Online Alignment of Process Analytical Models to Actual Process Operations

Abstract

Online alignment of process analytical models with actual process operations involves a batch modeling and analysis system that uses simple and computationally inexpensive techniques to combine data collected from ongoing, currently running, or online batch processes with The batch model built for the batch process is aligned such that the current operating status of the online batch process relative to the batch model can be reliably determined. This data alignment technology allows further statistical processing techniques, such as partial least squares (PLS) and principal component analysis (PCA) techniques, to be applied to online batch data to analyze the quality of the current running batch. In turn, these analyzes provide useful information to users such as batch operators so that they can determine the quality of the batch at that time based on the batch model and the likelihood that the desired batch output quality measure will be achieved at the end of the batch run. sex.

Description

technical field [0001] The present invention relates generally to process control systems, and more particularly to the implementation of data alignment techniques for a batch modeling system that aids in analyzing and predicting the quality of on-line or in-progress batch processes. Background technique [0002] Process control systems, such as those used in chemical, petroleum, or other processes, typically include one or more process controllers and communicatively coupled via an analog, digital, or combined analog / digital bus to at least one host computer or An input / output (I / O) device for an operator workstation and one or more field devices. Field devices, which can be, for example, valves, valve positioners, switches and sensors (eg, temperature, pressure, and flow rate sensors), perform process control functions in a process such as opening or closing valves and measuring process control parameters. Process controllers receive signals representing process measureme...

AI Technical Summary

Problems solved by technology

However, this assumption that the current batch has the same length of normalization as the aligned batches used to create the batch model is often wrong, and thus yields poor analytical results

Furthermore, such simplified heuristic rules are usually not sufficient for most applications

Therefore, these techniques lead to misleading results

Images