Радіоелектронні і комп'ютерні системи (Mar 2019)
EXPANSION METHOD OF LOGISTIC SUPPLY CHAIN MODEL, PRESENTED IN THE FORM OF A TWO-LEVEL NESTED PETRI NETS
Abstract
The issues related to the development of adequate network models of the processes of creating, deploying and supporting the functioning of the full logistics supply chain of dried fruit (SCDF) in Ukraine are considered. It is shown that the creation and operation of the SCDF, compared to other SCM class systems, raises a number of specific problems caused by the complexity of the interaction of raw material suppliers (fresh fruit), manufacturers of final products (drying, packaging), storage terminals, distributors, 3PL and 4PL providers (retailers). These problems are due to the fact that the interaction of participants in business processes in the SCDF generates a lot of material, financial and information flows, as well as flows of services from sources of raw materials to the final consumer. The variety of world regions from which dried fruit is delivered to Ukraine, a wide range of products supplied, yield, currency fluctuations, seasonality are the causes of a high level of uncertainty in the processes of formation and decision-making by the SCDF participants. The urgency of the problem is substantiated, the essence of which is to develop a temporal component in the network model of the SCDF, which adequately reflects in time the logistic processes that take place during the operation of the supply chain. A model of a full SCDF in the form of a two-level nested Petri net (NPN) is considered. At the same time, the network model of SCDF functioning in the form of a NPN includes two types of objects - the central link reflecting the activity of the focusing company on processing raw materials (dried fruits), and a number of subnets that simulate the activity of suppliers of raw materials and distributors of finished products. It is indicated that to predict the consequences of the current (or some predetermined) situation and to respond to inquiries about the future state of elements of the SCDF, a predictive model can be used that explicitly takes into account time dependencies. In this case, requests can be about the time of occurrence of certain events, and the fact that certain circumstances are present at a certain point in time. A method is described for extending the network model of the SCDF to comply with the 'just-in-time' principle when modeling business processes that take place supply chain.
Keywords
