Point cloud analysis method for high voltage breakdown and creepage problems in metal-encapsulated Microsystems
Abstract
In the face of the potential breakdown and creepage risks in metal encapsulated high-voltage isolation Microsystems, a new
reliability analysis method combining high-precision modeling and graph theory was established in this paper to provide a strong guarantee
for product design and testing. Firstly, the 3D model of the microsystem is transformed into a point cloud model, and a computable
numerical model is formed by fi ltering and surface reconstruction. Then, according to the geometric characteristics and physical relations,
the surface creepage problem and breakdown problem are equivalent to the calculation of geodesic path and Euclidean path respectively.
The Dijkstra algorithm is optimized according to the layout of microsystem, and the breakdown and creepage paths after packaging are
calculated. Finally, the risk level of the product is judged by reference to the experimental standard. The results of the control experiment
match with the calculation, and the precision is ideal, which shows the eff ectiveness of the method for microsystem-related problems.
reliability analysis method combining high-precision modeling and graph theory was established in this paper to provide a strong guarantee
for product design and testing. Firstly, the 3D model of the microsystem is transformed into a point cloud model, and a computable
numerical model is formed by fi ltering and surface reconstruction. Then, according to the geometric characteristics and physical relations,
the surface creepage problem and breakdown problem are equivalent to the calculation of geodesic path and Euclidean path respectively.
The Dijkstra algorithm is optimized according to the layout of microsystem, and the breakdown and creepage paths after packaging are
calculated. Finally, the risk level of the product is judged by reference to the experimental standard. The results of the control experiment
match with the calculation, and the precision is ideal, which shows the eff ectiveness of the method for microsystem-related problems.
Keywords
microsystem; Point cloud analysis; Dijkstra’s algorithm; Breakdown phenomenon; Creepage phenomenon
Full Text:
PDFReferences
[1].GuoYL,BennamounM,SohelF,etal.3Dobjectrecognitioninclutteredsceneswithlocalsurfacefeatures:asurvey[J].IEEETransaction SonPatternAnalysisandMac
hineIntelligence, 2014, 4 (11) : 2270-2287.
[2]. Jianwei Li,Jiawang Zhan. Research progress of 3D point cloud registration methods [J]. Journal of Image and Graphics, 2022,27 (02):349-367.
[3]. Xiang Zhang,Deyu Chen. Application of laser point cloud in 3D model reconstruction [J]. Laser Journal,2023,44(01):129-133.
[4]. Yanjin Du,Zhao Xu. Divisible and parallel reconstruction of massive point Clouds on Physical Surface [J]. Manufacturing Automati
on,2023,45(02):49-52+58.
[5]. Jie Yang,Xiaorong Wen,Qiulai Wang, etal. [J]. Journal of Northwest Forestry University,2022,37(06):129-137.
DOI: https://doi.org/10.18686/esta.v10i3.467
Refbacks
- There are currently no refbacks.
Copyright (c) 2023 Lu Wang,Han Bai
