Abstract

Effective planning strategies are essential to minimize high costs of production and inventory. Uncertainty and seasonal variation in product demand is a major issue that contributes to a substantial share of production planning costs. Hence, it is important to consider the uncertain information while designing a production planning model. This thesis is aimed at presenting a comparative analysis of deterministic and stochastic approaches towards finding optimal solutions for demand uncertainty problems. The first model is a generic mixed-integer programming model to maximize total profit. Decision variables are identified and random values are substituted by their expected values considering uncertainty to obtain the expected value solutions. Second model is formulated as a stochastic programming model by adding scenarios and probabilities in the deterministic model to explicitly account for the uncertainties in the product demand. The models are programmed and solved by LINGO optimization solver based on data collected from a brewing company. Several test problems are solved by varying the input parameters, product demand and probability of existence of scenarios to study the sensitivity of the models. A statistical comparative analysis is conducted on all the example problems by measuring the Expected Value of Perfect Information (EVPI), Value of Stochastic Solution (VSS) and the results are discussed.