Catastrophe Pre-Warning of Multi-Modular Floating Platforms with Ordinal Partition Networks

Abstract

Use of artificial networks to signify the onset of dynamic catastrophes in engineering systems is a simple and efficient strategy. Here an ordinal partition network and its construction from time series are introduced. The selection of mapping parameter is discussed in detail, which may significantly enhance the performance of the proposed method. Topological properties of the resulting network can sensitively detect the dynamic changes of original underlying systems, making the strategy workable. A Catastrophe Prediction Index (CPI) is proposed to serve as a monitoring indicator for pre-warning catastrophe events. The numerical results verify the feasibility of the proposed method.