Computer System And Method For Causality Analysis Using Hybrid First-Principles And Inferential Model

摘要

The present invention is directed to computer-based methods and system to perform root-cause analysis on an industrial process. The methods and system load process data for an industrial process from a historian database and build a hybrid first-principles and inferential model. The methods and system then executes the hybrid model to generate KPIs for the industrial process using the loaded process variables. The methods and system then selects a subset of the KPIs to represent an event occurring in the industrial process, and divides the data for the subset into multiple subset of time series. The system and methods select time intervals from the time series based on the data variability in the selected time intervals and perform a cross-correlation between the loaded process variables and the selected time interval, resulting in a cross-correlation score for each loaded process variable. The methods and system then select precursor candidates from the loaded process variables based on the cross-correlation scores and execute a parametric model for performing quantitative analysis of the selected precursor candidates, resulting in a strength of correlation score for each precursor candidate. The methods and system select root-cause variables from the selected precursor candidates based on the strength of correlation scores for analyzing the root-cause of the event.

主权项

1. A computer-implement method for performing root-cause analysis on an industrial process, the method comprising:
loading process data for a subject industrial process from a historian database; building a hybrid first-principles and inferential model for calculating and predicting the said KPIs; executing a hybrid model for generating continuous key performance indicators (KPIs) for the industrial process using the loaded process data, the hybrid model comprising a first principles model as a primary model and an inferential model as a secondary model of the industrial process; selecting a subset of the KPIs values to represent an event occurring in the industrial process, wherein the subset of the KPIs are selected based on correlation to the event; dividing process input data for the selected KPIs into multiple subsets of time series, wherein dividing includes selecting at least one time interval from the time series based on variability of the data in the at least one time interval; performing a cross-correlation between the loaded process variables and the at least one selected time interval, wherein the performed cross-correlation results in calculating a score for each loaded process variable; automatically selecting precursor candidates from the loaded process variables based on the calculated cross-correlation score for each loaded process variable; building a parametric model for performing quantitative analysis of the selected precursor candidates, the parametric model calculating a score for each selected precursor candidate based on strength of correlation to the subset of the KPIs; and automatically selecting root-cause variables from the selected precursor candidates based on the calculated strength of correlation score for each of the selected precursor candidate, resulting in output representing implications of the root-cause precursor variables being automatically programmed at a plant control system for diagnosing a root-cause of the event.