CN111400856A - Spatial traveling wave tube reliability assessment method based on multi-source data fusion - Google Patents
Spatial traveling wave tube reliability assessment method based on multi-source data fusion Download PDFInfo
- Publication number
- CN111400856A CN111400856A CN201910465869.6A CN201910465869A CN111400856A CN 111400856 A CN111400856 A CN 111400856A CN 201910465869 A CN201910465869 A CN 201910465869A CN 111400856 A CN111400856 A CN 111400856A
- Authority
- CN
- China
- Prior art keywords
- data
- node
- traveling wave
- wave tube
- test
- 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.)
- Granted
Links
Images
Landscapes
- Amplifiers (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
A space traveling wave tube reliability assessment method based on multi-source data fusion comprises the following steps: s1, processing the collected data and extracting features; s2, calculating the edge probability of the bottom-layer input node of the established Bayesian network model according to the result obtained in the step S1; s3, calculating the edge probability of the input nodes of other layers of the Bayesian network model according to the edge probability of the bottom layer input node obtained in the step S2; and S4, calculating a reliability point estimation result of the traveling wave tube according to the calculation result obtained in the step S3 and the traveling wave tube Bayes network model. The invention can utilize real-time on-orbit telemetering data to realize the running reliability evaluation of the traveling wave tube while carrying out static reliability evaluation on the product, and the result can be used as an important basis for the working state of a key single machine of the satellite and on-orbit health management.
Description
Technical Field
The invention relates to the field of evaluating reliability of spacecraft single-machine products by utilizing multi-source data, in particular to a method for evaluating reliability of a space traveling wave tube based on multi-source data fusion.
Background
The space traveling wave tube is a key component of a satellite-borne transponder and a satellite-borne synthetic aperture radar transmitter, the effect of the space traveling wave tube is mainly to realize the amplification of microwave power, and the reliability of the space traveling wave tube is very important for the stable operation of an entire satellite system. The traveling wave tube has a complex structure, many related processes and parts and high price, and belongs to a small sample product.
The reliability evaluation result obtained by the traditional statistical method mostly takes failure time as a statistical object, a large number of life tests or accelerated life tests are used for obtaining failure data of a product, then the reliability of the product at the end of the service life is evaluated by assuming that the life distribution obeys exponential distribution or Weibull distribution, and the evaluation result is conservative due to the short test time and the limited failure number. The traveling wave tube accumulates a great amount of data and quality problems in research, ground test and on-orbit flight stages, and research on a reliability evaluation method of the traveling wave tube fusing multi-source data is imperative in order to obtain a credible evaluation result by fully utilizing the data.
Disclosure of Invention
In view of the above, one of the main objectives of the present invention is to provide a method for evaluating reliability of a spatial traveling wave tube based on multi-source data fusion, so as to at least partially solve at least one of the above technical problems.
In order to achieve the above object, as an aspect of the present invention, a method for evaluating reliability of a spatial traveling wave tube based on multi-source data fusion is provided, including the following steps:
s1, processing the collected data and extracting features;
s2, calculating the edge probability of the bottom-layer input node of the established Bayesian network model according to the result obtained in the step S1;
s3, calculating the edge probability of the input nodes of other layers of the Bayesian network model according to the edge probability of the bottom layer input node obtained in the step S2;
and S4, calculating a reliability point estimation result of the traveling wave tube according to the calculation result obtained in the step S3 and the traveling wave tube Bayes network model.
Based on the technical scheme, compared with the prior art, the method for evaluating the reliability of the spatial traveling wave tube based on the multi-source data fusion has at least one of the following advantages:
(1) the invention makes full use of the data of the traveling wave tube in the development stage, the ground test stage and the on-orbit flight stage, and makes up the defects of the traditional reliability estimation method based on single data, in particular to the following steps:
firstly, establishing a hierarchical Bayesian network according to FMEA (failure mode and influence analysis), FTA (fault tree analysis) and collected fault information of a space traveling wave tube and a traveling wave tube amplifier and the composition structure of each component of the traveling wave tube, wherein a node represents the failure mode of the traveling wave tube, and an arrow represents the causal relationship (cause pointing result) between a lower node and an upper node;
secondly, the bottom node of the invention selects data (from different stages/processes) capable of representing the service life and reliability of the traveling wave tube, wherein cathode accelerated life test data is associated with a fault mode related to an electron gun assembly, performance data and on-orbit telemetering data are associated with the fault modes of each assembly of the traveling wave tube, and environment simulation test data is associated with the fault mode sensitive to thermal cycle and thermal vacuum environment;
finally, the invention can update the edge probability value of the bottom node according to whether the bottom data of the Bayesian network is updated or not, and further update the reliability evaluation result of the traveling wave tube;
(2) the invention can utilize real-time on-orbit telemetering data to realize the running reliability evaluation of the traveling wave tube while carrying out static reliability evaluation on the product, and the result can be used as an important basis for the working state of a key single machine of the satellite and on-orbit health management;
(3) compared with a reliability evaluation method by a series-parallel connection model adopted in a reliability block diagram, the method adopting the Bayesian network model has the advantages of conditional independence, polymorphic expression, uncertain relation expression, flexible modeling, bidirectional reasoning and the like, so that the data in the multiple-stage process of the traveling wave tube are fused by the Bayesian network model, and the incidence relation between the characteristic parameters representing the reliability of the traveling wave tube and the fault mode is established.
Drawings
FIG. 1 is a schematic diagram of reliability evaluation using a Bayesian network model in the present embodiment;
fig. 2 is a schematic diagram of a bayesian network model in the present embodiment.
Detailed Description
In order that the objects, technical solutions and advantages of the present invention will become more apparent, the present invention will be further described in detail with reference to the accompanying drawings in conjunction with the following specific embodiments.
The invention discloses a reliability evaluation method for a space traveling wave tube, which comprises the following steps:
s1, processing the collected data and extracting features;
s2, calculating the edge probability of the bottom-layer input node of the established Bayesian network model according to the result obtained in the step S1;
s3, calculating the edge probability of the input nodes of other layers of the Bayesian network model according to the edge probability of the bottom layer input node obtained in the step S2;
and S4, calculating a reliability point estimation result of the traveling wave tube according to the calculation result obtained in the step S3 and the traveling wave tube Bayes network model.
After step S4, the method further includes:
s5, collecting data again, if the collected data is updated, repeating the steps S1-S4, and updating the reliability point estimation result of the traveling wave tube;
and the data acquisition process is continuous, the step S5 is repeated after one period of data acquisition is finished, and the reliability point estimation result of the traveling wave tube is continuously updated.
The method for processing the collected data and extracting the features in step S1 includes: cleaning and processing data according to the test characteristics and the data types of each stage of the space traveling wave tube;
the method for cleaning and processing the data in the step S1 comprises wild value elimination, deficiency value supplement, normalization and characteristic value extraction;
the data collected in step S1 includes data of a ground test phase and data of an on-orbit telemetry phase.
The ground test stage data comprises any one or more of environment simulation test data, cathode accelerated life test data and performance test data;
wherein the environmental simulation test data comprises data of thermal cycle and thermal vacuum phases;
wherein, the cathode accelerated life test data is collected in the accelerated life test process of the short tube of the electron gun;
wherein the performance test data comprises the output power of the traveling wave tube.
The method for establishing the bayesian network model in step S2 includes: establishing a Bayesian network model according to the fault mode and influence analysis, fault tree analysis, weak link analysis and characteristic parameters related to the reliability of the traveling wave tube amplifier and the traveling wave tube;
the characteristic parameters related to the reliability of the traveling wave tube comprise cathode accelerated life test data, on-orbit remote measurement data, performance data and environment simulation test data.
Wherein, the bayesian network model in the step S2 is a hierarchical bayesian network model.
Wherein, the input nodes of the bayesian network model in the step S2 sequentially include a top node, an upper node, a middle node and a bottom node from top to bottom;
the top node is a traveling wave tube abnormal node;
the upper-layer nodes comprise electron gun abnormal nodes, slow wave structure abnormal nodes, magnetic focusing system abnormal nodes and/or collector abnormal nodes; the middle-layer nodes comprise filament abnormal nodes, cathode abnormal nodes, anode abnormal nodes, beam bunching electrode abnormal nodes, spiral flow abnormal nodes, interelectrode insulation deterioration nodes, heat dissipation abnormal nodes and/or mismatch nodes;
the bottom layer nodes comprise a first node, a second node, a third node and a fourth node;
wherein the first node is characterized by cathode accelerated life test data; the second node is characterized with orbit telemetry data; the third node is characterized by performance data; and the fourth node is characterized by environmental simulation test data.
In step S2, the conditional probabilities between the bottom-layer node and the middle-layer node, between the middle-layer node and the upper-layer node, and between the upper-layer node and the top-layer node of the bayesian network model are determined according to expert experience, fault statistical information, and a quality zeroing report.
The conditional probability calculation method of the first node is as follows:
aiming at long-term accelerated life test data developed by a cathode, an Arrhenius model is utilized
Respectively calculating the acceleration factors of each test temperature relative to the normal working temperature, converting the equivalent test time of the cathode at the normal working temperature according to the test time developed at each temperature, and utilizing
Obtaining the equivalent accumulated total test time of the accelerated life test; and according to the failure number counted in the test process, according to an exponential distribution model R (t) ═ e-λtObtaining reliability point estimation of the end of the service life;
wherein A isi-acceleration factor, T0Temperature of the environment of use, Ti-the temperature of the acceleration is such that,
-accelerating the equivalent cumulative test time of the life test,
-cumulative accelerated life test time under ith stress, ni-the amount of test sample at the i-th stress,
-test time of jth sample under ith stress;
the conditional probability calculation method of the second node is as follows:
according to analysis of the variation trend of the on-orbit positive pressure and spiral flow degradation tracks and analysis of the product characteristics of the traveling wave tube amplifier, it is deduced that the degradation reasons of spiral flow and positive pressure remote measurement have correlation; first according to the covariance matrix
Judging the correlation between the degradation amounts; then adopting binary joint distribution for the spiral flow and the positive pressure, assuming that the product degradation obeys normal distribution, according to
Calculating the correlation coefficient between the spiral flow and the male pressure degradation amount to obtain a binary parameter joint probability density function
Finally, the upper limit of the degradation failure of the spiral current is assumed to be 5mA, and the positive voltage rises to degrade the failureUpper limit of 350V, using
Obtaining the reliability under the correlation of the spiral flow and the positive pressure;
the conditional probability calculation method of the third node is as follows:
for product performance parameters influenced by various random factors, the product performance parameters are generally considered to be in accordance with normal distribution, so that the output power of the traveling wave tube is assumed to be in accordance with normal distribution; for performance data, namely output power, of a plurality of traveling wave tubes in a decimetric wave band, judging whether the acquired output power comes from the same large sample matrix or not by utilizing normality test and variance homogeneity test; for samples from the same parent, the mean and variance of the parent of the large sample are respectively obtained and utilized
And
obtaining reliability point estimation based on the performance normality;
wherein mu is the mean value of the performance data, s is the variance of the performance data, K is the tolerance coefficient, L is the lower limit of the output power, and phi (K) is normal distribution;
wherein the decimetric waves include L waves and S waves;
the conditional probability calculation method of the fourth node is as follows:
mainly considering thermal cycle and thermal vacuum phase in the environment simulation test process of the sample piece, respectively utilizing Norris-L and zberg model in the thermal cycle phase
And the thermal vacuum stage using the Arrhenius model
Finding out an environmental factor according to two phasesThe reliability point estimation at the end of the service life is obtained by utilizing Weibull distribution;
wherein N isHL、NOThermal cycling, number of failure cycles under in-orbit stress, Δ THL、ΔTOThermal cycling, temperature range under in-orbit stress, fHL、fOThermal cycling, cycling frequency under in-orbit stress, TmHL、TmO-maximum temperature in thermal cycling, cycling under in-orbit stress, LL、LHTime to failure at low, high stress, TL、TH-temperatures at low, high stress.
Wherein, the step S4 specifically includes: and calculating a reliability point estimation result of the traveling wave tube by using Bayesian fusion software according to the updated Bayesian network model and the edge probability of each layer of input nodes.
In one embodiment of the present invention, for example, the following technical solutions are adopted:
in order to fully utilize data and quality problem information of each stage, the invention discloses a spatial traveling wave tube reliability assessment method based on multi-source data fusion, which associates data with a fault mode and improves the assessment precision of a domestic spatial traveling wave tube.
A traveling wave tube reliability assessment method based on multi-source data fusion comprises the following steps: building a Bayesian network model by using a space traveling wave tube, a space traveling wave tube amplifier FMEA, a space traveling wave tube FTA and quality problems, fusing ground test data (cathode accelerated life test data, performance test data and environment simulation test data) and on-orbit remote measurement data (traveling wave tube amplifier spiral flow and positive voltage), and according to the characteristics of the test and the data at each stage, adopting a corresponding evaluation method to respectively obtain reliability point estimation which is used as the edge probability of a Bayesian network model bottom node to be input into a network; and then, combining with expert experience, and obtaining a reliability evaluation result of the traveling wave tube by utilizing Bayesian reasoning.
The method comprises the following steps:
step one, establishing a Bayesian network model according to an amplifier of the traveling wave tube, FMEA and FTA of the traveling wave tube, weak link analysis and characteristic parameters capable of representing the reliability of the traveling wave tube for evaluating the reliability of the traveling wave tube;
collecting data of a ground test stage (an environment simulation test, a cathode accelerated life test and a performance test) and an on-orbit remote measuring stage, and performing data processing and feature extraction;
selecting a proper method (an expert experience method, an accelerated life test method, a data driving method, an environmental factor folding method and the like) according to the characteristics of the data of each phase, obtaining the edge probability of the input node at the bottom layer of the Bayesian network model, and determining the conditional probability of each input node to the upper node, namely a CPT table (conditional probability table) by combining a quality zeroing report, a fault problem list and expert experience;
step four, judging whether Bayesian updating is carried out or not according to whether the collected on-orbit data, cathode accelerated life test data, fault statistics and other data are updated or not;
and fifthly, calculating a reliability evaluation result of the traveling wave tube by using Bayes fusion software Genie.
Furthermore, the cathode accelerated life test data in the second step are acquired in the process that the M-type cathode is loaded on the electron gun short tube to perform the accelerated life test, and a reliability evaluation method based on the accelerated life test is adopted for the partial data. And respectively calculating acceleration factors at different temperatures by using an Arrhenius model to obtain equivalent total test time, and obtaining a reliability point estimation result at the end of the service life by using exponential distribution in combination with statistical failure data.
Further, the performance test data in the second step is the output power of the traveling wave tube. For product performance parameters influenced by various random factors, the product performance parameters are generally considered to be subjected to normal distribution, and a performance data-output power reliability evaluation method is utilized to carry out reliability evaluation on the product performance parameters;
and further, according to different characteristics of the two processes, the environmental factors of the two stages are calculated by respectively adopting a Norris-L andzberg model and an Arrhenius model, the equivalent total test time is calculated by combining the test time of the traveling wave tubes in the two processes, and then the reliability point estimation corresponding to the environmental simulation test data is obtained by utilizing Weibull distribution according to the counted number and failure number of the traveling wave tubes for a certain series of satellites.
Furthermore, in the in-orbit telemetry data in the second step, a binary parameter joint degradation model is constructed by using the spiral flow and the positive pressure data of a certain series of space traveling wave tube amplifiers for satellites, the degradation model is used for predicting the pseudo life data at the end of the life, and then the reliability point estimation result of the part is obtained.
Further, the evaluation result is input as the edge probability of the bottom node of the Bayesian network model, and the conditional probability of the upper node is determined according to the statistical fault information and the quality problem by combining with expert experience. And finally, updating the edge probability of the bottom node of the Bayesian network according to whether the data of each phase is updated or not, and further updating the reliability evaluation result of the traveling wave tube.
The technical solution of the present invention is further illustrated by the following specific embodiments and the accompanying drawings. It should be noted that the following specific examples are given by way of illustration only and the scope of the present invention is not limited thereto.
The traditional reliability assessment is to use a statistical method to obtain the failure data and failure time of the product through a large number of life and reliability tests. For reliability evaluation of the space traveling wave tube, at present, few documents and reports can be inquired, most of the documents and reports are a life test method directly used, but the method is time-consuming and labor-consuming, and has the problems of less accumulated reliability data, small statistical samples and the like. In order to fully utilize data of each stage of travelling wave tube development, production, test and on-orbit operation and expand reliability data and sample capacity, a method based on multi-source data fusion is needed to be adopted to associate the data of each stage with a fault mode and comprehensively evaluate the reliability of the travelling wave tube.
As shown in fig. 1 and 2, the invention discloses a method for evaluating reliability of a spatial traveling wave tube based on multi-source data fusion, which mainly comprises the following steps:
step 1, establishing a Bayesian network according to an amplifier of the traveling wave tube, FMEA and FTA of the traveling wave tube, weak link analysis and characteristic parameters capable of representing the reliability of the traveling wave tube for evaluating the reliability of the traveling wave tube;
in the step, by combing key factors influencing the reliability of the traveling wave tube, combining an FMEA (failure mode effect), an FTA (fiber to the array) and a quality zero report of the traveling wave tube amplifier and the traveling wave tube, a hierarchical Bayesian network is built, and a system is combined to obtain a Bayesian network model of the reliability of the traveling wave tube, wherein the lowest node of the model corresponds to test process data experienced by the traveling wave tube.
Step 2, collecting data of a ground test stage (an environment simulation test, a cathode accelerated life test and a performance test) and an on-orbit remote measuring stage, and performing data processing and feature extraction;
according to the test characteristics and the data types of the traveling wave tube in each stage, the data are cleaned and processed, and the steps comprise abnormal value removal, missing value supplement, normalization, characteristic value extraction and the like.
Step 3, selecting a proper method (an expert experience method, an accelerated life test method, a data driving method, an environmental factor folding method and the like) according to the characteristics of each stage of data, obtaining the edge probability of the input nodes of the Bayesian network, and determining the conditional probability of each input node to the upper nodes by combining a quality zeroing report, a fault problem list and expert experience;
in the step 2, the results of data processing and feature extraction are utilized, and by combining the type of the developed test and the characteristics of the collected data, reliability point estimation corresponding to each test stage is respectively obtained by adopting different methods, and the reliability point estimation is input into the Bayesian network as the edge probability value of the node at the bottommost layer. The conditional probability of the upper node of the bayesian network is determined based on expert experience, fault statistics, quality zeroing reports, etc. The edge probability value of the bottom layer node is determined as follows:
step 3.1 reliability assessment based on cathode accelerated life test
Aiming at long-term accelerated life test data developed by a cathode, an Arrhenius model is utilized
Respectively calculating the acceleration factors of each test temperature relative to the normal working temperature, converting the equivalent test time of the cathode at the normal working temperature according to the test time developed at each temperature, and utilizing
And obtaining the equivalent accumulated total test time of the accelerated life test. And according to the failure number counted in the test process, according to an exponential distribution model R (t) ═ e-λtAnd obtaining the reliability point estimation of the end of the service life.
Wherein A isi-acceleration factor, T0Temperature of the environment of use, Ti-the temperature of the acceleration is such that,
-accelerating the equivalent cumulative test time of the life test,
-cumulative accelerated life test time under ith stress, ni-the amount of test sample at the i-th stress,
-test time for jth sample under ith stress.
Step 3.2 Performance Normal-based reliability assessment
For the performance data-output power of L wave bands of a plurality of traveling wave tubes, the normality test and the variance homogeneity test are needed to be used for judging whether the output power is from the same large sample parent, for the samples from the same parent, the mean value and the variance of the parent of the large sample are obtained, and the mean value and the variance of the parent of the large sample are used
And
and solving reliability point estimation based on the performance normality.
Wherein μ is the mean of the performance data, s is the variance of the performance data, K is the tolerance coefficient, L is the lower limit of the output power, and Φ (K) is the normal distribution.
Step 3.3 telemetry parameter based on-track reliability assessment
According to the on-orbit positive pressure and spiral flow degradation trajectory change trend analysis and the traveling wave tube amplifier product characteristic analysis, the method deduces that the degradation reasons of spiral flow and positive pressure remote measurement have correlation. First according to the covariance matrix
A correlation between the degradation amounts is judged. Then adopting binary joint distribution for the spiral flow and the positive pressure, assuming that the product degradation obeys normal distribution, according to
Calculating the correlation coefficient between the spiral flow and the male pressure degradation amount to obtain a binary parameter joint probability density function
Finally, assuming that the upper limit of the degradation failure of the spiral current is 5mA and the upper limit of the degradation failure of the positive voltage rise is 350V, utilizing
And obtaining the reliability under the correlation of the spiral flow and the positive pressure.
Step 3.4 reliability assessment based on environmental factor conversion
The step mainly considers the environmental simulation test of the sample pieceThermal cycling and thermal vacuum phases of the process, respectively, in the thermal cycling phase using the Norris-L and zberg model
And the thermal vacuum stage using the Arrhenius model
And (5) solving an environmental factor, and solving reliability point estimation at the end of the service life by utilizing Weibull distribution according to the test time and failure data of the two stages.
Wherein N isHL、NOThermal cycling, number of failure cycles under in-orbit stress, Δ THL、ΔTOThermal cycling, temperature range under in-orbit stress, fHL、fOThermal cycling, cycling frequency under in-orbit stress, TmHL、TmO-maximum temperature in thermal cycling, cycling under in-orbit stress, LL、LHTime to failure at low, high stress, TL、TH-temperatures at low, high stress.
Step 4, judging whether Bayesian updating is carried out or not according to whether the collected on-orbit data, cathode accelerated life test data, fault statistics and other data are updated or not;
the step judges whether Bayesian updating is carried out or not according to whether the data source of the Bayesian network bottom node is updated or not, so that the final traveling wave tube reliability evaluation result is updated. The data which can be updated is on-track telemetering data generally, and the running wave tube operation reliability can be evaluated according to the partial data.
And 5, outputting a reliability point estimation result of the traveling wave tube by utilizing Genie software.
The reliability point estimation result of the traveling wave tube is calculated by using Bayesian fusion software Genie according to the built traveling wave tube Bayesian network reliability evaluation model and the CPT table.
The method provided by the present invention is further illustrated by the following examples.
For example, for a traveling wave tube for a certain satellite, data of multiple stages are fully fused according to a built Bayesian network model, and a correlation relation is established between a reliability characterization parameter of the traveling wave tube and a fault mode. The edge probability value of the bayesian network bottom node corresponding to the cathode accelerated life test data obtained in step 3.1 is 0.925, the edge probability value of the bayesian network bottom node corresponding to the performance data-output power obtained in step 3.2 is 0.9907, the edge probability value of the bayesian network bottom node corresponding to the in-orbit telemetry data-spiral current and positive voltage obtained in step 3.3 is 0.938, and the edge probability value of the bayesian network bottom node corresponding to the environment simulation test data obtained in step 3.4 is 0.92. And finally, carrying out causal reasoning of the Bayesian network by combining a CPT table, and obtaining a reliability point of the traveling wave tube at the end of 12-year service life, wherein the reliability point is estimated to be 0.94. This assessment result is more illustrative because the data and quality issue information incorporating multiple stages is more than the end-of-12-year reliability point estimate of 0.938 using traditional statistical methods.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention and are not intended to limit the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A reliability evaluation method for a space traveling wave tube is characterized by comprising the following steps:
s1, processing the collected data and extracting features;
s2, calculating the edge probability of the bottom-layer input node of the established Bayesian network model according to the result obtained in the step S1;
s3, calculating the edge probability of the input nodes of other layers of the Bayesian network model according to the edge probability of the bottom layer input node obtained in the step S2;
and S4, calculating a reliability point estimation result of the traveling wave tube according to the calculation result obtained in the step S3 and the traveling wave tube Bayes network model.
2. The evaluation method according to claim 1,
after step S4, the method further includes:
s5, collecting data again, if the collected data is updated, repeating the steps S1-S4, and updating the reliability point estimation result of the traveling wave tube;
preferably, the data acquisition process is continuous, and step S5 is repeated after the data acquisition cycle is finished, so as to continuously update the reliability point estimation result of the traveling wave tube.
3. The evaluation method according to claim 1,
the method for processing the collected data and extracting the features in the step S1 includes: cleaning and processing data according to the test characteristics and the data types of each stage of the space traveling wave tube;
preferably, the method for cleaning and processing data in step S1 includes wild value elimination, missing value supplement, normalization and feature value extraction;
preferably, the data collected in step S1 includes data from a ground test phase and data from an on-track telemetry phase.
4. The evaluation method according to claim 3,
the ground test stage data comprises any one or more of environment simulation test data, cathode accelerated life test data and performance test data;
preferably, the environmental simulation test data comprises data of thermal cycle and thermal vacuum phase;
preferably, the cathode accelerated life test data is collected in the accelerated life test process of the short tube of the electron gun;
preferably, the performance test data includes output power of the traveling wave tube.
5. The evaluation method according to claim 1,
the method for establishing the bayesian network model in the step S2 includes: establishing a Bayesian network model according to the fault mode and influence analysis, fault tree analysis, weak link analysis and characteristic parameters related to the reliability of the traveling wave tube amplifier and the traveling wave tube;
preferably, the characteristic parameters related to the reliability of the traveling wave tube comprise cathode accelerated life test data, on-orbit telemetry data, performance data and environment simulation test data.
6. The evaluation method according to claim 1,
the bayesian network model in said step S2 is a hierarchical bayesian network model.
7. The evaluation method according to claim 6,
the input nodes of the bayesian network model in the step S2 sequentially include a top node, an upper node, a middle node and a bottom node from top to bottom;
the top node is a traveling wave tube abnormal node;
the upper-layer nodes comprise electron gun abnormal nodes, slow wave structure abnormal nodes, magnetic focusing system abnormal nodes and/or collector abnormal nodes;
the middle-layer nodes comprise filament abnormal nodes, cathode abnormal nodes, anode abnormal nodes, beam bunching electrode abnormal nodes, spiral flow abnormal nodes, interelectrode insulation deterioration nodes, heat dissipation abnormal nodes and/or mismatch nodes;
the bottom layer nodes comprise a first node, a second node, a third node and a fourth node;
preferably, the first node is characterized by cathode accelerated life test data; the second node is characterized with orbit telemetry data; the third node is characterized by performance data; and the fourth node is characterized by environmental simulation test data.
8. The evaluation method according to claim 7,
the conditional probabilities between the bottom-layer node and the middle-layer node, between the middle-layer node and the upper-layer node, and between the upper-layer node and the top-layer node of the bayesian network model in the step S2 are all determined according to expert experience, fault statistical information, and quality zeroing reports.
9. The evaluation method according to claim 7,
the conditional probability calculation method of the first node is as follows:
aiming at long-term accelerated life test data developed by a cathode, an Arrhenius model is utilized
Respectively calculating the acceleration factors of each test temperature relative to the normal working temperature, converting the equivalent test time of the cathode at the normal working temperature according to the test time developed at each temperature, and utilizing
Obtaining the equivalent accumulated total test time of the accelerated life test; and according to the failure number counted in the test process, according to an exponential distribution model R (t) ═ e-λtObtaining reliability point estimation of the end of the service life;
wherein A isi-acceleration factor, T0Temperature of the environment of use, Ti-the temperature of the acceleration is such that,
-accelerating the equivalent cumulative test time of the life test,
-cumulative accelerated life test time under ith stress, ni-the amount of test sample at the i-th stress,
-stress of the ithTest time for the next j sample;
the conditional probability calculation method of the second node is as follows:
according to analysis of the variation trend of the on-orbit positive pressure and spiral flow degradation tracks and analysis of the product characteristics of the traveling wave tube amplifier, it is deduced that the degradation reasons of spiral flow and positive pressure remote measurement have correlation; first according to the covariance matrix
Judging the correlation between the degradation amounts; then adopting binary joint distribution for the spiral flow and the positive pressure, assuming that the product degradation obeys normal distribution, according to
Calculating the correlation coefficient between the spiral flow and the male pressure degradation amount to obtain a binary parameter joint probability density function
Finally, assuming that the upper limit of the degradation failure of the spiral current is 5mA and the upper limit of the degradation failure of the positive voltage rise is 350V, utilizing
Obtaining the reliability under the correlation of the spiral flow and the positive pressure;
the conditional probability calculation method of the third node is as follows:
for product performance parameters influenced by various random factors, the product performance parameters are generally considered to be in accordance with normal distribution, so that the output power of the traveling wave tube is assumed to be in accordance with normal distribution; for performance data, namely output power, of a plurality of traveling wave tubes in a decimetric wave band, judging whether the acquired output power comes from the same large sample matrix or not by utilizing normality test and variance homogeneity test; for samples from the same parent, the mean and variance of the parent of the large sample are respectively obtained and utilized
And
obtaining reliability point estimation based on the performance normality;
wherein mu is the mean value of the performance data, s is the variance of the performance data, K is the tolerance coefficient, L is the lower limit of the output power, and phi (K) is normal distribution;
preferably, the decimeter waves include L waves and S waves;
the conditional probability calculation method of the fourth node is as follows:
mainly considering thermal cycle and thermal vacuum phase in the environment simulation test process of the sample piece, respectively utilizing Norris-L and zberg model in the thermal cycle phase
And the thermal vacuum stage using the Arrhenius model
Obtaining an environmental factor, and obtaining reliability point estimation at the end of the service life by utilizing Weibull distribution according to the test time and failure data of the two stages;
wherein N isHL、NOThermal cycling, number of failure cycles under in-orbit stress, Δ THL、ΔTOThermal cycling, temperature range under in-orbit stress, fHL、foThermal cycling, cycling frequency under in-orbit stress, TmHL、TmO-maximum temperature in thermal cycling, cycling under in-orbit stress, LL、LHTime to failure at low, high stress, TL、TH-temperatures at low, high stress.
10. The evaluation method according to claim 1,
the step S4 specifically includes: and calculating a reliability point estimation result of the traveling wave tube by using Bayesian fusion software according to the updated Bayesian network model and the edge probability of each layer of input nodes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910465869.6A CN111400856B (en) | 2019-05-30 | 2019-05-30 | Spatial traveling wave tube reliability assessment method based on multi-source data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910465869.6A CN111400856B (en) | 2019-05-30 | 2019-05-30 | Spatial traveling wave tube reliability assessment method based on multi-source data fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111400856A true CN111400856A (en) | 2020-07-10 |
| CN111400856B CN111400856B (en) | 2022-05-27 |
Family
ID=71435869
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910465869.6A Active CN111400856B (en) | 2019-05-30 | 2019-05-30 | Spatial traveling wave tube reliability assessment method based on multi-source data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111400856B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115203059A (en) * | 2022-09-09 | 2022-10-18 | 中国西安卫星测控中心 | Method and system for evaluating reliability of aerospace measurement and control software |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120209575A1 (en) * | 2011-02-11 | 2012-08-16 | Ford Global Technologies, Llc | Method and System for Model Validation for Dynamic Systems Using Bayesian Principal Component Analysis |
| CN109284535A (en) * | 2018-08-14 | 2019-01-29 | 北京空间飞行器总体设计部 | Space travelling wave tube amplifier reliability appraisal procedure and device based on telemetry parameter |
| US20190087294A1 (en) * | 2018-05-11 | 2019-03-21 | Beihang University | Method for establishing fault diagnosis technique based on contingent Bayesian networks |
-
2019
- 2019-05-30 CN CN201910465869.6A patent/CN111400856B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120209575A1 (en) * | 2011-02-11 | 2012-08-16 | Ford Global Technologies, Llc | Method and System for Model Validation for Dynamic Systems Using Bayesian Principal Component Analysis |
| US20190087294A1 (en) * | 2018-05-11 | 2019-03-21 | Beihang University | Method for establishing fault diagnosis technique based on contingent Bayesian networks |
| CN109284535A (en) * | 2018-08-14 | 2019-01-29 | 北京空间飞行器总体设计部 | Space travelling wave tube amplifier reliability appraisal procedure and device based on telemetry parameter |
Non-Patent Citations (2)
| Title |
|---|
| 伏洪勇 等: "基于性能退化模型的空间电子学设备加速寿命试验研究与应用", 《载人航天》 * |
| 厉海涛等: "动量轮在轨状态可靠性贝叶斯网络建模与评估", 《航空学报》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115203059A (en) * | 2022-09-09 | 2022-10-18 | 中国西安卫星测控中心 | Method and system for evaluating reliability of aerospace measurement and control software |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111400856B (en) | 2022-05-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108664700B (en) | Accelerated degradation information fusion modeling method based on uncertain data envelope analysis | |
| CN106248381B (en) | A kind of rolling bearing life dynamic prediction method based on multiple features and phase space | |
| CN112949209B (en) | Degradation rate-fluctuation combined updating method for evaluating storage life of elastic sealing rubber | |
| CN107918103B (en) | Lithium ion battery residual life prediction method based on gray particle filtering | |
| CN110059337B (en) | Gyro system life prediction method based on Bayes multi-source data fusion | |
| CN103699118A (en) | Method and device for analyzing abnormal state of spacecraft in operating process | |
| CN107944594B (en) | Short-term load prediction method based on spearman grade and RKELM microgrid | |
| CN109543743B (en) | Multi-sensor fault diagnosis method for refrigerating unit based on reconstructed prediction residual error | |
| CN110472822B (en) | Intelligent power distribution network power supply reliability evaluation system and method | |
| CN111400856B (en) | Spatial traveling wave tube reliability assessment method based on multi-source data fusion | |
| Bozorg et al. | Bayesian bootstrapping in real-time probabilistic photovoltaic power forecasting | |
| Zhang et al. | Evolving Elman neural networks based state-of-health estimation for satellite lithium-ion batteries | |
| CN108763250B (en) | Photovoltaic power station monitoring data restoration method | |
| CN115630490A (en) | Motor reliability modeling and service life prediction method based on multi-performance parameter degradation | |
| CN114252103A (en) | Fusion power station operation fault prediction method | |
| Nelega et al. | Prediction of power generation of a photovoltaic power plant based on neural networks | |
| Bottone et al. | Failure prediction and diagnosis for satellite monitoring systems using Bayesian networks | |
| Aunali et al. | Comparison of bayesian method and classical charts in detection of small shifts in the control charts | |
| Wang et al. | Large-scale field data-based battery aging prediction driven by statistical features and machine learning | |
| CN114169128A (en) | Reliability enhancement test quantitative evaluation method based on Bayes analysis | |
| Gou et al. | Remaining useful life prediction for lithium-ion battery using ensemble learning method | |
| CN116258067A (en) | Traveling wave tube service life assessment method integrating multi-source heterogeneous data | |
| CN106404712A (en) | Adaptive model correcting method and system based on GT-KF-PLC near infrared spectrum | |
| Lliuyacc-Blas et al. | PV hosting capacity estimation in low voltage feeders through bayesian statistical inference | |
| CN113496255B (en) | Power distribution network mixed observation point distribution method based on deep learning and decision tree driving |
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 |