中文摘要：

文章针对目前动态过程质量异常模式的识别精度不高的问题,提出一种基于统计特征的动态过程质量异常模式识别方法。该方法首先提取出样本数据的16个统计特征,再通过相关性分析筛选出相关性较小的统计特征;然后将筛选后的相关性较小的统计特征输入支持向量机（SVM）分类器进行识别。通过仿真实验进行验证,实验结果表明,基于统计特征的异常模式识别模型能够提高整体的识别精度,可适用于生产现场的质量监控。

英文摘要：

To solve the problem of low recognition accuracy in dynamic process of quality abnormal pattern, this paper pro- poses an identification method of dynamic quality abnormal pattern based on statistical features. In the proposed method, 16 statistical features are extracted from the sample data, and then the less relevant statistical features are screened out via correlation analysis; finally the less relevant statistical features are inputted into the support vector machine （SVM） for recognition. Through simulation experiment the paper comes to a conclusion that the quality abnormal pattern recognition model based on statistical fea- tures can improve integral recognition accuracy, and can be applied to the quality control in the production site.