CN111832226B - IGBT residual life estimation method based on auxiliary particle filtering - Google Patents
IGBT residual life estimation method based on auxiliary particle filtering Download PDFInfo
- Publication number
- CN111832226B CN111832226B CN202010666870.8A CN202010666870A CN111832226B CN 111832226 B CN111832226 B CN 111832226B CN 202010666870 A CN202010666870 A CN 202010666870A CN 111832226 B CN111832226 B CN 111832226B
- Authority
- CN
- China
- Prior art keywords
- transistor
- performance
- estimation
- residual life
- act
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 239000002245 particle Substances 0.000 title claims description 53
- 238000001914 filtration Methods 0.000 title claims description 21
- 230000015556 catabolic process Effects 0.000 claims abstract description 41
- 238000006731 degradation reaction Methods 0.000 claims abstract description 40
- 238000011156 evaluation Methods 0.000 claims abstract description 19
- 230000007246 mechanism Effects 0.000 claims abstract description 15
- 238000005259 measurement Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 13
- 230000007704 transition Effects 0.000 claims description 8
- 238000012952 Resampling Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 abstract description 9
- 238000012360 testing method Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 6
- 239000002243 precursor Substances 0.000 description 5
- 238000005070 sampling Methods 0.000 description 5
- 238000005315 distribution function Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 239000000523 sample Substances 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 229910000679 solder Inorganic materials 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000001351 cycling effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- 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
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Testing Of Individual Semiconductor Devices (AREA)
Abstract
The disclosure provides a residual life estimation method and device. Acquiring statistical data of the residual life of the switch; determining a performance index based on the statistical data in combination with the performance of the transistor, wherein the performance index is a degradation mechanism and an index directly related to degradation; establishing an estimation system of the residual life based on the performance index; and in addition, the residual life of the switch is evaluated on the premise that the estimation system is combined with the operating environment of the transistor, and the evaluation is carried out according to the evaluation result, so that a more accurate residual life curve of the switch can be determined, and a maintenance decision scheme can be scientifically and reasonably formulated.
Description
Technical Field
The present invention relates to estimation models in intelligent decision making, and in particular, to a method and apparatus for estimating remaining life.
Background
Insulated Gate Bipolar Transistors (IGBTs) are widely used in various power electronic systems, and the remaining life of the switches of the Insulated Gate Bipolar Transistors (IGBTs) is called RUL, from which the health of the IGBTs can be evaluated to avoid unexpected faults of the converter. However, due to the presence of random noise, the estimation of RUL typically has a large variance, resulting in a large error rate of the IGBT's estimation.
Disclosure of Invention
The disclosure aims to provide a residual life estimation method and device, so as to reduce the error rate of the estimation of the residual life.
According to an aspect of the present disclosure, there is provided a remaining life estimation method including: acquiring statistical data of the residual life of the switch; determining a performance index based on the statistical data in combination with the performance of the transistor, wherein the performance index is a degradation mechanism and an index directly related to degradation; establishing an estimation system of the residual life based on the performance index; the remaining lifetime of the switch is evaluated on the premise that the evaluation system combines with the operating environment of the transistor.
In one embodiment, after determining the performance index based on the statistical data and the performance of the transistor, the method further comprises: an estimate of the remaining lifetime of the particle filter based switch is established from the estimate of the auxiliary particle filter.
In one embodiment, the determining a performance index based on the statistics in combination with the performance of the transistor, the performance index being a degradation mechanism and an index directly related to degradation includes: determining the stress of the transistor in fault according to the operation performance of the transistor and the statistical data; judging a fault condition of the transistor based on the stress of the transistor at the time of fault; and determining corresponding early warning parameters according to the fault condition of the transistor.
In one embodiment, after determining the corresponding early warning parameter according to the fault condition of the transistor, the method further includes: acquiring corresponding early warning parameters; and forming a track according to the corresponding early warning parameters so as to predict the faults of the transistors.
In one embodiment, the forming a track according to the corresponding pre-warning parameter to predict the fault of the transistor further includes: introducing a weight value into the corresponding early warning parameter; resampling is employed to adjust the weight values.
According to an aspect of the present disclosure, there is provided an apparatus of a remaining life estimation method, including: the acquisition module is used for acquiring statistical data of the residual life of the switch; a combination module, configured to combine the performance of the transistor based on the statistical data, and determine a performance index, where the performance index is an index directly related to degradation and a degradation mechanism; the establishing module is used for establishing an estimation system of the residual life based on the performance index; and the evaluation module is used for prolonging the residual service life of the switch on the premise that the evaluation system is combined with the operation environment of the transistor.
According to an aspect of the present disclosure, there is provided a computer readable program medium storing computer program instructions which, when executed by a computer, cause the computer to perform a method according to the above.
According to an aspect of the present disclosure, there is provided an electronic apparatus including: a processor; and a memory having stored thereon computer readable instructions which, when executed by the processor, implement the method described above.
The technical scheme provided by the embodiment of the invention can comprise the following beneficial effects:
in some embodiments of the present invention, statistical data of remaining lifetime of a switch is obtained; determining a performance index based on the statistical data in combination with the performance of the transistor, wherein the performance index is a degradation mechanism and an index directly related to degradation; establishing an estimation system of the residual life based on the performance index; and in addition, the residual life of the switch is evaluated on the premise that the estimation system is combined with the operating environment of the transistor, and the evaluation is carried out according to the evaluation result, so that a more accurate residual life curve of the switch can be determined, and a maintenance decision scheme can be scientifically and reasonably formulated.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
Fig. 1 is a flowchart illustrating a remaining life estimation method according to an exemplary embodiment.
Fig. 2 is a detailed flow chart of the determination of performance metrics based on the statistics in combination with the performance of the transistors, according to an exemplary embodiment.
Fig. 3 is a flowchart illustrating a method of estimating remaining lifetime of a transistor based on auxiliary particle filtering, according to an exemplary embodiment.
Fig. 4 is a schematic diagram of an experimental apparatus showing a method for estimating a remaining lifetime of a transistor based on auxiliary particle filtering according to an exemplary embodiment.
Fig. 5 is a temperature cycling test of stress experienced by a transistor in an industrial application, according to an example embodiment.
Fig. 6 is SIRPF traces of an APF filter and VCE, ON for a transistor shown according to an example embodiment.
Figure 7 is a schematic diagram of APF simulated trajectories for different numbers of particles, according to an example embodiment.
Fig. 8 is a block diagram illustrating an apparatus of a remaining life estimation method according to an exemplary embodiment.
Fig. 9 is a hardware diagram of an electronic device, according to an example embodiment.
Fig. 10 is a computer-readable storage medium illustrating a remaining life estimation method according to an exemplary embodiment.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the invention. Rather, they are merely examples of apparatus and methods consistent with aspects of the invention as detailed in the accompanying claims.
Furthermore, the drawings are merely schematic illustrations of the present disclosure and are not necessarily drawn to scale. Some of the block diagrams shown in the figures are functional entities and do not necessarily correspond to physically or logically separate entities.
Insulated Gate Bipolar Transistors (IGBTs) are widely used in various power electronic systems. Since IGBTs are one of the weaker elements in power electronic converters, the residual life (RUL) of the switch is estimated to be a hot spot of research in recent years. Such RUL estimation helps to schedule maintenance according to the health of the IGBTs to avoid unexpected failure of the converter. However, due to the presence of random noise, RUL estimates typically have a large variance, leading to erroneous results for IGBT prognosis. In order to reduce such variance in the RUL estimation, a particle filtering method including sequence importance sampling and sequence importance resampling has recently been adopted. However, these methods result in non-negligible estimated variances due to degradation and depletion of the samples.
In the related art, due to the presence of random noise, the estimation of RUL generally has a large variance, resulting in a large error rate of the estimation of IGBT.
According to an embodiment of the present disclosure, there is provided a remaining life estimating method, as shown in fig. 1 and 9, including:
step S110, acquiring statistical data of the residual life of the switch;
Step S120, determining performance indexes based on the statistical data and the performance of the transistor, wherein the performance indexes are indexes directly related to degradation mechanisms and degradation;
Step S130, establishing an estimation system of the residual life based on the performance index;
step S140, evaluating the residual life of the switch on the premise that the evaluation system is combined with the operation environment of the transistor;
In some embodiments of the invention, statistics of the remaining lifetime of the switch are obtained based on the foregoing scheme; determining a performance index based on the statistical data in combination with the performance of the transistor, wherein the performance index is a degradation mechanism and an index directly related to degradation; establishing an estimation system of the residual life based on the performance index; and in addition, the residual life of the switch is evaluated on the premise that the estimation system is combined with the operating environment of the transistor, and the evaluation is carried out according to the evaluation result, so that a more accurate residual life curve of the switch can be determined, and a maintenance decision scheme can be scientifically and reasonably formulated. In addition, the estimation variance is substantially reduced by increasing the dimension of the samples and maintaining the diversity of the samples.
These steps are described in detail below.
In step S110, statistics of the remaining lifetime of the switch are obtained.
In the embodiment of the disclosure, the statistical data is taken as the past working data of the residual life and has certain referential property. By estimating the exact remaining lifetime, an optimized maintenance plan is determined in order to repair and replace a failed or degraded switch in time.
As shown in fig. 2, step S120 includes:
S121, determining stress of the transistor during fault according to the operation performance of the transistor and the statistical data;
S122, judging the fault condition of the transistor based on the stress of the transistor during fault;
s123, determining corresponding early warning parameters according to the fault condition of the transistor;
S124, acquiring corresponding early warning parameters;
s125, forming a track according to the corresponding early warning parameters so as to predict the faults of the transistors.
In S121, the operation performance of the transistor is presented by using actual data and corresponds to the statistical data, and the stress of the transistor in the fault is obtained by actually comparing the operation performance of the transistor with the statistical data, so as to improve the accuracy of determining the stress of the transistor in the fault.
Judging the final result of mechanical stress, the heat generated in the on-off state can not be eliminated because the transistor usually works under the high frequency condition, and the steady-state temperature of the module is increased. Temperatures approaching the nominal characteristic temperature are more damaging than similar variations in lower temperatures.
VCE, ON, IC, VGE, VGE, TH, ton, toff together with the node temperature (T j) are used as an indicator parameter for the IGBT fault precursor. Wherein T j is the most influencing parameter for degradation, VCE, ON is more sensitive to degradation and has better ON-line measurement capability and accuracy. VGE, TH are effective failure precursors to gate oxide failure. Ton and Toff are precursors to failure when the IGBT is shorted, where VCE, ON is susceptible to industrial noise.
In addition, according to the running performance of the transistor and the detected statistical data, a degradation mechanism and performance indexes directly related to degradation are determined, so that the degradation performance of the transistor is adjusted and the influence of variance is reduced according to the performance indexes.
According to the method for estimating the residual life of the transistor capable of judging the degradation degree, specifically, according to the estimation of the auxiliary particle filtering of the particle filter, the estimation of the residual life of the switch based on the particle filtering is established, so that the accuracy of the estimation is improved.
In S122, the fault condition of the transistor is determined based on the stress of the transistor during the fault, and the corresponding fault condition of the transistor is determined by grading the stress value, so as to facilitate distinguishing the emergency degree of the fault, so that the user can reasonably arrange the corresponding maintenance plan according to different fault degrees.
In S123, corresponding early warning parameters are determined according to the fault condition of the transistor, and the early warning parameters can intuitively display the corresponding emergency degree, so that the user can check the emergency degree conveniently.
In S124, corresponding early warning parameters are acquired, and the data is fed back to the control system, so that the control system can record and react.
In S125, the corresponding early warning parameters are corresponding to data points, and a plurality of data points form a track in the corresponding coordinate system, so that the fault state of the transistor is predicted through the extension of the track, so that the fault can be prevented and regulated in time.
In addition, before S125, it includes: introducing a weight value into the corresponding early warning parameter; resampling is employed to adjust the weight values.
The method can effectively reduce the estimated error rate through the setting of the weight value, and adjust important fault factors, in addition, resampling is adopted for the adjustment of the weight value, and the weight value is adjusted through repeated sampling data so as to facilitate the improvement of the rationality of the weight value, thereby improving the accuracy of the estimation.
In summary, in a specific embodiment, a method for estimating the remaining lifetime of a transistor capable of determining the degradation is found according to the performance index directly related to the degradation. The process of estimating the remaining lifetime of a transistor using a particle filtering method updates a data-driven model using current data, and the updated model can be used to simulate a trajectory.
The trajectories of VCE, ON can be estimated using a bayesian tracking system as follows:
VCE,ON,pre,n＝f(VCE,ON,pre,n-1)+vn-1
VCE,ON,act,n＝f(VCE,ON,pre,n)+mn
The above equations are state transition and measurement models, where VCE, ON, act, n are measured VCE, measured value of ON at time n, VCE, ON, pre, n are predicted VCE, predicted value of ON at time n, v is process noise, m is measurement noise, and n is an integer. These models are built from time series data using an Autoregressive (AR) method. These models are updated in real time as the data is updated to the current data.
The estimation of VCE, ON, pre, n-1 and VCE, ON, act, n-1 based VCE, ON, pre, n is a stochastic estimation problem that is caused by noise and is denoted p (VCE, ON, pre, n|VCE, ON, act, n). Such a distribution is called a posterior distribution and can be defined by measurement noise, mn and statistics of the measurement model. At time n=0, the probability is expressed as follows:
p(VCE,ON,pre,0|VCE,ON,act,0)＝VCE,ON,pre,0
If no a priori knowledge is available in the above equation, it may be chosen as a constant probability distribution. Then, the distribution in the above equation can be updated as long as there is a new measurement result. Over time, this distribution will approach the actual p (VCE, ON, pre, n|vce, ON, act, n). This approximation is called the important density function, q (VCE, ON, pre, n|vce, ON, act, n).
In the prediction step, VCE, ON, act, n is predicted using the Chapman-Kolmogorov equation based ON previous measurements, as follows:
p(VCE,ON,pre,n|VCE,ON,act,n-1)
＝∫p(VCE,ON,pre,n|VCE,ON,pre,n-1)×p(VCE,ON,pre,n-1|VCE,ON,act,n-1)dVCE,ON,pre,n-1
Where p (VCE, ON, pre, n|VCE, ON, pre, n-1) is the transition Probability Distribution Function (PDF) and p (VCE, ON, pre, n-1|VCE, ON, act, n-1) is the posterior probability distribution function of time n-1. The transition probability distribution function is defined by a state transition model.
In the updating step, p (VCE, ON, pre, n|vce, ON, act, n) may be written using bayesian rules as follows:
Where p (VCE, ON, act, n|VCE, ON, pre, n) is a likelihood function and p (VCE, ON, act, n|VCE, ON, act, n-1) is a normalization parameter. The normalization constants are defined as follows:
p(VCE,ON,act,n|VCE,ON,act,n-1)
＝∫p(VCE,ON,act,n|VCE,ON,pre,n)×p(VCE,ON,pre,n|VCE,ON,act,n-1)dVCE,ON,pre,n
In PF, the posterior probability density is represented by random samples and the correlation weights, i.e. importance weights, are calculated as follows:
delta (·) is a dirac trigonometric function, Is the important weight of the i-th sample at time n, i is the index of the i-th sample extracted from the posterior PDF. These random samples are called particles. The importance weights of the particles are calculated using importance sampling principles
The importance weights of the ith particle are defined as follows:
Wherein the method comprises the steps of The importance weight calculation is the ratio of the posterior probability and the importance probability of the point. The importance of PDF is a gaussian approximation of a non-gaussian PDF. The importance of PDF is analyzed as follows:
Combining the above three formulas
Wherein:
In the method, in the process of the invention, Is the incremental importance weight that is recalculated at each step. In the above equation, importance weights are estimated by a recursive method. This PF method is called SISPF.
And a resampling link is introduced in an SIS algorithm, so that the problem of weight degradation is solved. The basic principle of the resampling step is to replicate particles with higher weights and eliminate particles with lower weights so that the weight of each particle becomes the same. According to this principle it is possible to:
In step S130, an estimation system of the remaining lifetime is established based on the performance index; the method for estimating the residual life of the transistor capable of judging the degradation degree is found according to the performance index directly related to the degradation, and a particle filter is adopted for estimating the system.
Specifically, in the auxiliary particle filtering, the index of the sampling particle is introduced as an auxiliary variable. Auxiliary particle filtering generates samples in pairs { V CE,ON,pre,n i,ji }, where j is the index of the ith particle at time n-1.
The joint probability of a particle pair { V CE,ON,pre,n i,ji } is proportional to the transition probability distribution function, p (V CE,ON,pre,n|VCE,ON,act,n) and the importance weight of these particles. This relationship can be expressed as:
In auxiliary particle filtering, particles are first extracted from this joint probability, and then the index is discarded to sample from the edge PDF, p (V CE,ON,pre,n,j|VCE,ON,act,n). Particles are extracted from the important PDF due to the non-gaussian nature of the edge PDF. The important PDF, q (V CE,ON,pre,n,j|VCE,ON,act,n), can be expressed as:
In the method, in the process of the invention, Is the average value of q (V CE,ON,pre,n|VCE,ON,act,n), defined as follows:
The importance weight of the i-th particle can be calculated as follows:
work estimation algorithm based on auxiliary particle filtering:
And establishing a state transition model and a measurement model according to experimental data of the transistor, selecting a proper importance PDF, and predicting the importance PDF-V CE,ON according to the models. The predicted values are compared to V CE,ON,act and the model is adjusted based on these measurements. From these adjusted models, N tracks of V CE,ON were simulated. The remaining life is estimated by calculating the time required from these trajectories to the critical value. The remaining lifetime of time n is calculated as follows:
RULn＝nf-n
RUL n is the estimated remaining life at time n, n f is the time from the current time n to the critical V CE,ON of the failure precursor. In fig. 9, the remaining life estimate is depicted at time n. For each particle, the trajectory of the fault precursor can be modeled. For the first particle, the remaining lifetime is For the nth particle, the remaining lifetime isFor the actual remaining life is/>Where n r is the actual time at which the transistor under test actually fails.
In the sampling example, the state transition model and the measurement model may be updated using a residual life estimation algorithm based on auxiliary particle filtering. And correcting the ballistic projection according to the updated model.
The error in the remaining life estimate is a criterion for performance comparison. Typically, the Mean Absolute Error (MAE) is used in cases where the error is constantly distributed, while RMSE is used in cases where the error distribution is not constant. In addition, RMSE takes into account the effects of outliers, where MAE eliminates the effects of outliers, and thus RMSE is more applicable herein. The error of the remaining life estimate is calculated as follows:
Wherein RUL r is the actual remaining life and RUL i is the ith grain derived from the importance density function
Residual life estimate of the child.
To further improve the accuracy of the residual life estimation based on auxiliary particle filtering, auxiliary particle filtering is applied based on the approximate conditions of the transistor, dividing the degradation situation of the transistor and the resulting change in V CE,ON into three regions, namely a healthy region, a constant growth region and an exponentially growing region.
In step S140, an estimate of the remaining lifetime based on the particle filter is established from the estimate of the auxiliary particle filter, which is evaluated in combination with the operating conditions of the transistor.
In some embodiments of the present invention, referring to fig. 3, a transistor remaining lifetime estimation method based on auxiliary particle filtering includes the steps of:
Determining an optimized maintenance plan by estimating the exact remaining lifetime in order to repair and replace a failed or degraded switch in time;
determining a degradation mechanism and performance indexes directly related to degradation according to the running performance of the transistor and the detected statistical data;
According to the performance index of the degradation direct correlation, searching a transistor residual life estimation method capable of judging the degradation degree, and adopting a Particle Filter (PF) as a nonlinear system and non-Gaussian noise estimation;
establishing residual life estimation based on particle filtering according to the estimation of the particle filter;
The remaining lifetime estimate based on particle filtering is evaluated in combination with the operating conditions of the transistor.
Experimental tests were performed below to determine WBLO and solder fatigue mechanisms, power cycle tests were performed to verify the performance of the proposed APF-based residual life estimation, and a block diagram of the test apparatus is shown in fig. 4. The test bed comprises a direct current power supply, a driving circuit and a resistance load. The switching of the transistors is controlled by a gate drive circuit employing TI-DSPF 28335. TC was collected using a type k thermocouple and monitored with a differential voltage probe. During power cycle testing, the transistor is subjected to temperature cycles that simulate the stresses experienced by the transistor in an industrial application, as shown in fig. 5. When TC reaches 150 ℃, the switch is stopped to cool the transistor to 25 ℃. The temperature dependent process is automated using a Texas Instruments (TI) Digital Signal Processor (DSP). VCE, ON, IC and TC were obtained using NI DAQUSB 9001. These tests were performed on transistors of the Infrax technology (model: FS30R06W1E 3). These transistors have a breakdown voltage of 600 volts and a maximum current of 20A. The experiment was performed at a voltage of 60V and a current of 30A.
Residual life estimation using APF method: starting from the transistor 1 shown in fig. 6, the 30-particle SIR-PF and APF methods are applied to VCE, ON. The root mean square error of the 30 particle SIR-PF and APF were 22% and 17.8%, respectively. Figure 7 shows APF simulated trajectories for different numbers of particles. For an APF of 100 particles, the simulated trajectory is very close to the actual trajectory and the variance is small. The root mean square error decreases with increasing APF and SIR-PF particle numbers. And APF shows better performance.
At very low process noise levels, the residual life estimation error is high, independent of the measured noise level. When the standard deviation of the measurement noise and the process noise is smaller than 0.8v and in the range of 0.05v to 0.25v, respectively, the percentage error of the residual life estimation is about 8%. For SIR-PF, the residual life estimation error may be as low as 13%. The performance of APF is significantly better than SIR-PF at different process and measurement noise.
The technical scheme provided by the embodiment of the invention can comprise the following beneficial effects:
In some embodiments of the present invention, statistical data of remaining lifetime of a switch is obtained; determining a performance index based on the statistical data in combination with the performance of the transistor, wherein the performance index is a degradation mechanism and an index directly related to degradation; establishing an estimation system of the residual life based on the performance index; and in addition, the residual life of the switch is evaluated on the premise that the estimation system is combined with the operating environment of the transistor, and the evaluation is carried out according to the evaluation result, so that a more accurate residual life curve of the switch can be determined, and a maintenance decision scheme can be scientifically and reasonably formulated. In addition, the method substantially reduces the estimation variance by increasing the dimensionality of the samples and maintaining the diversity of the samples. Furthermore, a simple slope-based method is proposed to identify regions of significant degradation in transistors. When the transistor s under test enters this region, the APF method is adopted. This step can reduce variations in the RUL estimation and calculation costs.
In addition, the invention has the following advantages:
1. According to the invention, according to a power cycle test, the performance of the proposed APF-based residual life estimation is verified, WBLO and a solder fatigue mechanism are determined, the randomness generated by a fault form and the correlation among performance indexes of different fault forms are synthesized, evaluation is carried out according to an evaluation result, a more accurate residual life curve can be determined, and a maintenance decision scheme is definitely, scientifically and reasonably formulated;
2. The method simulates the temperature cycle of the stress experienced by the transistor in industrial application, has strict and reliable logic and higher feasibility, and comprehensively and reasonably considers various performance index factors possibly influencing the performance of the transistor;
3. The invention can enable the residual life to approach to the minimum error according to the estimation result of the residual life of the transistor based on auxiliary particle filtering, so that the service life of the transistor device can be predicted by the staff most accurately, and the invention has great significance for ensuring the safe, stable and economic operation of the power grid and has obvious social and economic benefits.
As shown in fig. 8, in one embodiment, the apparatus 200 of the remaining life estimation method further includes:
An obtaining module 210, configured to obtain statistics of remaining lifetime of the switch;
a combination module 220, configured to combine the performances of the transistors based on the statistical data, and determine performance indexes, where the performance indexes are indexes directly related to degradation mechanism and degradation;
a building module 230, configured to build an estimation system of the remaining lifetime based on the performance index;
an evaluation module 240 for evaluating the remaining lifetime of the switch on the premise that the evaluation system incorporates the operating environment of the transistor.
An electronic device 40 according to this embodiment of the present invention is described below with reference to fig. 9. The electronic device 40 shown in fig. 4 is merely an example and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.
As shown in fig. 9, the electronic device 40 is in the form of a general purpose computing device. Components of electronic device 40 may include, but are not limited to: the at least one processing unit 41, the at least one memory unit 42, a bus 43 connecting the different system components, including the memory unit 42 and the processing unit 41.
Wherein the storage unit stores program code that is executable by the processing unit 41 such that the processing unit 41 performs the steps according to various exemplary embodiments of the present invention described in the above-described "example methods" section of the present specification.
The memory unit 42 may include readable media in the form of volatile memory units, such as Random Access Memory (RAM) 421 and/or cache memory 422, and may further include Read Only Memory (ROM) 423.
The storage unit 42 may also include a program/utility 424 having a set (at least one) of program modules 425, such program modules 425 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment.
The bus 43 may be one or more of several types of bus structures including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.
Electronic device 40 may also communicate with one or more external devices (e.g., keyboard, pointing device, bluetooth device, etc.), one or more devices that enable a user to interact with electronic device 40, and/or any device (e.g., router, modem, etc.) that enables electronic device 40 to communicate with one or more other computing devices. Such communication may be through an input/output (I/O) interface 45. Also, electronic device 40 may communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet, through network adapter 46. As shown, the network adapter 46 communicates with other modules of the electronic device 40 over the bus 43. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with electronic device 40, including, but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage systems, and the like.
From the above description of embodiments, those skilled in the art will readily appreciate that the example embodiments described herein may be implemented in software, or may be implemented in software in combination with the necessary hardware. Thus, the technical solution according to the embodiments of the present disclosure may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (may be a CD-ROM, a U-disk, a mobile hard disk, etc.) or on a network, including several instructions to cause a computing device (may be a personal computer, a server, a terminal device, or a network device, etc.) to perform the method according to the embodiments of the present disclosure.
According to an embodiment of the present disclosure, there is also provided a computer-readable storage medium having stored thereon a program product capable of implementing the method described above in the present specification. In some possible embodiments, the various aspects of the invention may also be implemented in the form of a program product comprising program code for causing a terminal device to carry out the steps according to the various exemplary embodiments of the invention as described in the "exemplary methods" section of this specification, when said program product is run on the terminal device.
Referring to fig. 10, a program product 50 for implementing the above-described method according to an embodiment of the present invention is described, which may employ a portable compact disc read only memory (CD-ROM) and include program code, and may be run on a terminal device, such as a personal computer. However, the program product of the present invention is not limited thereto, and in this document, a readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium would include the following: an electrical connection having one or more wires, a portable disk, a hard disk, random Access Memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The computer readable signal medium may include a data signal propagated in baseband or as part of a carrier wave with readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A readable signal medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Program code for carrying out operations of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of remote computing devices, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., connected via the Internet using an Internet service provider).
Furthermore, the above-described drawings are only schematic illustrations of processes included in the method according to the exemplary embodiment of the present invention, and are not intended to be limiting. It will be readily appreciated that the processes shown in the above figures do not indicate or limit the temporal order of these processes. In addition, it is also readily understood that these processes may be performed synchronously or asynchronously, for example, among a plurality of modules.
It is to be understood that the invention is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the invention is limited only by the appended claims.
Claims (6)
1. A remaining life estimation method, comprising:
step S110, acquiring statistical data of the residual life of the switch;
Step S120, determining performance indexes based on the statistical data and the performance of the transistor, wherein the performance indexes are indexes directly related to degradation mechanisms and degradation; step S120 includes: determining the stress of the transistor in fault according to the operation performance of the transistor and the statistical data; judging a fault condition of the transistor based on the stress of the transistor at the time of fault; determining corresponding early warning parameters according to the fault condition of the transistor; acquiring corresponding early warning parameters; forming a track according to the corresponding early warning parameters so as to predict the faults of the transistors;
step S130, establishing a residual life estimation system based on the performance index;
Step S140, evaluating the residual life of the switch on the premise that the evaluation system is combined with the operation environment of the transistor;
Estimating the remaining life of the transistor using a particle filtering method, updating a data driving model using current data, and simulating a trajectory using the updated model; the trajectory of V CE,ON can be estimated using a bayesian tracking system as follows:
VCE,ON,pre,n＝f(VCE,ON,pre,n-1)+vn-1
VCE,ON,act,n＝f(VCE,ON,pre,n)+mn
The above equation is state transition model and measurement model, where V CE,ON,act,n is the measured value of measured V CE,ON at time n, V CE,ON,pre,n is the predicted value of predicted V CE,ON at time n, V is process noise, m is measurement noise, and n is an integer; these models are built from time series data using an Autoregressive (AR) method; when the data is updated to be the current data, the models are updated in real time; the estimation of V CE,ON,pre,n based on V CE,ON,pre,n-1 and V CE,ON,act,n-1 is a random estimation problem, which is caused by noise, denoted p (V CE,ON,pre,n|VCE,ON,act,n); such a distribution is called a posterior distribution and can be defined by measurement noise, m n and statistics of the measurement model.
2. The method of claim 1, wherein after determining the performance index based on the statistics in combination with the performance of the transistor, further comprising:
an estimate of the remaining lifetime of the particle filter based switch is established from the estimate of the auxiliary particle filter.
3. The method of claim 1, wherein the forming a trace based on the corresponding pre-alarm parameters to predict a failure of the transistor further comprises:
Introducing a weight value into the corresponding early warning parameter;
resampling is employed to adjust the weight values.
4. An apparatus for a remaining life estimation method, comprising:
the acquisition module is used for acquiring statistical data of the residual life of the switch;
a combination module, configured to combine the performance of the transistor based on the statistical data, and determine a performance index, where the performance index is an index directly related to degradation and a degradation mechanism; step S120 includes: determining the stress of the transistor in fault according to the operation performance of the transistor and the statistical data; judging a fault condition of the transistor based on the stress of the transistor at the time of fault; determining corresponding early warning parameters according to the fault condition of the transistor; acquiring corresponding early warning parameters; forming a track according to the corresponding early warning parameters so as to predict the faults of the transistors;
the establishing module is used for establishing an estimation system of the residual life based on the performance index;
the evaluation module is used for evaluating the residual service life of the switch on the premise that the evaluation system is combined with the operation environment of the transistor; estimating the remaining life of the transistor using a particle filtering method, updating a data driving model using current data, and simulating a trajectory using the updated model; the trajectories of VCE, ON can be estimated using a bayesian tracking system as follows:
VCE,ON,pre,n＝f(VCE,ON,pre,n-1)+vn-1
VCE,ON,act,n＝f(VCE,ON,pre,n)+mn
The above equation is state transition model and measurement model, where V CE,ON,act,n is the measured value of measured V CE,ON at time n, V CE,ON,pre,n is the predicted value of predicted V CE,ON at time n, V is process noise, m is measurement noise, and n is an integer; these models are built from time series data using an Autoregressive (AR) method; when the data is updated to be the current data, the models are updated in real time; the estimation of V CE,ON,pre,n based on V CE,ON,pre,n-1 and V CE,ON,act,n-1 is a random estimation problem, which is caused by noise, denoted p (V CE,ON,pre,n|VCE,ON,act,n); such a distribution is called a posterior distribution and can be defined by measurement noise, m n and statistics of the measurement model.
5. A computer readable program medium, characterized in that it stores computer program instructions, which when executed by a computer, cause the computer to perform the method according to any of claims 1 to 3.
6. An electronic device, comprising:
A processor;
A memory having stored thereon computer readable instructions which, when executed by the processor, implement the method of any of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010666870.8A CN111832226B (en) | 2020-07-13 | 2020-07-13 | IGBT residual life estimation method based on auxiliary particle filtering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010666870.8A CN111832226B (en) | 2020-07-13 | 2020-07-13 | IGBT residual life estimation method based on auxiliary particle filtering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111832226A CN111832226A (en) | 2020-10-27 |
| CN111832226B true CN111832226B (en) | 2024-04-26 |
Family
ID=72900959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010666870.8A Active CN111832226B (en) | 2020-07-13 | 2020-07-13 | IGBT residual life estimation method based on auxiliary particle filtering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111832226B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112507597B (en) * | 2020-11-02 | 2021-10-22 | 中国南方电网有限责任公司超高压输电公司广州局 | Multi-cluster particle filter based multi-terminal flexible direct current transmission system state evaluation method |
| CN112632843A (en) * | 2020-12-30 | 2021-04-09 | 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）） | Electronic component fault management method and device |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| SU788078A1 (en) * | 1979-02-08 | 1980-12-15 | Поволжское Отделение Всесоюзного Государственного Проектно-Изыскательского И Научно-Исследовательского Института "Сельэнергопроект" | Integrating indicator of interrepair service life of contact system |
| CN102567560A (en) * | 2010-12-27 | 2012-07-11 | 北京圣涛平试验工程技术研究院有限责任公司 | Method and system for estimating service life of MOS (Metal Oxide Semiconductor) device |
| CN104463331A (en) * | 2014-12-29 | 2015-03-25 | 北京航空航天大学 | Accelerated degradation experiment modeling method based on fuzzy theory |
| CN105373660A (en) * | 2015-11-12 | 2016-03-02 | 成都嘉石科技有限公司 | Equivalent circuit-based transistor reliability representation method |
| CN107315138A (en) * | 2016-04-21 | 2017-11-03 | 深圳市智鼎自动化技术有限公司 | Power MOSFET failure predication and health treatment method and test system |
| US9869722B1 (en) * | 2016-09-22 | 2018-01-16 | Rockwell Automation Technologies, Inc. | Method and apparatus for electrical component life estimation |
| CN108051722A (en) * | 2017-12-22 | 2018-05-18 | 中国电子产品可靠性与环境试验研究所 | The lifetime estimation method and system of hot carrier injection effect |
| CN109738773A (en) * | 2018-06-19 | 2019-05-10 | 北京航空航天大学 | IGBT module life-span prediction method under a kind of non-stationary operating condition |
| CN109885849A (en) * | 2018-05-07 | 2019-06-14 | 长春工业大学 | Trolley coach microswitch method for predicting residual useful life based on Strong tracking filter |
| CN110175388A (en) * | 2019-05-21 | 2019-08-27 | 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）） | Switching Power Supply failure prediction method, device, computer equipment and storage medium |
| CN110457789A (en) * | 2019-07-25 | 2019-11-15 | 桂林电子科技大学 | A kind of lithium ion battery residual life prediction technique merged based on improvement particle filter with double exponential decay experience physical models |
| CN110750848A (en) * | 2018-07-20 | 2020-02-04 | 南京航空航天大学 | Method for estimating remaining life of software-hardware degradation system by considering software operation |
| CN111323663A (en) * | 2020-02-26 | 2020-06-23 | 中南大学 | Electromagnetic valve service life prediction method and device based on current feature extraction |
-
2020
- 2020-07-13 CN CN202010666870.8A patent/CN111832226B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| SU788078A1 (en) * | 1979-02-08 | 1980-12-15 | Поволжское Отделение Всесоюзного Государственного Проектно-Изыскательского И Научно-Исследовательского Института "Сельэнергопроект" | Integrating indicator of interrepair service life of contact system |
| CN102567560A (en) * | 2010-12-27 | 2012-07-11 | 北京圣涛平试验工程技术研究院有限责任公司 | Method and system for estimating service life of MOS (Metal Oxide Semiconductor) device |
| CN104463331A (en) * | 2014-12-29 | 2015-03-25 | 北京航空航天大学 | Accelerated degradation experiment modeling method based on fuzzy theory |
| CN105373660A (en) * | 2015-11-12 | 2016-03-02 | 成都嘉石科技有限公司 | Equivalent circuit-based transistor reliability representation method |
| CN107315138A (en) * | 2016-04-21 | 2017-11-03 | 深圳市智鼎自动化技术有限公司 | Power MOSFET failure predication and health treatment method and test system |
| US9869722B1 (en) * | 2016-09-22 | 2018-01-16 | Rockwell Automation Technologies, Inc. | Method and apparatus for electrical component life estimation |
| CN108051722A (en) * | 2017-12-22 | 2018-05-18 | 中国电子产品可靠性与环境试验研究所 | The lifetime estimation method and system of hot carrier injection effect |
| CN109885849A (en) * | 2018-05-07 | 2019-06-14 | 长春工业大学 | Trolley coach microswitch method for predicting residual useful life based on Strong tracking filter |
| CN109738773A (en) * | 2018-06-19 | 2019-05-10 | 北京航空航天大学 | IGBT module life-span prediction method under a kind of non-stationary operating condition |
| CN110750848A (en) * | 2018-07-20 | 2020-02-04 | 南京航空航天大学 | Method for estimating remaining life of software-hardware degradation system by considering software operation |
| CN110175388A (en) * | 2019-05-21 | 2019-08-27 | 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）） | Switching Power Supply failure prediction method, device, computer equipment and storage medium |
| CN110457789A (en) * | 2019-07-25 | 2019-11-15 | 桂林电子科技大学 | A kind of lithium ion battery residual life prediction technique merged based on improvement particle filter with double exponential decay experience physical models |
| CN111323663A (en) * | 2020-02-26 | 2020-06-23 | 中南大学 | Electromagnetic valve service life prediction method and device based on current feature extraction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111832226A (en) | 2020-10-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Celaya et al. | Prognostics of power MOSFETs under thermal stress accelerated aging using data-driven and model-based methodologies | |
| Alaswad et al. | A review on condition-based maintenance optimization models for stochastically deteriorating system | |
| Wang et al. | A two-stage data-driven-based prognostic approach for bearing degradation problem | |
| CN111832226B (en) | IGBT residual life estimation method based on auxiliary particle filtering | |
| CN109242135B (en) | Model operation method, device and business server | |
| Zhang et al. | A Wiener process model with dynamic covariate for degradation modeling and remaining useful life prediction | |
| CN110852509A (en) | Fault prediction method and device of IGBT module and storage medium | |
| CN110309577A (en) | A kind of submarine pipeline method for predicting residual useful life based on IM and LMLE-BU algorithm | |
| US8793106B2 (en) | Continuous prediction of expected chip performance throughout the production lifecycle | |
| Ye et al. | A new class of multi-stress acceleration models with interaction effects and its extension to accelerated degradation modelling | |
| US20220222527A1 (en) | Method for predicting hardware fault using machine learning | |
| Daroogheh et al. | An improved particle filtering-based approach for health prediction and prognosis of nonlinear systems | |
| CN102799946A (en) | Method and device for dynamically predicting reliability of electronic product | |
| CN110441669A (en) | The gradual failure diagnosis of uncertain sophisticated circuitry system and life-span prediction method | |
| Rizvi et al. | Real-time zip load parameter tracking using sensitivity-based adaptive window and variable elimination with realistic synchrophasor data | |
| CN117250490A (en) | LSTM-based residual current operated circuit breaker service life prediction method | |
| KR20170140753A (en) | Method and apparatus for analyzing safety of software | |
| McMenemy et al. | A machine learning approach for adaptive classification of power mosfet failures | |
| US10884405B2 (en) | Method for assessing health conditions of industrial equipment | |
| CN115640542A (en) | Intelligent electric energy meter state evaluation method and evaluation device based on Bayesian theory | |
| Haque et al. | RUL estimation of power semiconductor switch using evolutionary time series prediction | |
| Vaccaro et al. | Remaining Useful Lifetime Prediction of Discrete Power Devices by Means of Artificial Neural Networks | |
| Mojarad et al. | Markov-based anomaly correction in embedded systems | |
| CN112581727A (en) | Displacement drift early warning method, device, equipment and storage medium of bridge | |
| Okazaki et al. | Prediction of the number of defects in image sensors by VM using equipment QC data |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |