Electronic component fault diagnosis model based on data analysis and prediction

Abstract

The invention provides an electronic component fault diagnosis model based on data analysis and prediction. The electronic component fault diagnosis model comprises a data acquisition module, a data preprocessing module, a fault diagnosis model construction module, a prediction module and a prediction result output module. The fault diagnosis model construction module receives the 'clean' structured data set and then performs clustering to obtain a plurality of clusters. For each cluster, fractal geometry is applied to quantify the fractal dimension value of the cluster, the fractal dimensionvalue of each cluster is the fault feature of the cluster, and the value is used for judging the fault type existing in the batch components. Input data is stored in a storage module in a classified mode according to fault types. According to the method, the manual repetitive detection work is reduced, the music detection period is shortened, the screening cost is reduced, the production quality control potential of electronic components of manufacturers is excavated, the continuous improvement of the music ability and benefits is realized, high-quality and high-reliability electronic components are provided for models, and the reliability and stability of aerospace model products are ensured.

Description

technical field [0001] The invention relates to the field of electronic components, in particular to a fault diagnosis model of electronic components based on data analysis and prediction. Background technique [0002] Electronic components are important components of aerospace weapons and equipment, and the reliability and stability of their quality may determine the success or failure of aerospace product missions. According to statistics, 80% of aerospace quality problems are related to the quality of components. [0003] At present, the component quality and reliability assurance business mainly relies on the traditional test and detection technology, that is, the "one case, one discussion" small sample test and identification technology. Through the test and test of each type of component, each specification, and each production batch, Such as appearance inspection, pre-screening test, temperature shock, sealing inspection, X-ray inspection, vibration, impact, stress, ...

AI Technical Summary

Problems solved by technology

In the current new environment of parallel development of multi-model tasks and high-intensity launches, traditional test and test technology methods are under increasing pressure on efficiency and cost, such as time-consuming, labor-intensive, and material-intensive task cycles, which cannot keep up with the fast pace Research and development requirements; test data is not centrally stored and utilized, and business value cannot be mined; business processes rely heavily on manual processing, data quality is inconvenient to quote, process traceability is difficult, and processing efficiency is low, etc.

Images