论文部分内容阅读
核概率主元分析(kernel probabilistic principal component analysis,KPPCA)能够有效去除过程的非线性。但是KPPCA仅构造了生产过程的静态线性关系,处理具有较强动态特性的实际工业生产过程效果较差。为克服上述缺点,提出一种基于动态KPPCA的过程监测方法,利用核函数将经过压缩的动态增广数据映射到高维空间,然后利用PPCA对满足线性关系的过程变量映射值进行监测。仿真结果表明:该方法监测指标对故障的灵敏度高,误报率和漏检率较小,故障状况与正常状况很明显的分离开来。
Kernel probabilistic principal component analysis (KPPCA) can effectively remove the nonlinearity of the process. However, KPPCA only constructs the static linear relationship of the production process, and the actual industrial production process with strong dynamic characteristics is less effective. In order to overcome the shortcomings mentioned above, a dynamic KPPCA-based process monitoring method is proposed, which maps the compressed dynamic augmented data into high-dimensional space by kernel function, and then uses PPCA to monitor the process variable mapping values that satisfy the linear relationship. The simulation results show that the method has high sensitivity to faults, false alarm rate and missed detection rate, and fault status is clearly separated from normal status.