Advanced Engine Diagnostics Using Universal Process Modeling

There are many benefits in monitoring the health of an internal combustion engine, whether it be a spark-ignition, compression-ignition, or gas turbine. These benefits include detecting faulty or failing engine sensors and detecting subtle problems which could, if left undetected, cause serious engine damage. While there are many benefits to monitoring engine health, the challenge becomes how to determine if an engine is healthy or not.

Universal Process Modeling (UPM) is a new and unique multivariate modeling technique which can accurately determine the health of complex equipment such as an internal combustion engine. UPM is an inductive technique which uses a reference library of example data to describe how an engine normally operates. UPM calculates an overall “system” health and the health of each variable monitored. The healths are expressed in how many standard deviations they are away from their expected values. Engine-related problems are revealed when the system health falls outside a statistical limit. Insight into the nature of the problem may be gained by examining those variables which have deviated the most from their expected values. This paper discusses the features and benefits of UPM and describes its application to advanced engine diagnostics.