Intelligent adaptive control in an on-line cane transport scheduler : final report UCQ2S

Abstract

The vast majority of the Australian sugar crop is transported from the field to the sugar mill by privately owned cane railways. The total length of the permanent way exceeds 3000 km and approximately 500 x 106 tonne kilometres of cane are hauled during the crushing season, making the combined operation the third largest rail transport system in Australia.

Transporting cane from the field to the factory is an expensive process. Both capital and operating costs are large and cane transport is the largest cost unit in the manufacturing of raw sugar accounting for about one third of the total manufacturing costs.

Cane railway systems perform two major tasks. Firstly, they take empty bins from the mill and deliver them to the growers where they are filled with chopped cane, and secondly, they collect the full bins from the growers and return them to the mill. At the mill, the full bins are weighed and then move onto a tip where the cane is removed. The now empty bins are ready for delivery to the growers.

The Cane Railway Scheduling Problem is to design a set of locomotive runs, or a schedule, to satisfy both the mill and the growers. A substantial set of constraints has to be considered and there's no definitive objective function that can be used to assess the merits of the schedules produced.

A system, the Automatic Cane Railway Scheduling System (ACRSS) has been developed to generate schedules solving the cane transport problem.