Satellite electronic equipment reliability assessment method based on data sparse sampling

Abstract

The invention provides a satellite electronic equipment reliability assessment method and device based on data sparse sampling, a medium and electronic equipment, and the method comprises the steps: obtaining a degradation model of a preset performance index of the electronic equipment, and building a reliability calculation model based on the degradation model; under the condition that degradation model parameters of the degradation model have prior information, performance data of the electronic equipment are sampled based on a data sparse sampling strategy of Bayesian estimation; under the condition that degradation model parameters of the degradation model do not have prior information, performance data of the electronic equipment are sampled based on a data sparse sampling strategy of adaptive maximum likelihood estimation; and based on the sampling result and the reliability calculation model, obtaining a reliability evaluation result of the electronic equipment. According to whether the degradation model parameters have prior information or not, the sampling strategy adopting Bayesian estimation or adaptive maximum likelihood estimation is determined to perform efficient and accurate sampling, and the problem of low processing efficiency in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of satellite electronic equipment reliability evaluation, in particular to a satellite electronic equipment reliability evaluation method and device based on data sparse sampling, a medium and electronic equipment. BACKGROUND

[0002] Satellites have long on-orbit time and complex space environment, and therefore have extremely high requirements for electronic equipment reliability. With the scaling of aerospace products and the extension of test cycles, a large amount of data is generated. How to efficiently extract key data of equipment performance degradation and improve processing efficiency to support reliability evaluation is a core problem.

[0003] In related technologies, for the reliability evaluation needs of satellite electronic equipment, the following two methods are mainly used: one is full sampling method (complete analysis of degradation performance data), and the other is conventional sampling method (such as equidistant sampling or time period average). However, both of them have low data processing efficiency and are difficult to adapt to the needs of batch production and rapid evaluation.

[0004] Therefore, the existing satellite electronic equipment reliability evaluation scheme has high data storage cost (full sampling occupies a large amount of storage), low reliability evaluation efficiency (slow data search, slow processing, high computing power), and low data sampling efficiency (waste of times in sparse areas and missing of key data in dense areas, which is difficult to accurately obtain the state of the equipment), which cannot meet the needs of high reliability evaluation. SUMMARY

[0005] The application aims to provide a satellite electronic equipment reliability evaluation method and device based on data sparse sampling, a medium and electronic equipment, which can solve at least one of the above technical problems. The specific scheme is as follows:

[0006] According to the specific embodiment of the application, in a first aspect, the application provides a satellite electronic equipment reliability evaluation method based on data sparse sampling, which comprises:

[0007] Obtaining a degradation model of a preset performance index of an electronic device, and constructing a reliability calculation model based on the degradation model; wherein the degradation model is used to indicate the degradation law of the performance data of the electronic device with time; and the reliability calculation model is used to calculate the reliability of the electronic device;

[0008] In the case where the degradation model parameters of the degradation model have prior information, a data sparse sampling strategy based on Bayesian estimation is used to sample the performance data of the electronic device;

[0009] In the case where the degradation model parameters of the degradation model have no prior information, a data sparse sampling strategy based on adaptive maximum likelihood estimation is used to sample the performance data of the electronic device;

[0010] Based on the sampling result and the reliability calculation model, a reliability evaluation result of the electronic device is obtained.

[0011] In some possible embodiments, the data sparse sampling strategy based on Bayesian estimation performs sampling on the performance data of the electronic device, including:

[0012] Based on prior information of a first degradation model parameter of the degradation model and first performance data of the electronic device, a posterior distribution of the first degradation model parameter is obtained.

[0013] Based on the posterior distribution and the first performance data, a reliability parameter estimation value of a preset candidate time point is generated.

[0014] Based on the reliability parameter estimation value and a first constraint condition of a to-be-solved reliability parameter estimation result, a first sparse sampling node is determined.

[0015] Based on the first sparse sampling node, sampling is performed on the performance data of the electronic device.

[0016] In some possible embodiments, the first constraint condition is that an estimation variance of the to-be-solved reliability parameter is minimum or not greater than a first preset value.

[0017] The determination of the first sparse sampling node based on the reliability parameter estimation value and the first constraint condition of the to-be-solved reliability parameter estimation result includes:

[0018] Based on the posterior distribution of the first degradation model parameter and the first performance data, a plurality of performance data samples of a candidate time point are generated.

[0019] Based on the performance data samples and the posterior distribution of the first degradation model parameter, a to-be-solved reliability parameter estimation value corresponding to each performance data sample is obtained.

[0020] Based on the to-be-solved reliability parameter estimation value, a reliability parameter estimation variance of the candidate time point is solved.

[0021] Based on the reliability parameter estimation variance and the first constraint condition, a sampling node for performing data sparse sampling is determined.

[0022] In some possible embodiments, the generation of the plurality of performance data samples of the candidate time point based on the posterior distribution of the first degradation model parameter and the first performance data includes:

[0023] Based on the posterior distribution of the first degradation model parameter, M first degradation model parameter samples are extracted by using a random sampling method.

[0024] obtain performance data conditional distribution of the candidate time point based on the first degradation model parameter sample, the first performance data and the degradation model; wherein each conditional distribution corresponds to a first degradation model parameter sample;

[0025] extract N performance data samples from the performance data conditional distribution corresponding to each candidate time point.

[0026] In some possible embodiments, the data sparse sampling strategy based on adaptive maximum likelihood estimation performs sampling on the performance data of the electronic device, including:

[0027] obtain the maximum likelihood estimation value of the second degradation model parameter of the degradation model based on the second performance data of the electronic device;

[0028] obtain the to-be-solved reliability parameter estimation value corresponding to the preset candidate time point based on the maximum likelihood estimation value and the reliability calculation model;

[0029] determine a second sparse sampling node based on the reliability parameter estimation value and the second constraint condition of the to-be-solved reliability parameter estimation result;

[0030] perform sampling on the performance data of the electronic device based on the second sparse sampling node.

[0031] In some possible embodiments, the obtaining the maximum likelihood estimation value of the second degradation model parameter of the degradation model based on the second performance data of the electronic device includes:

[0032] calculate the degradation rate corresponding to each time series data point based on the second performance data of the electronic device;

[0033] if the degradation rate of the time series data point is less than a preset rate threshold, construct a likelihood function item with a first normal distribution as an error distribution;

[0034] if the degradation rate of the data point is greater than or equal to the preset rate threshold, construct a likelihood function item with a second normal distribution as an error distribution; wherein the variance of the second error distribution is greater than the variance of the first error distribution;

[0035] construct a segmented likelihood function of the second degradation parameter of the degradation model based on the error distribution likelihood function items corresponding to different time series data points;

[0036] obtain the maximum likelihood estimation value of the second degradation parameter of the degradation model based on the segmented likelihood function and the second performance data.

[0037] In some possible embodiments, the method further includes adaptive noise filtering preprocessing of the performance data, the adaptive noise filtering preprocessing including:

[0038] Adaptive noise filtering preprocessing of the performance data by using a Kalman filter based on time sequence fluctuation characteristics of the performance data;

[0039] Wherein, the process noise covariance matrix and the measurement noise covariance matrix of the Kalman filter are dynamically adjusted based on the time sequence fluctuation characteristics of the performance data.

[0040] According to the specific embodiments of the present application, in a second aspect, the present application further provides a satellite-based electronic device reliability evaluation device based on data sparse sampling, the device comprising:

[0041] A model construction unit is configured to acquire a degradation model of a preset performance index of an electronic device, and construct a reliability calculation model based on the degradation model; wherein the degradation model is used to indicate the degradation law of performance data of the electronic device over time; and the reliability calculation model is used to calculate the reliability of the electronic device.

[0042] A first sampling unit is configured to, in the case that there is prior information of degradation model parameters of the degradation model, perform sampling on the performance data of the electronic device based on a data sparse sampling strategy of Bayesian estimation.

[0043] A second sampling unit is configured to, in the case that there is no prior information of degradation model parameters of the degradation model, perform sampling on the performance data of the electronic device based on a data sparse sampling strategy of adaptive maximum likelihood estimation.

[0044] A result output unit is configured to obtain a reliability evaluation result of the electronic device based on a sampling result and the reliability calculation model.

[0045] According to the specific embodiments of the present application, in a third aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the method of any one of the above.

[0046] According to the specific embodiments of the present application, in a fourth aspect, the present application further provides an electronic device, comprising: one or more processors; a storage device configured to store one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of the above.

[0047] The above scheme of the embodiments of the present application has at least the following beneficial effects compared with the prior art:

[0048] According to whether there is prior information of the degradation model parameters, sampling strategies are selected in different cases. For sampling with prior information, Bayesian estimation is used, and for sampling without prior information, adaptive maximum likelihood estimation is used. In this way, when there is prior information, the key data reflecting the degradation of electronic equipment can be accurately found with the help of the prior information, and when there is no prior information, effective information can also be automatically captured, without wasting sampling times in data sparse places and missing key data in data dense places as in conventional sampling, so that the efficiency of sampling can be greatly improved.

[0049] Sampling data sparsely can obviously reduce the amount of data to be processed without the need to collect and analyze all performance data of electronic equipment. Therefore, the occupation cost of storage resources can be reduced, and the problems of slow search, slow processing, and high requirement for computing power caused by too much data can be avoided, and the data processing efficiency is naturally improved.

[0050] First, a reliability calculation model is constructed according to the degradation model, and then the reliability evaluation result is directly obtained combined with the sampling result. Since the degradation model describes the degradation law of performance data of electronic equipment over time, this law can help to lock the key data and reduce the amount of invalid data processing. With less data, the pressure of subsequent storage and calculation is naturally reduced, and the processing efficiency is improved. Based on the degradation model, the reliability calculation model is constructed, the correlation logic between performance degradation and reliability is connected, the calculation process is simplified, and the problem of low processing efficiency in the prior art is solved. BRIEF DESCRIPTION OF DRAWINGS

[0051] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the specification, serve to explain the principles of the application. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained from these drawings without creative labor for those skilled in the art. In the drawings:

[0052] Figure 1 Flowchart of the satellite-based electronic equipment reliability evaluation method provided by the present application based on data sparse sampling Figure 1 ;

[0053] Figure 2 Flowchart of the satellite-based electronic equipment reliability evaluation method provided by the present application based on data sparse sampling Figure 2 ;

[0054] Figure 3 Flowchart of the satellite-based electronic equipment reliability evaluation method provided by the present application based on data sparse sampling Figure 3 ;

[0055] Figure 4A schematic diagram of a satellite-based electronic device reliability evaluation apparatus based on data sparse sampling is provided in the present application;

[0056] Figure 5 An electronic device structure schematic diagram is shown in the present application embodiment. DETAILED DESCRIPTION

[0057] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0058] The terms used in the embodiments of the present application are only for the purpose of describing the specific embodiments, and are not intended to limit the present application. The singular form of one, the and the used in the embodiments of the present application and the appended claims are intended to include the plural form, unless the context clearly indicates otherwise, and the plurality generally contains at least two.

[0059] It should be understood that the terms and / or used herein are only to describe the associated relationship between the associated objects, and it is indicated that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character / in this paper generally represents the relationship of one or the other of the associated objects before and after.

[0060] It should be understood that although the terms first, second, third, etc. may be used in the embodiments of the present application, these descriptions should not be limited to these terms. These terms are only used to distinguish the description. For example, without departing from the scope of the embodiments of the present application, the first can also be called the second, and similarly, the second can also be called the first.

[0061] It should also be noted that the term includes, includes or any other variant thereof is intended to cover non-exclusive inclusion, so that the goods or devices including a series of elements not only include those elements, but also include other elements not explicitly listed, or also include elements inherent to such goods or devices. Without more limitation, the element limited by the sentence including one does not exclude the presence of another identical element in the goods or devices including the element.

[0062] Satellite electronic equipment is various, covering power system, communication system, control system, data processing system and other key fields, which plays an important role in the normal operation of satellite. Due to the long time of satellite in-orbit operation, the electronic equipment needs to work stably in complex space environment for a long time, so the reliability requirement of the equipment is very high. Therefore, long time life test or degradation test and a large number of reliability screening test need to be carried out during the development stage of satellite electronic equipment, and the key performance parameters of satellite electronic equipment are monitored during the test, so as to evaluate the life and reliability of satellite electronic equipment. With the large-scale and batch production of aerospace products and long time test monitoring, a large amount of data will be generated. How to efficiently and quickly extract the key data reflecting the performance degradation characteristics of satellite electronic equipment from the data to improve the data processing efficiency and support the reliability evaluation of satellite electronic equipment has become a key problem to be solved.

[0063] In view of the above key problems, two methods are usually adopted at present: one is the full sampling method, that is, the performance data with degradation characteristics are directly analyzed to carry out reliability and life evaluation; the other is the conventional sampling method, such as equal interval sampling (extracting one data point every hour or day) or calculating the average value in this period, and then carrying out reliability and life evaluation based on the sampled data. At present, the data processing efficiency of the two methods is not high, which is difficult to meet the actual needs of batch production and rapid reliability evaluation.

[0064] The disadvantages of the existing satellite electronic equipment reliability evaluation technology mainly include high data storage cost, low reliability evaluation efficiency, low data sampling efficiency, etc. Specifically, in the face of massive electronic equipment monitoring data, if the full sampling method is adopted, the performance data with degradation characteristics are all sampled and analyzed, which will occupy a large amount of storage resources, increase the storage cost, and due to the large amount of data, it may cause inconvenience in data searching, slow data processing speed, high demand for data computing power, and finally low reliability evaluation efficiency; if the conventional sampling method (such as equal interval sampling) is adopted, a large number of sampling times will be wasted in the data sparse interval, while some key data will be missed in the data intensive interval, which cannot effectively and accurately collect the information reflecting the state of satellite electronic equipment, so the sampling efficiency needs to be improved, which cannot meet the analysis needs of satellite electronic equipment reliability evaluation.

[0065] In addition, satellite electronic equipment itself faces stringent life and reliability requirements. In order to ensure that the satellite electronic equipment can be stably operated in orbit for a long time, some equipment needs to carry out long-time life test or degradation test during the development and design stage, and the key parameters of the equipment are continuously monitored during the test process, so as to accurately evaluate the state. However, due to the large amount of data generated by test monitoring, the cost of data storage is high, it is difficult to quickly and accurately extract the key data for reliability evaluation from the data, and the efficiency of reliability evaluation is low.

[0066] The optional embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0067] Figure 1 The flowchart of the construction method of the sample data set provided by the present application is shown in Figure Figure 1 As shown in the figure, the satellite electronic equipment reliability evaluation method based on data sparse sampling provided by the present embodiment comprises:

[0068] S101, obtaining a degradation model of a preset performance index of an electronic equipment, and constructing a reliability calculation model based on the degradation model; wherein the degradation model is used to indicate the degradation law of the performance data of the electronic equipment with time; and the reliability calculation model is used to calculate the reliability of the electronic equipment;

[0069] It can be understood that the performance index refers to a quantifiable key characteristic parameter that the electronic equipment must meet to realize the specified function (such as power supply, communication, signal processing), which is the basic standard for measuring whether the equipment is working normally or not and whether it is degraded. The performance index of different electronic equipment is different, for example, the output voltage (such as the rated value 5V, which needs to be stabilized at 4.8-5.2V) and the ripple coefficient (such as ≤0.1%) of the power module of the satellite; the signal transmission power (such as the rated value 10W, which needs to be ≥9W) and the bit error rate (such as ≤10 -6 ) of the communication module of the satellite.

[0070] The degradation model is a mathematical model constructed based on the time sequence monitoring data of the performance index, which is used to quantitatively describe the law that the performance index gradually deviates from the rated value (i.e. degradation) with time, and its function is to convert the physical process of performance degradation into a calculable mathematical expression. Common model types include linear model, exponential model, Weibull model, etc. Through the degradation model, the performance index value at any time t in the future can be predicted (such as predicting that the on-resistance will reach 12Ω after 100 days), which provides a performance prediction basis for subsequent reliability calculation.

[0071] The reliability calculation model is a mathematical model for calculating the probability of electronic equipment not failing (i.e., reliable) within a specified time, which is based on the prediction results of the degradation model and combined with the failure threshold of the electronic equipment. The core is to convert whether the performance index meets the standard into a reliability value. The failure threshold is the critical value of the performance index, and the equipment cannot work normally if it exceeds the threshold. The reliability calculation model is an output tool for evaluating results, which can convert the abstract performance degradation trend into intuitive reliability values (such as 90%, 50%).

[0072] S102, in the case that the degradation model parameter of the degradation model has prior information, based on the data sparse sampling strategy of Bayesian estimation, sampling is performed on the performance data of the electronic equipment;

[0073] The prior information refers to the effective information obtained before sampling the performance data of the current electronic equipment, which is directly related to the degradation model parameter of the degradation model. Its role is to quantify the initial uncertainty of the degradation model parameter, and provide the basis for the initial probability distribution of the parameter for subsequent Bayesian estimation.

[0074] The specific source of the prior information can be the historical test / in-orbit data of the same type of electronic equipment, i.e., other satellite equipment with the same model, consistent functions and similar working conditions as the current electronic equipment, ground accelerated degradation test data in the research and development stage, or performance monitoring data of the equipment already in orbit. For example, the current equipment is a satellite power module, and the degradation model parameter is the output voltage degradation rate. The prior information can come from the accelerated test data of multiple power modules of the same batch in a high-temperature environment on the ground. By statistically analyzing these data, the historical distribution of the output voltage degradation rate is obtained. The specific source of the prior information can also be experts in the field of satellite electronic equipment reliability, who evaluate the reasonable value range of the degradation model parameter based on long-term engineering experience, and form quantitative information through comparison and verification with historical data.

[0075] S103, in the case that the degradation model parameter of the degradation model has no prior information, based on the data sparse sampling strategy of adaptive maximum likelihood estimation, sampling is performed on the performance data of the electronic equipment;

[0076] When the degradation model parameter of the degradation model has no known prior information, the data sparse sampling strategy of adaptive maximum likelihood estimation is used to collect the performance data of the electronic equipment. The core logic of maximum likelihood estimation is to find the model parameter that can produce the observation data with the maximum probability only through the current observation data, without relying on any prior information. In the absence of prior information, the parameter uncertainty cannot be initialized by Bayesian estimation (which requires prior distribution), while maximum likelihood estimation can directly estimate the degradation model parameter only with performance data, avoiding the dilemma that the parameter cannot be estimated without prior information.

[0077] S104, obtain the reliability evaluation result of the electronic device based on the sampling result and the reliability calculation model.

[0078] The sampling result refers to the performance data of the key node collected through the sparse sampling strategy (Bayesian estimation or adaptive maximum likelihood estimation) described above. The sampling node is the most critical moment (such as the time point of the rapid performance degradation stage, the key node of the satellite task) for reliability evaluation after screening, rather than redundant data sampled randomly.

[0079] The reliability calculation model is a performance-reliability conversion tool constructed in advance based on the degradation model, which converts the performance degradation state reflected by the sampling result into a probabilistic reliability value.

[0080] In this embodiment, according to whether the degradation model parameters have prior information, the sampling strategy is selected according to the situation. For sampling with prior information, Bayesian estimation is used, and for sampling without prior information, adaptive maximum likelihood estimation is used. In this way, when there is prior information, the key data reflecting the degradation of the electronic device can be accurately found with the help of the prior information, and when there is no prior information, effective information can also be automatically captured, without wasting sampling times in data sparse places and missing key data in data intensive places as in conventional sampling, thereby greatly improving the efficiency of sampling.

[0081] Sparse sampling of sampling data does not require all performance data of the electronic device to be collected and analyzed, which can significantly reduce the amount of data to be processed. Thus, the storage resource occupation cost can be reduced, and the problems of slow search, slow processing, and high computing power requirement caused by too much data can be avoided, and the data processing efficiency is naturally improved.

[0082] First, the reliability calculation model is constructed according to the degradation model, and then the reliability evaluation result is directly obtained in combination with the sampling result. Since the degradation model describes the degradation law of the performance data of the electronic device over time, this law can help to lock the key data and reduce the amount of invalid data processing. With less data, the subsequent storage and computing pressure is naturally reduced, and the processing efficiency is improved. Based on the degradation model, the reliability calculation model is constructed, the correlation logic between performance degradation and reliability is connected, the calculation process is simplified, and thus the problem of low processing efficiency in the prior art is solved.

[0083] Figure 2 The construction method of the sample data set provided in the present application Figure 2 As shown in Figure 2 , the electronic device reliability evaluation method of the satellite based on data sparse sampling provided in the present embodiment comprises:

[0084] S201, based on the time sequence fluctuation characteristics of the performance data, the performance data is adaptively noise filtered and preprocessed by using a Kalman filter;

[0085] wherein, based on the time sequence fluctuation characteristics of the performance data, the process noise covariance matrix and the measurement noise covariance matrix of the Kalman filter are dynamically adjusted.

[0086] It can be understood that the Kalman filter is a commonly used time sequence data processing tool, which can extract the real change rule of the data in the presence of noise interference (such as data fluctuation caused by measurement error and environmental interference).

[0087] The performance data (such as on-off resistance and working current) of the satellite electronic device is susceptible to sudden noise caused by electromagnetic interference and measurement error when collected in orbit, and the data has dynamic characteristics of uneven time sequence fluctuation (such as small fluctuation in the stable stage and large jump when disturbed). The Kalman filter with fixed parameters is difficult to balance noise suppression and real degradation trend reservation, so the adaptive noise filtering preprocessing is needed to solve this problem.

[0088] The specific scheme of the adaptive noise filtering preprocessing is: based on the time sequence fluctuation characteristics of the performance data (that is, the standard deviation of the difference value of the performance data at adjacent time points), the process noise covariance matrix and the measurement noise covariance matrix of the Kalman filter are dynamically adjusted.

[0089] For example, the fluctuation threshold is set to 0.05 times the rated value of the performance indicator, and the specific adjustment logic is as follows:

[0090] When the standard deviation of the difference value of the performance data at adjacent time points is greater than 0.05 times the rated value of the performance indicator (the data fluctuation is large), the weight of the measurement noise covariance matrix is increased to effectively suppress the sudden noise interference;

[0091] When the standard deviation of the difference value of the performance data at adjacent time points is less than or equal to 0.05 times the rated value of the performance indicator (the data is in a stable state), the weight of the process noise covariance matrix is reduced to better preserve the real degradation trend of the data.

[0092] S202, based on the prior information of the first degradation model parameter of the degradation model and the first performance data of the electronic device, the posterior distribution of the first degradation model parameter is obtained;

[0093] It can be understood that the prior information (such as the initial judgment of the key parameters of the device degradation obtained by the previous small batch test and the same type of device) and the first performance data after filtering by S201 can reflect the effective data of the real degradation of the device); Using these two information combined with the Bayesian method, the initial parameter judgment (prior) is corrected to a parameter value range that is more consistent with the actual situation (that is, the posterior distribution, which is not a fixed value, but a parameter probability interval); The corrected parameter range is an important basis for calculating the reliability of the device, which can make the subsequent evaluation more accurate.

[0094] S203, generating a reliability parameter estimation value of the preset candidate time point based on the posterior distribution and the first performance data;

[0095] It can be understood that the approximate values of the reliability related parameters of several key time points are calculated in advance by using the posterior distribution in combination with the first performance data. Specifically, the candidate time point is a key time point (such as 100 days in orbit, 200 days in orbit, or a stage in which performance may rapidly degrade) that needs to be evaluated in advance to evaluate the reliability of the equipment; the approximate estimation value of the reliability parameter at each candidate time point is calculated by using the posterior distribution (the accurate range of the key parameters of the equipment degradation) and the first performance data and substituting into the reliability related calculation logic (such as the reliability of 100 days in orbit is about 98% and the reliability of 200 days in orbit is about 95%). This is done to understand the reliability at different time points in advance, to lay a foundation for selecting the key time points of the data in the subsequent sampling, and to avoid the subsequent sampling without direction.

[0096] S204, determining a first sparse sampling node based on the reliability parameter estimation value and a first constraint condition of an estimation result of the to-be-solved reliability parameter;

[0097] It can be understood that the key time points that need to be sampled are selected by using the calculated approximate values of the reliability at each time point in combination with the first constraint condition during the evaluation. The purpose of this is to sample only at necessary time points, which not only wastes resources but also ensures that the evaluation result meets the requirements.

[0098] The reliability parameter estimation value is the reliability at each candidate time point calculated by S203 (such as a time point with large reliability fluctuation or a time point close to the failure threshold); the first constraint condition is a hard requirement for evaluation. In combination with the two, the time points that are most critical to meet the constraint condition are selected from the candidate time points (such as a time point with large reliability fluctuation must be sampled and a time point with stable and small error can not be sampled). These points are the first sparse sampling node.

[0099] In some embodiments, the first constraint condition is that the estimation variance of the to-be-solved reliability parameter is minimum or not greater than a first preset value;

[0100] Determining the first sparse sampling node based on the reliability parameter estimation value and the first constraint condition of the estimation result of the to-be-solved reliability parameter includes:

[0101] Generating a plurality of performance data samples of the candidate time point based on the posterior distribution of the first degradation model parameter and the first performance data;

[0102] Obtaining a to-be-solved reliability parameter estimation value corresponding to each performance data sample based on the performance data sample and the posterior distribution of the first degradation model parameter;

[0103] Solve the reliability parameter estimation variance of the candidate time point based on the to-be-solved reliability parameter estimation value;

[0104] Determine the sampling node of performing data sparse sampling based on the reliability parameter estimation variance and the first constraint condition.

[0105] It can be understood that the performance data of the candidate time point is simulated first. The candidate time point is a time point that may be sampled (such as 10 days, 20 days in the future). In combination with the existing information (posterior distribution of the first degradation model parameter, the first performance data that has been sampled), a plurality of performance data that may occur at these candidate time points (such as simulating 100 groups of performance values at each candidate time point, covering the possibility of data fluctuation) are simulated. Then, the reliability estimation value corresponding to each simulated data is calculated. Each group of performance data simulated in the first step is substituted into the calculation logic to obtain the reliability estimation value corresponding to different simulated data at each candidate time point (such as 100 reliability estimation values corresponding to 100 groups of performance data at the candidate time point of 10 days). Then, the reliability estimation variance of the candidate time point is calculated. The reliability estimation variance can reflect the fluctuation size of the reliability estimation value, wherein a small variance indicates that the reliability estimation is accurate, and a large variance indicates that the estimation is inaccurate. The plurality of reliability estimation values of the same candidate time point are used to calculate the estimation variance of this time point (each candidate time point corresponds to a variance). Finally, the sampling node is selected according to the variance standard. According to the first constraint condition (either the time point with the smallest variance is selected, or the time point with a variance not exceeding a preset value is selected), the candidate time points are screened: only the time points that meet the variance standard are determined as the nodes that are actually sampled, so that the reliability estimation accuracy is ensured, and redundant data is not sampled.

[0106] In some embodiments, based on the posterior distribution of the first degradation model parameter and the first performance data, a plurality of performance data samples of the candidate time point are generated, comprising:

[0107] Based on the posterior distribution of the first degradation model parameter, M first degradation model parameter samples are extracted by using a random sampling method;

[0108] Based on the first degradation model parameter sample, the first performance data, and the degradation model, a performance data conditional distribution of the candidate time point is obtained; wherein each conditional distribution corresponds to a first degradation model parameter sample;

[0109] From the performance data conditional distribution corresponding to each candidate time point, N performance data samples are extracted.

[0110] S205, based on the first sparse sampling node, performing sampling on the performance data of the electronic device.

[0111] S206, based on the sampling result and the reliability calculation model, obtaining the reliability evaluation result of the electronic device.

[0112] In this embodiment, based on the time sequence fluctuation characteristics of the performance data, adaptive filtering is performed by dynamically adjusting the process noise covariance matrix and the measurement noise covariance matrix of the Kalman filter, effectively solving the problem that the performance data of the satellite electronic device is affected by electromagnetic interference and measurement error. Both the sudden noise is suppressed when the data fluctuates greatly, and the real degradation trend is retained in the stable data stage, providing accurate data support for subsequent parameter estimation and reliability calculation; then, combining the prior information of the degradation model parameters (such as initial judgment obtained from small batch test and similar device data) and the first performance data after filtering, the initial parameter prior judgment is corrected to a posterior distribution (not a fixed value, but a parameter value range with high probability) more close to the actual situation through the Bayesian method, greatly improving the accuracy of the degradation model parameter value, and laying a scientific foundation for reliability evaluation; on this basis, the reliability parameter estimation value at the preset candidate time (such as the key time point in orbit and the stage where the performance may rapidly degrade) is generated in advance, the general situation of reliability at different time points is clearly understood, the direction for subsequent sampling node selection is clear, and blind sampling is avoided; then, combining the reliability parameter estimation value and the first constraint condition (such as minimum estimation variance or not exceeding the first preset value) of the to-be-solved reliability parameter estimation result, the performance data samples at the candidate time are simulated, the reliability estimation value and the variance corresponding to each sample are calculated, and the first sparse sampling node which is most critical to meet the constraint condition is screened out, and sampling is only performed at necessary time, which not only avoids the waste of satellite data transmission, device energy consumption and ground monitoring resources, but also ensures the reliability evaluation accuracy; finally, the evaluation result is obtained based on the sampling result and the reliability calculation model.

[0113] Figure 3 Flowchart of the construction method of the sample data set provided in the present application Figure 3 As shown in Figure 4 , the satellite electronic device reliability evaluation method based on data sparse sampling provided in this embodiment comprises:

[0114] S301, based on the time sequence fluctuation characteristics of the performance data, the performance data is adaptively noise filtered and preprocessed by using a Kalman filter;

[0115] Exemplarily, this step participates in the above step S201, and will not be repeated.

[0116] S302, based on the second performance data of the electronic device, the maximum likelihood estimation value of the second degradation model parameter of the degradation model is obtained;

[0117] The second performance data is performance data (e.g., resistance value, battery capacity, etc.) obtained by the electronic device in observation, and is an actual observation result reflecting the current degradation state of the device. The core logic of the maximum likelihood estimate is to select a set of parameter values from all possible parameter values, which has the maximum probability of observing the current set of second performance data. For example, assuming that the degradation model is performance = A - B x time (A is the initial value, and B is the degradation rate), and there are three second data (e.g., performance of 80 at 100 hours, 75 at 200 hours, and 70 at 300 hours). The maximum likelihood estimate is to find the specific values of A and B, so that the performance values calculated by substituting the set of A and B into the model are closest to the observed 80, 75, and 70, and the probability of this occurrence is the maximum. The set of A and B is the maximum likelihood estimate value.

[0118] In some embodiments, obtaining the maximum likelihood estimate value of the second degradation model parameter of the degradation model based on the second performance data of the electronic device includes:

[0119] Based on the second performance data of the electronic device, the degradation rate corresponding to each time series data point is calculated.

[0120] If the degradation rate of the time series data point is less than a preset rate threshold, a likelihood function item with a first normal distribution as the error distribution is constructed.

[0121] If the degradation rate of the data point is greater than or equal to the preset rate threshold, a likelihood function item with a second normal distribution as the error distribution is constructed; and the variance of the second normal distribution is greater than the variance of the first normal distribution.

[0122] Based on the error distribution likelihood function items corresponding to different time series data points, a segmented likelihood function of the second degradation parameter of the degradation model is constructed.

[0123] Based on the segmented likelihood function and the second performance data, the maximum likelihood estimate value of the second degradation parameter of the degradation model is obtained.

[0124] It can be understood that the degradation rate corresponding to each time point is calculated first, such as the performance of 90 at the 100th hour and 85 at the 200th hour, and the degradation rate between the two points is "(90-85) / (200-100) = 0.05", and the degradation rate of each time series data point is calculated. According to the degradation speed, two cases are handled to advance a preset rate threshold (such as 0.03) to distinguish between slow degradation and fast degradation: if the degradation rate of a certain point is < threshold (slow degradation): data fluctuation is usually small (such as performance measurement value is stable during slow degradation), and the first normal distribution with small variance is used to describe the error between data and model prediction value; if the rate is ≥ threshold (fast degradation): data fluctuation is easy to become large (such as performance is high and low at different times when it is fast), and the second normal distribution with large variance is used to describe the error. For each data point, a probability small formula (likelihood function term) is built. For each time series data point, according to whether it is slow degradation or fast degradation, the corresponding normal distribution (first or second) is selected to build a small formula describing how large the probability of observing this data is, which is the likelihood function term. The small formula is spliced to form a segmented total formula (segmented likelihood function). The likelihood function terms of all data points are spliced together to form a whole function.

[0125] S303, based on the maximum likelihood estimate value and the reliability calculation model, obtaining the preset candidate time corresponding to the to-be-solved reliability parameter estimate value;

[0126] It can be understood that first, it is determined whether the device is fast or slow at each time point; when slow degradation, data fluctuation is small (small error standard is used); when fast degradation, data fluctuation is large (large error standard is used), and error rules are determined for each data point according to "fast and slow"; from all possible model parameters, a set of parameters that "use this set of parameters + corresponding error rules, and can best match the newly measured data" is selected, which is the estimated value to be found.

[0127] S304, based on the reliability parameter estimate value and the second constraint condition of the to-be-solved reliability parameter estimate result, determining a second sparse sampling node;

[0128] S305, based on the second sparse sampling node, performing sampling on the performance data of the electronic device.

[0129] S306, based on the sampling result and the reliability calculation model, obtaining the reliability evaluation result of the electronic device.

[0130] In this embodiment, when calculating the degradation model parameters based on the second performance data, the error distribution is constructed according to the comparison between the degradation rate of the time sequence data point and the preset threshold in two scenarios of slow degradation and fast degradation. In the slow degradation stage, the data fluctuation is small, and the first normal distribution with smaller variance is adopted. In the fast degradation stage, the data fluctuation is large, and the second normal distribution with larger variance is adopted. Then, the maximum likelihood estimation value of the parameters is derived through the piecewise likelihood function. This parameter calculation method that fits the real degradation law of the device can improve the parameter estimation accuracy. When determining the sampling nodes, instead of blindly selecting, the current reliability parameter estimation value and the second constraint condition (such as the sampling number, cost limit, etc.) of the to-be-sought reliability parameter estimation result are combined to only select the nodes that are critical to improving the reliability evaluation accuracy as the second sparse sampling nodes, thereby reducing unnecessary sampling operations and saving the resource cost of satellite data transmission, device energy consumption and ground monitoring. The whole process forms a coherent closed loop from data preprocessing, parameter estimation, sampling optimization to the final reliability evaluation, each step relies on the actual performance data and degradation characteristics of the electronic device, and the finally output reliability evaluation result can be practically used to judge the stable working time of the satellite electronic device.

[0131] The application also provides a device embodiment for implementing the method steps of the above embodiments, based on the same explanation of the same name meaning and the same technical effects as the above embodiments, which will not be described here.

[0132] As shown in Figure 5 A satellite electronic device reliability evaluation device based on data sparse sampling, the device comprises:

[0133] A model construction unit 401 is configured to obtain a degradation model of a preset performance index of an electronic device, and construct a reliability calculation model based on the degradation model. The degradation model is used to indicate the degradation law of the performance data of the electronic device over time. The reliability calculation model is used to calculate the reliability of the electronic device.

[0134] A first sampling unit 402 is configured to, in the case that the degradation model parameters of the degradation model have prior information, perform sampling on the performance data of the electronic device based on a data sparse sampling strategy of Bayesian estimation.

[0135] A second sampling unit 403 is configured to, in the case that the degradation model parameters of the degradation model do not have prior information, perform sampling on the performance data of the electronic device based on a data sparse sampling strategy of adaptive maximum likelihood estimation.

[0136] A result output unit 404 is configured to obtain a reliability evaluation result of the electronic device based on the sampling result and the reliability calculation model.

[0137] In some embodiments, the first sampling unit 402 is specifically configured to:

[0138] obtain a posterior distribution of the first degradation model parameter based on prior information of the first degradation model parameter and first performance data of the electronic device;

[0139] generate a reliability parameter estimation value of a preset candidate time point based on the posterior distribution and the first performance data;

[0140] determine a first sparse sampling node based on the reliability parameter estimation value and a first constraint condition of a to-be-solved reliability parameter estimation result;

[0141] perform sampling on performance data of the electronic device based on the first sparse sampling node.

[0142] In some embodiments, the first constraint condition is that an estimation variance of the to-be-solved reliability parameter is minimum or not greater than a first preset value; the first sampling unit 402 is specifically configured to:

[0143] generate a plurality of performance data samples of the candidate time point based on the posterior distribution of the first degradation model parameter and the first performance data;

[0144] obtain a to-be-solved reliability parameter estimation value corresponding to each performance data sample based on the performance data sample and the posterior distribution of the first degradation model parameter;

[0145] obtain a reliability parameter estimation variance of the candidate time point based on the to-be-solved reliability parameter estimation value;

[0146] determine a sampling node of performing data sparse sampling based on the reliability parameter estimation variance and the first constraint condition.

[0147] In some embodiments, the first sampling unit 402 is specifically configured to:

[0148] extract M first degradation model parameter samples based on the posterior distribution of the first degradation model parameter by using a random sampling method;

[0149] obtain a performance data conditional distribution of the candidate time point based on the first degradation model parameter sample, the first performance data, and the degradation model; wherein each conditional distribution corresponds to a first degradation model parameter sample;

[0150] extract N performance data samples from the performance data conditional distribution corresponding to each candidate time point.

[0151] In some embodiments, the second sampling unit 403 is specifically configured to:

[0152] obtain a maximum likelihood estimation value of a second degradation model parameter of the degradation model based on second performance data of the electronic device;

[0153] The preset candidate time corresponding to the to-be-solved reliability parameter estimation value is obtained based on a maximum likelihood estimation value and a reliability calculation model;

[0154] The second sparse sampling node is determined based on the reliability parameter estimation value and a second constraint condition of the to-be-solved reliability parameter estimation result.

[0155] The performance data of the electronic device is sampled based on the second sparse sampling node.

[0156] In some embodiments, the second sampling unit 403 is specifically configured to:

[0157] The degradation rate corresponding to each time series data point is calculated based on the second performance data of the electronic device.

[0158] If the degradation rate of the time series data point is less than a preset rate threshold, a likelihood function item with a first normal distribution as an error distribution is constructed.

[0159] If the degradation rate of the data point is greater than or equal to the preset rate threshold, a likelihood function item with a second normal distribution as an error distribution is constructed; wherein the variance of the second normal distribution is greater than the variance of the first normal distribution.

[0160] The segmented likelihood function of the second degradation parameter of the degradation model is constructed based on the error distribution likelihood function items corresponding to different time series data points.

[0161] The maximum likelihood estimation value of the second degradation parameter of the degradation model is obtained based on the segmented likelihood function and the second performance data.

[0162] In some embodiments, the first sampling unit 402 and the second sampling unit 403 are specifically configured to:

[0163] Adaptive noise filtering preprocessing is performed on the performance data based on the time series fluctuation characteristics of the performance data using a Kalman filter.

[0164] The process noise covariance matrix and the measurement noise covariance matrix of the Kalman filter are dynamically adjusted based on the time series fluctuation characteristics of the performance data.

[0165] As shown in Figure 5 The embodiment provides an electronic device, which comprises at least one processor and a memory in communication connection with the at least one processor; the memory stores instructions executable by the processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method steps of the above embodiment.

[0166] The embodiment provides a non-volatile computer storage medium, which stores computer executable instructions; the computer executable instructions are executable to perform the method steps of the above embodiment.

[0167] The following description refers to Figure 5 which shows a structural schematic of an electronic device suitable for implementing embodiments of the present application. The terminal device in embodiments of the present application can include, but is not limited to, mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet PCs), PMPs (Portable Multimedia Players), car terminals (e.g., car navigation terminals), and the like, as well as fixed terminals such as digital TVs, desktop computers, and the like. Figure 5 The electronic device shown is merely an example and should not impose any limitation on the functions and use range of embodiments of the present application.

[0168] As Figure 5 shown, the electronic device can include a processing device (e.g., a central processor, a graphic processor, etc.) 501 that can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 502 or programs loaded into a random access memory (RAM) 503 from a storage device 508. In the RAM 503, various programs and data required for the operation of the electronic device are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0169] In general, the following devices can be connected to the I / O interface 505: input devices 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, and the like; output devices 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, and the like; storage devices 508 including, for example, a magnetic tape, a hard disk, and the like; and communication devices 509. The communication devices 509 can allow the electronic device to communicate wirelessly or wiredly with other devices to exchange data. Although ​ An electronic device having various devices is shown, but it is understood that all of the shown devices are not required to be implemented or provided. More or less devices can alternatively be implemented or provided.

[0170] In particular, according to embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, embodiments of the present application include a computer program product including a computer program carried on a computer readable medium, the computer program containing program codes for executing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network through the communication devices 509, or installed from the storage devices 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-described functions defined in the methods of embodiments of the present application are performed.

[0171] Note that the computer readable medium described above can be a computer readable signal medium or a computer readable storage medium or any combination thereof. The computer readable storage medium can be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or any suitable combination of the foregoing. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, the computer readable storage medium can be any tangible medium that contains or stores a program used by an instruction execution system, apparatus or device, and can be used by or in connection with the instruction execution system, apparatus or device. In the present application, the computer readable signal medium can include a computer readable program code carried in a baseband or as a part of a carrier wave, in which the computer readable program code is carried. Such a propagated computer readable signal medium can take various forms, including but not limited to, an electromagnetic signal, an optical signal or any suitable combination of the foregoing. The computer readable signal medium can also be any computer readable medium that is not a computer readable storage medium and that can be used to carry or transmit a program for use by or in connection with an instruction execution system, apparatus or device. The program code contained in the computer readable medium can be transmitted by any suitable medium, including but not limited to, wire, cable, RF (radio frequency), etc., or any suitable combination of the foregoing.

[0172] The computer readable medium described above can be included in the electronic device described above; or can exist separately from the electronic device and can be accessed via the electronic device.

[0173] Computer program code for carrying out operations of the present application can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0174] The flow diagrams and the block diagrams in the drawings are illustrations of architectures, functionalities, and operations of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flow diagrams or block diagrams can represent a module, a segment, or a portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block can occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and / or flow diagrams, and combinations of blocks in the block diagrams and / or flow diagrams, can be implemented by special purpose hardware-based systems that perform the specified functions or operations, or combinations of special purpose hardware and computer instructions.

[0175] The units described in the embodiments of the present disclosure can be implemented by software or by hardware. In some cases, the name of the unit does not constitute a limitation on the unit itself.

Claims

1. A method for evaluating reliability of an electronic device of a satellite based on data sparse sampling, characterized by, The method comprises: obtaining a degradation model of a preset performance index of an electronic device, and constructing a reliability calculation model based on the degradation model; wherein the degradation model is used to indicate the degradation law of performance data of the electronic device over time; and the reliability calculation model is used to calculate the reliability of the electronic device; in the case that the degradation model parameters of the degradation model have prior information, performing sampling on the performance data of the electronic device based on a data sparse sampling strategy of Bayesian estimation; in the case that the degradation model parameters of the degradation model do not have prior information, performing sampling on the performance data of the electronic device based on a data sparse sampling strategy of adaptive maximum likelihood estimation; obtaining a reliability evaluation result of the electronic device based on the sampling result and the reliability calculation model.

2. The method of claim 1, wherein, The data sparse sampling strategy based on Bayesian estimation performs sampling on the performance data of the electronic device, comprising: obtaining a posterior distribution of a first degradation model parameter of the degradation model based on prior information of the first degradation model parameter and first performance data of the electronic device; generating a reliability parameter estimation value of a preset candidate time based on the posterior distribution and the first performance data; determining a first sparse sampling node based on the reliability parameter estimation value and a first constraint condition of a to-be-solved reliability parameter estimation result; performing sampling on the performance data of the electronic device based on the first sparse sampling node.

3. The method of claim 2, wherein, The first constraint condition is that the estimation variance of the to-be-solved reliability parameter is minimum or not greater than a first preset value; The first constraint condition of the to-be-solved reliability parameter estimation result based on the reliability parameter estimation value, comprising: generating a plurality of performance data samples of a candidate time based on the posterior distribution of the first degradation model parameter and the first performance data; obtaining a to-be-solved reliability parameter estimation value corresponding to each performance data sample based on the performance data sample and the posterior distribution of the first degradation model parameter; solving the reliability parameter estimation variance of the candidate time based on the to-be-solved reliability parameter estimation value; determining a sampling node of data sparse sampling based on the reliability parameter estimation variance and the first constraint condition.

4. The method of claim 3, wherein, The posterior distribution of the first degradation model parameter and the first performance data, comprising: extracting M first degradation model parameter samples by using a random sampling method based on the posterior distribution of the first degradation model parameter; obtaining a performance data conditional distribution of a candidate time based on the first degradation model parameter sample, the first performance data and the degradation model; wherein each conditional distribution corresponds to a first degradation model parameter sample; extracting N performance data samples from the performance data conditional distribution corresponding to each candidate time.

5. The method of claim 1, wherein, The data sparse sampling strategy based on adaptive maximum likelihood estimation performs sampling on the performance data of the electronic device, comprising: obtaining a maximum likelihood estimation value of a second degradation model parameter of the degradation model based on second performance data of the electronic device; Based on the maximum likelihood estimate value and the reliability calculation model, an estimated value of a preset candidate time point corresponding to a to-be-solved reliability parameter is obtained; Based on the reliability parameter estimate value and a second constraint condition of a to-be-solved reliability parameter estimate result, a second sparse sampling node is determined; Based on the second sparse sampling node, sampling is performed on the performance data of the electronic device.

6. The method of claim 1, wherein, The maximum likelihood estimate value of the second degradation model parameter of the degradation model based on the second performance data of the electronic device includes: Based on the second performance data of the electronic device, a degradation rate corresponding to each time series data point is calculated; If the degradation rate of the time series data point is less than a preset rate threshold, a first likelihood function item with a normal distribution as an error distribution is constructed; If the degradation rate of the data point is greater than or equal to the preset rate threshold, a second likelihood function item with a second normal distribution as an error distribution is constructed; wherein the variance of the second error distribution is greater than the variance of the first error distribution; Based on the error distribution likelihood function items corresponding to different time series data points, a segmented likelihood function of the second degradation parameter of the degradation model is constructed; Based on the segmented likelihood function and the second performance data, the maximum likelihood estimate value of the second degradation parameter of the degradation model is obtained.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes adaptive noise filtering preprocessing of the performance data, and the adaptive noise filtering preprocessing includes: Based on the time series fluctuation characteristics of the performance data, a Kalman filter is used to perform adaptive noise filtering preprocessing on the performance data; Wherein, based on the time series fluctuation characteristics of the performance data, the process noise covariance matrix and the measurement noise covariance matrix of the Kalman filter are dynamically adjusted.

8. A satellite-based electronic device reliability evaluation apparatus based on data sparse sampling, characterized by, The device includes: A model construction unit is configured to obtain a degradation model of a preset performance indicator of an electronic device, and construct a reliability calculation model based on the degradation model; wherein the degradation model is used to indicate the degradation law of the performance data of the electronic device over time; and the reliability calculation model is used to calculate the reliability of the electronic device; A first sampling unit is configured to, in a case where the degradation model parameter of the degradation model has prior information, perform sampling on the performance data of the electronic device based on a data sparse sampling strategy of Bayesian estimation; A second sampling unit is configured to, in a case where the degradation model parameter of the degradation model does not have prior information, perform sampling on the performance data of the electronic device based on a data sparse sampling strategy of adaptive maximum likelihood estimation; A result output unit is configured to obtain a reliability evaluation result of the electronic device based on a sampling result and the reliability calculation model.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The program, when executed by a processor, implements the method of any one of claims 1 to 7.

10. An electronic device, comprising: Comprise: One or more processors; A storage device is configured to store one or more programs, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the method of any one of claims 1 to 7.

Images