Optimization of purchasing plans based on forecasted demand for resources

Abstract

The challenge of enhancing purchasing strategies within a large organization, taking into account non-linear constraints, has been thoroughly examined and formalized. The increase in demand for resources over time, changes in prices and the costs of tender procedures are taken into account. The purchasing strategy integrates forecasts derived from historical data and is in accordance with the capacity plan. A simple, linear autoregressive model is used to predict demand changes and a predictive control technique with a moving horizon. Furthermore, the findings from experiments utilizing the genetic algorithm are presented. Finally, important open problems are discussed, the solution of which would expand the scope of applicability and universality of the developed tool.