首页|期刊导航|安庆师范大学学报(自然科学版)|非正态下权重累加集成芯片封装质量判断方法

非正态下权重累加集成芯片封装质量判断方法OA

Quality Evaluation Method for Integrated Chip Packaging Test Data under Non-Normal Distribution Using Line Chart Vertex Weight Accumulation

中文摘要英文摘要

现有过程能力指数主要应用于集成芯片封装测试初期,受数据量少影响,数据多呈现非正态分布,从而导致集成芯片封装测试质量难以判断.文章提出了一种独立于均值和方差的折线图顶点权重累加法,采用99.73%的区间跨度点滤除偏移量过大的干扰数据,选择折线顶点权重累加均值替代过程均值,最终实现在无需均值和方差的前提下对集成芯片封装质量的精准判断.仿真结果表明,相较于现有各类非正态改进算法,该方法求解过程能力指数稳定性更优,可在测试初期及时发现集成芯片封装质量问题,节约封装和测试成本.

In the early stage of integrated chip(IC)packaging and testing,the limited amount of data often leads to non-normal distribution,making it difficult to evaluate quality using existing process capability indices.This paper proposes a Line Chart Vertex Weight Accumulation Method that does not rely on the mean or variance.By filtering out extreme deviation data points using the 99.73%interval range and replacing the traditional process mean with the weight-accumulated mean of line chart vertices,the method enables accurate quality evaluation of IC packaging without depending on distribution assumptions.Simulation results show that this method provides more stable process capability index calculations compared to other non-nor-mal distribution approaches and can effectively detect quality issues at the early testing stage,thereby reducing packaging and testing costs.

张珺瑚;刘建;章礼华;詹文法

安庆师范大学 智能制造与机器人学院,安徽 安庆 246133安庆师范大学 智能制造与机器人学院,安徽 安庆 246133安庆师范大学 智能制造与机器人学院,安徽 安庆 246133安庆师范大学 智能制造与机器人学院,安徽 安庆 246133

信息技术与安全科学

集成芯片封装质量折线图顶点权重累加法过程能力指数非正态分布

integrated chippackaging qualityline chart vertex weight accumulation methodprocess capability indexnon-normal distribution

《安庆师范大学学报(自然科学版)》 2026 (2)

33-38,6

安徽省高等学校科学研究项目(2023AH050481)和安徽省高等学校质量工程项目(2023zygzts036,2024zybj022)

10.13757/j.cnki.cn34-1328/n.2026.02.005

评论