Assessment of Equipment Health Index for Automated Process Control Systems Using Dynamical System Attractor Reconstruction

A method for calculating the health index of dynamic equipment in process control systems is proposed, based on phase attractor reconstruction. The method is intended as a computational module within an operator decision support system. System dynamics are reconstructed from a limited set of observed variables using Takens’ theorem. The deviation of the current state from the reference state is estimated via the Mahalanobis distance, which is then converted into a health index. The approach was tested on a Simulink model of a three-plunger pump, simulating three fault types and their combinations. The health index changes as degradation increases. Combined defects can be assessed both qualitatively and in approximate quantitative terms.