Computer aided optimisation of an agro-industrial complex consisting of processes and inventories by means of a custom-developed hybrid genetic algorithm

Abstract

Mixing is a very common industrial process; the resulting mixture may be either a final or an intermediate product, of predetermined content for each constituent, according to certain quality specifications. During the production process within an agro-industrial complex, raw materials (usually highly inhomogeneous and sensitive by nature) from external inventories, distributed in the time/space domain, should cater for the internal ones that feed the mixer. The aim of the present work is the computer-aided optimisation of such a combined industrial process by (a) continuously locating transhipment points, (b) determining the optimal routing of raw materials, (c) reshuffling the external inventory list and (d) obtaining the minimal operating cost provided through a custom-developed module, namely the external inventory cost optimiser which is fed with real-time data obtained by the coupled use of a geographical information system (GIS) with a Global Positioning System (GPS). To accomplish such tasks, a hybrid genetic algorithm (GA) is employed featuring specific myopic rules. A case study is also presented referring to the design of an ethanol production unit located in the Thessaly plane of Central Greece where the number of inventories varies. In addition, the impacts of certain unpredictable events on raw material concentration (e.g. deterioration of residues quality, etc) are discussed.