基于红外热图序列的板级芯片故障诊断研究

Research on Faults Diagnosis of Board Chip Based on Infrared Images Series

摘要:

提出一种基于红外热图序列的板级芯片开/短路缺陷检测方法。首先记录芯片关键区域在上电程序响应过程的温度均值序列,运用Savitzky Golay卷积平滑法对其平滑滤波后提取时域特征参量,利用主成分分析法优选关键特征；然后构建支持向量机分类模型,利用粒子群算法优化支持向量机模型参数,使其能有效区分不同的电路板故障类型。为验证提出的方法在芯片开/短路缺陷检测中的有效性,在开发板上的主控芯片上进行了多种焊球开/短路模拟实验。结果表明,优化后的分类模型在测试集的交叉验证分类准确率为96.90%,证明了该方法诊断芯片开/短路缺陷的有效性。

Abstract:

A chip open/short circuit defects inspection method is proposed based on infrared images series. Firstly, the mean temperature series of the critical area of chip during the response process of the power-on procedure was recorded, the Savitzky Golay convolution smoothing method was applied to extract the time domain feature parameters after smoothing and filtering, and the principal component analysis method was uesd to select the key features. Then a support vector machine classification model was constructed, whose parameters were optimized by particle swarm algorithm to effectively distinguish different types of circuit board defects. In order to prove the validness of the method proposed, a variety of solder ball open/short circuit experiments were carried out on a CPU chip of circuit board. The research results show that the cross-validation classification accuracy of the optimized SVM model in the test dataset is 96.90%, which proves the validness of the method for detecting the chip open/short defects in this paper.