Dynamic allocation of predicted quantities to forecast intervals for pan stage supervisory support system process models

Abstract

This paper describes a critical feature for the industrial process models of an expert advisory system and their integration within a knowledge based supervisory support system (KBSSS) for advice on best practices and management of a sugar mill crystallization stage. This functionality works cooperatively to translate pan stage industrial process models, used during the forward prediction of pan stage operating conditions, to a time scale basis by dynamically allocating forecast processing quantities to predefined intervals over the prediction horizon. The innovative dynamic allocation procedure outlined underpins the prediction ability of the process models, acting in a backbone capacity, to establish forecasting capabilities for the system. The primary topic of this paper will be a description of the approach and how it supports the predictive modelling with focus on: (1) design features, (2) implementation and (3) application to the prediction of syrup quantities to the pan stage from cane receival and juice processing information.