CN-121997173-A - Quality and reliability evaluation method and system based on small sample data

CN121997173ACN 121997173 ACN121997173 ACN 121997173ACN-121997173-A

Abstract

The invention relates to the technical field of electronic information, in particular to a quality and reliability evaluation method and a system based on small sample data, which are implemented by acquiring reliability and quality evaluation data of an evaluation object; the method comprises the steps of calling a classical probability distribution life model based on a maximum likelihood estimation method to carry out reliability assessment on an assessment object to obtain a first fitting goodness test result, calling a multi-source information fusion reliability assessment model based on a BAYES method to carry out reliability assessment on the assessment object to obtain a second fitting goodness test result, comparing the first fitting goodness test result with the second fitting goodness test result, selecting a probability distribution model and a distribution parameter with high final fitting goodness as a reliability assessment mathematical model, extracting the reliability assessment result of the reliability assessment mathematical model, and carrying out quality analysis and assessment on the quality condition of the assessment object. The method provided by the invention can be used for comparing the model more suitable for small sample data, and simultaneously improves the accuracy and reliability of analysis and evaluation.

Inventors

ZHENG SHUYANG
Request for anonymity
WANG LEI
Request for anonymity
Request for anonymity

Assignees

空军装备部驻成都地区军事代表局驻绵阳地区第一军事代表室

Dates

Publication Date: 20260508
Application Date: 20260123

Claims (10)

1. The quality and reliability evaluation method based on the small sample data is characterized by comprising the following steps: Acquiring reliability and quality evaluation data of an evaluation object; Invoking a classical probability distribution life model based on a maximum likelihood estimation method to perform reliability evaluation on the evaluation object to obtain a first fitting goodness test result; calling a multisource information fusion reliability assessment model based on a BAYES method to carry out reliability assessment on the assessment object, and obtaining a second fitting goodness test result; Comparing the first fitting goodness test result with the second fitting goodness test result, and selecting a probability distribution model and a distribution parameter with high final fitting goodness as a reliability evaluation mathematical model; And extracting a reliability evaluation result of the reliability evaluation mathematical model, and carrying out quality analysis evaluation on the quality condition of the evaluation object based on the reliability evaluation result.
2. The method for evaluating the quality and the reliability based on the small sample data according to claim 1, wherein the step of invoking a classical probability distribution life model based on the maximum likelihood estimation method to evaluate the reliability of the evaluation object to obtain a first goodness-of-fit test result comprises: Acquiring reliability evaluation data of an evaluation object within a preset time period; Calling a classical probability distribution life model, and solving a distribution parameter and a fitting goodness test result by adopting a maximum likelihood estimation method; and obtaining a first goodness-of-fit test result and a first average fault interval time based on the distribution parameters and the goodness-of-fit test result.
3. The method for evaluating the quality and the reliability based on the small sample data according to claim 2, wherein the step of calling the classical probability distribution life model and solving the distribution parameters and the fitting goodness-of-fit test result by using a maximum likelihood estimation method comprises the following steps: respectively selecting a probability distribution model as an exponential distribution and a Wilbecco distribution, and respectively solving the distribution parameters of the exponential distribution and the distribution parameters of the Wilbecco distribution by adopting a maximum likelihood estimation method; respectively carrying out fitting goodness test on the distribution parameters of the exponential distribution and the distribution parameters of the Wilby distribution by using a K-S method; and determining probability distribution with high fitting goodness test as probability distribution of the classical probability distribution life model according to the result of the fitting goodness test.
4. The small sample data based quality and reliability evaluation method according to claim 1, wherein the performing reliability evaluation on the evaluation object by invoking a multi-source information fusion reliability evaluation model based on the bailes method to obtain a second goodness-of-fit test result comprises: Screening and cleaning the reliability and quality data in the reliability and quality evaluation database of the evaluation object; Calling a multisource information fusion reliability assessment model, and solving distribution parameters and a fitting goodness test result by adopting a BAYES method; And obtaining a second fitting goodness test result and a second average fault interval time based on the distribution parameters and the fitting goodness test result.
5. The method for evaluating the quality and the reliability based on the small sample data according to claim 4, wherein the steps of calling the multisource information fusion reliability evaluation model and solving the distribution parameters and the fitting goodness test result by adopting a BAYES method comprise the following steps: Invoking a quality and reliability evaluation database storing the reliability data of the evaluation object, and selecting similar products and weight scores thereof of the evaluation object by a BWM method; based on the similar products and the weight scores thereof, respectively selecting probability distribution models as exponential distribution and Wilbon distribution, and acquiring initial prior distribution of the evaluation object; According to the prior information of the initial prior distribution, a posterior calculation formula of the evaluation object is obtained by using a BAYES method; solving a non-analytic solution of the BAYES method by using an MCMC method; Taking the posterior mean value of the non-analytic solution as a posterior reliability characteristic parameter to carry out fitting goodness test; and determining probability distribution with high fitting goodness test as probability distribution of the multisource information fusion reliability evaluation model according to the fitting goodness test result.
6. The method for evaluating quality and reliability based on small sample data according to claim 5, wherein the selecting similar products and weight scores thereof of the evaluation object by BWM method comprises: performing compatibility test on the reliability data of the selected similar products by using a rank sum test method, and screening out similar product types meeting the requirements; and determining the weight scores of the screened similar product models.
7. The method for evaluating quality and reliability based on small sample data according to claim 5, wherein the obtaining an initial prior distribution of the evaluation object based on the similar products and weight scores thereof comprises: adopting conjugate prior distribution as prior distribution of exponential distribution; The normal distribution or the uniform distribution is selected as the prior distribution of the Wilby distribution; And carrying out weighted fusion on the prior distribution of the index distribution and the prior distribution of the Weilbu distribution based on similar product score weights to obtain the initial prior distribution of the evaluation object.
8. The method for evaluating quality and reliability based on small sample data according to claim 1, wherein the evaluating quality of an evaluation object based on the reliability evaluation result comprises: Calculating batch fault rate of the evaluation object, annual zeroing proportion, fault proportion of each component unit of the evaluation object, fault cause proportion of each fault cause of the evaluation object, fault quantity proportion of each geographical environment condition of the evaluation object, fitting a curve of time-dependent change of the fault quantity, fitting time-dependent change of the fault quantity of each fault cause, and obtaining a statistical analysis result; and taking the reliability evaluation result as a central main index, and carrying out quality analysis and evaluation by combining the statistical analysis result.
9. The method for evaluating the quality and the reliability based on the small sample data according to claim 8, wherein the step of performing the quality analysis evaluation by combining the statistical analysis result comprises the following steps: Combining statistical analysis results of different dimensions, and analyzing and evaluating the quality characteristics of the object; Analyzing the quality control condition according to the change of the fault trend; Analyzing the design improvement key points of the product according to the fault type distribution condition; and analyzing the outfield use guarantee capacity according to the maintenance efficiency.
10. The small sample data-based quality and reliability evaluation system applied to the small sample data-based quality and reliability evaluation method according to any one of claims 1 to 9, characterized by comprising: the data acquisition module is used for acquiring the reliability and quality evaluation data of the evaluation object; The first test result module is used for calling a classical probability distribution life model based on a maximum likelihood estimation method to carry out reliability evaluation on the evaluation object so as to obtain a first fitting goodness test result; The second test result module is used for calling a multi-source information fusion reliability evaluation model based on the BAYES method to evaluate the reliability of the evaluation object so as to obtain a second fitting goodness test result; The result comparison module is used for comparing the first fitting goodness test result with the second fitting goodness test result, and selecting a probability distribution model and a distribution parameter with high final fitting goodness as a reliability evaluation mathematical model; and the evaluation analysis module is used for extracting the reliability evaluation result of the reliability evaluation mathematical model and carrying out quality analysis and evaluation on the quality condition of the evaluation object based on the reliability evaluation result.

Description

Quality and reliability evaluation method and system based on small sample data Technical Field The invention relates to the technical field of electronic information, in particular to a quality and reliability evaluation method and system based on small sample data. Background At present, reliability evaluation of avionic equipment is mainly carried out by fitting a distribution model of fault time data, and adopting a classical probability statistical method such as a least square method, a maximum likelihood estimation method and the like to obtain characteristic parameters, wherein the classical probability statistical method needs a large amount of use test data. Although the aviation equipment underwriting units develop for decades, the production guarantees equipment of a plurality of batches and models, and abundant multi-source data information is accumulated, because the reliability information of each stage is not effectively utilized, the utilization degree of similar product data is low, the data among the stages do not form dynamic link relation, the reliability evaluation work among the stages is relatively independent, and the like, the reliability test data accumulated in advance of the aviation electronic equipment cannot be effectively utilized in the reliability evaluation work of the existing aviation electronic equipment, and the engineering development information and the test data are underutilized In addition, the avionic device has high development cost, high price and high cost of acquiring reliability test data due to few test prototypes, so that the avionic device can collect very little reliability test data in the same time period and is in a small sample state, and it is well known that the accuracy and precision of a reliability evaluation result are seriously influenced by the size of the data quantity, namely the richness of the reliability information. Therefore, how to accurately evaluate the equipment reliability level under the data condition of a small sample by using the existing data is a problem to be solved in the equipment reliability work. Disclosure of Invention In view of the above, the invention provides a quality and reliability evaluation method and system based on small sample data, which aims to realize quality and reliability evaluation of cross-stage multi-source information fusion by adopting small samples and further improve the accuracy and reliability of analysis and evaluation. In order to solve the technical problems, the technical scheme of the invention is to provide a quality and reliability evaluation method based on small sample data, which comprises the following steps: Acquiring reliability and quality evaluation data of an evaluation object; Invoking a classical probability distribution life model based on a maximum likelihood estimation method to perform reliability evaluation on the evaluation object to obtain a first fitting goodness test result; calling a multisource information fusion reliability assessment model based on a BAYES method to carry out reliability assessment on the assessment object, and obtaining a second fitting goodness test result; Comparing the first fitting goodness test result with the second fitting goodness test result, and selecting a probability distribution model and a distribution parameter with high final fitting goodness as a reliability evaluation mathematical model; And extracting a reliability evaluation result of the reliability evaluation mathematical model, and carrying out quality analysis evaluation on the quality condition of the evaluation object based on the reliability evaluation result. As an implementation manner, the invoking the classical probability distribution lifetime model based on the maximum likelihood estimation method to perform reliability evaluation on the evaluation object to obtain a first goodness-of-fit test result includes: Acquiring reliability evaluation data of an evaluation object within a preset time period; Calling a classical probability distribution life model, and solving a distribution parameter and a fitting goodness test result by adopting a maximum likelihood estimation method; and obtaining a first goodness-of-fit test result and a first average fault interval time based on the distribution parameters and the goodness-of-fit test result. As an implementation manner, the method for calling the classical probability distribution life model and solving the distribution parameters and the fitting goodness test result by adopting the maximum likelihood estimation method comprises the following steps: respectively selecting a probability distribution model as an exponential distribution and a Wilbecco distribution, and respectively solving the distribution parameters of the exponential distribution and the distribution parameters of the Wilbecco distribution by adopting a maximum likelihood estimation method; respectively carrying out fitting goodness test on the dis