Root cause diagnosis of disturbances propagation paths by using improved convergent cross mapping

摘要

Modern plants' measurement devices and control systems are cross-linked with each other. A disturbance in major equipment may lead to plant-wide disturbances, which makes it difficult to find the root cause. This paper presents a data-driven method to detect the root cause by finding the disturbance propagation path. The major contribution is to improve the convergent cross mapping by revising the shadow manifold construction. We apply the improved convergent cross mapping (ICCM) to a chemical process and benchmark systems, which shows the ICCM's advantages on causality detection.