that is independent of the initial chain value θ(0)(ii) a The constructed Markov chain has f (theta)t+1|θ(t)) Easy to generate, and its equilibrium distribution is the posterior distribution of the distribution of interest f (θ | y); first, an initial value theta is selected(0)When the sample volume T is given when the balanced distribution is achieved, the convergence of the algorithm is monitored by using convergence diagnosis; if convergence, eliminating the first B observed values and dividing the { theta(B+1),θ(B+2),…,θ(T)Taking the sample as a posterior analysis sample, and generating more observed values if the sample is not the posterior analysis sample; then posterior distribution mapping is carried out, and single edge distribution is focused; finally, all parameters of posterior distribution including Monte Carlo error are obtained.
7. The method for analyzing the reliability of the LED integrated driving power supply based on the accelerated performance degradation parameters of claim 2, wherein the step E is used for establishing a model of the mean and variance of the performance degradation increment by improving the distribution model in the step A according to the test data and the Arrhenius model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611235962.0A CN107238765A (en) | 2016-12-28 | 2016-12-28 | LED integrated driving power supply reliability analysis methods based on acceleration degradation parameter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611235962.0A CN107238765A (en) | 2016-12-28 | 2016-12-28 | LED integrated driving power supply reliability analysis methods based on acceleration degradation parameter |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107238765A true CN107238765A (en) | 2017-10-10 |
Family
ID=59982972
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611235962.0A Pending CN107238765A (en) | 2016-12-28 | 2016-12-28 | LED integrated driving power supply reliability analysis methods based on acceleration degradation parameter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107238765A (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107506337A (en) * | 2017-10-12 | 2017-12-22 | 中国人民解放军海军航空工程学院 | Reliability statistics estimating method based on polynary acceleration degraded data |
| CN107967406A (en) * | 2017-12-18 | 2018-04-27 | 广东科鉴检测工程技术有限公司 | Medical instrument control panel accelerated test analysis method for reliability and system |
| CN108107375A (en) * | 2017-12-20 | 2018-06-01 | 中国电子产品可靠性与环境试验研究所 | Obtain the method and device of LED drive power life information |
| CN108135052A (en) * | 2017-12-20 | 2018-06-08 | 中国电子产品可靠性与环境试验研究所 | The status information monitoring method of LED lamp, apparatus and system |
| CN108446475A (en) * | 2018-03-14 | 2018-08-24 | 哈尔滨理工大学 | Based on the electromagnetic relay reliability estimation method for accelerating to degenerate |
| CN108536918A (en) * | 2018-03-16 | 2018-09-14 | 北京航天控制仪器研究所 | A kind of the determination method, apparatus and electronic equipment of resin adhesive storage life |
| CN109683040A (en) * | 2018-12-25 | 2019-04-26 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Reliability checking method, device and the equipment of flexible direct current transmission converter valve |
| CN110210117A (en) * | 2019-05-31 | 2019-09-06 | 西安工程大学 | A kind of prediction technique and system of spinning frame residue service life |
| CN110334389A (en) * | 2019-05-22 | 2019-10-15 | 航天科工防御技术研究试验中心 | The method of inspection and device of accelerated life test and outfield result consistency |
| CN110737987A (en) * | 2019-10-16 | 2020-01-31 | 北京航空航天大学 | Method for evaluating expected life of LED lighting products |
| CN110851980A (en) * | 2019-11-11 | 2020-02-28 | 中国人民解放军火箭军工程大学 | Method and system for predicting residual life of equipment |
| CN113376549A (en) * | 2021-05-26 | 2021-09-10 | 国网上海能源互联网研究院有限公司 | Pilot protection method and system for flexible direct current power distribution network |
| CN113435057A (en) * | 2021-07-12 | 2021-09-24 | 西安工程大学 | Reliability evaluation method for spinning machine performance degradation |
| CN113703424A (en) * | 2021-08-27 | 2021-11-26 | 哈尔滨市科佳通用机电股份有限公司 | Mean Time Between Failures (MTBF) test evaluation method for servo driver |
| CN114117759A (en) * | 2021-11-12 | 2022-03-01 | 大连海事大学 | Large ship shafting inherent frequency uncertainty analysis method based on nonparametric model |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101710368A (en) * | 2009-12-21 | 2010-05-19 | 北京航空航天大学 | Bayesian reliability comprehensive estimation method based on multisource degraded data |
| CN105468866A (en) * | 2015-12-15 | 2016-04-06 | 长春工业大学 | Method for predicting remaining life of LED driving power of railway vehicles |
| CN105468907A (en) * | 2015-11-23 | 2016-04-06 | 北京航空航天大学 | Accelerated degradation data validity testing and model selection method |
-
2016
- 2016-12-28 CN CN201611235962.0A patent/CN107238765A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101710368A (en) * | 2009-12-21 | 2010-05-19 | 北京航空航天大学 | Bayesian reliability comprehensive estimation method based on multisource degraded data |
| CN105468907A (en) * | 2015-11-23 | 2016-04-06 | 北京航空航天大学 | Accelerated degradation data validity testing and model selection method |
| CN105468866A (en) * | 2015-12-15 | 2016-04-06 | 长春工业大学 | Method for predicting remaining life of LED driving power of railway vehicles |
Non-Patent Citations (1)
| Title |
|---|
| 栗阳: ""基于加速试验的LED驱动电源寿命预测及对整灯寿命影响分析"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107506337A (en) * | 2017-10-12 | 2017-12-22 | 中国人民解放军海军航空工程学院 | Reliability statistics estimating method based on polynary acceleration degraded data |
| CN107967406A (en) * | 2017-12-18 | 2018-04-27 | 广东科鉴检测工程技术有限公司 | Medical instrument control panel accelerated test analysis method for reliability and system |
| CN108107375A (en) * | 2017-12-20 | 2018-06-01 | 中国电子产品可靠性与环境试验研究所 | Obtain the method and device of LED drive power life information |
| CN108135052A (en) * | 2017-12-20 | 2018-06-08 | 中国电子产品可靠性与环境试验研究所 | The status information monitoring method of LED lamp, apparatus and system |
| CN108446475A (en) * | 2018-03-14 | 2018-08-24 | 哈尔滨理工大学 | Based on the electromagnetic relay reliability estimation method for accelerating to degenerate |
| CN108536918A (en) * | 2018-03-16 | 2018-09-14 | 北京航天控制仪器研究所 | A kind of the determination method, apparatus and electronic equipment of resin adhesive storage life |
| CN108536918B (en) * | 2018-03-16 | 2022-04-08 | 北京航天控制仪器研究所 | Method and device for determining storage life of resin type adhesive, and electronic device |
| CN109683040A (en) * | 2018-12-25 | 2019-04-26 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Reliability checking method, device and the equipment of flexible direct current transmission converter valve |
| CN109683040B (en) * | 2018-12-25 | 2021-10-15 | 中国电子产品可靠性与环境试验研究所（（工业和信息化部电子第五研究所）（中国赛宝实验室）） | Method, device and equipment for detecting reliability of flexible direct-current transmission converter valve |
| CN110334389A (en) * | 2019-05-22 | 2019-10-15 | 航天科工防御技术研究试验中心 | The method of inspection and device of accelerated life test and outfield result consistency |
| CN110210117A (en) * | 2019-05-31 | 2019-09-06 | 西安工程大学 | A kind of prediction technique and system of spinning frame residue service life |
| CN110210117B (en) * | 2019-05-31 | 2023-03-31 | 西安工程大学 | Method and system for predicting remaining operation life of spinning frame |
| CN110737987A (en) * | 2019-10-16 | 2020-01-31 | 北京航空航天大学 | Method for evaluating expected life of LED lighting products |
| CN110851980A (en) * | 2019-11-11 | 2020-02-28 | 中国人民解放军火箭军工程大学 | Method and system for predicting residual life of equipment |
| CN113376549A (en) * | 2021-05-26 | 2021-09-10 | 国网上海能源互联网研究院有限公司 | Pilot protection method and system for flexible direct current power distribution network |
| CN113376549B (en) * | 2021-05-26 | 2024-04-19 | 国网上海能源互联网研究院有限公司 | Pilot protection method and pilot protection system for flexible direct-current power distribution network |
| CN113435057A (en) * | 2021-07-12 | 2021-09-24 | 西安工程大学 | Reliability evaluation method for spinning machine performance degradation |
| CN113435057B (en) * | 2021-07-12 | 2023-01-13 | 西安工程大学 | Reliability evaluation method for spinning machine performance degradation |
| CN113703424A (en) * | 2021-08-27 | 2021-11-26 | 哈尔滨市科佳通用机电股份有限公司 | Mean Time Between Failures (MTBF) test evaluation method for servo driver |
| CN114117759A (en) * | 2021-11-12 | 2022-03-01 | 大连海事大学 | Large ship shafting inherent frequency uncertainty analysis method based on nonparametric model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107238765A (en) | LED integrated driving power supply reliability analysis methods based on acceleration degradation parameter | |
| CN107436963B (en) | O-shaped rubber sealing ring service life prediction method based on Copula function multi-degradation | |
| Zhang et al. | A Wiener process model with dynamic covariate for degradation modeling and remaining useful life prediction | |
| CN112926144B (en) | Multi-stress accelerated life test coupling effect analysis and life prediction method | |
| CN108037460B (en) | Real-time capacity evaluation method for lithium ion batteries produced in batches | |
| CN111859658B (en) | Product storage life and reliability assessment method | |
| JP2021521646A (en) | Integrated circuit profiling and anomaly detection | |
| Zhou et al. | Exponential dispersion process for degradation analysis | |
| CN107885928B (en) | Stepping stress acceleration performance degradation reliability analysis method considering measurement error | |
| CN102252898A (en) | Method for testing accelerated life of electronic product based on life-stress model | |
| CN108804806B (en) | Simplified MLE (Multi-level engineering) method for parameters in Weibull distributed comprehensive stress constant addition test | |
| CN111612142B (en) | BP neural network-based high-power LED life prediction method | |
| CN107730127B (en) | Relay storage degradation data prediction method based on output characteristic initial distribution | |
| CN109918776B (en) | Fatigue crack prediction method of mechanical product based on two-step least square method | |
| Kong et al. | A remote estimation method of smart meter errors based on neural network filter and generalized damping recursive least square | |
| Srivastava et al. | Optimum multi-objective ramp-stress accelerated life test with stress upper bound for Burr type-XII distribution | |
| Guan et al. | Objective Bayesian analysis for competing risks model with Wiener degradation phenomena and catastrophic failures | |
| CN116070302A (en) | Cable insulation state prediction method and device | |
| CN109214112B (en) | Reliability simulation analysis method and system of fuel cell system | |
| CN112182883B (en) | Photovoltaic module parameter estimation method based on self-adaptive LM | |
| CN113609456A (en) | Reliability analysis method, system, medium and terminal for photovoltaic charging station of electric vehicle | |
| CN109766518A (en) | Consider the uncertain acceleration degeneration modeling and analysis method of individual of sample difference | |
| CN112782603B (en) | Lithium ion battery cycle life distribution fitting method based on interval truncation data | |
| Al Haddad et al. | Parametric degradation model of OLED using Design of Experiments (DoE) | |
| CN111061708A (en) | Electric energy prediction and restoration method based on LSTM neural network |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20171010 |