Planning under non-stationary stochastic demand and finite capaci

In this assignment we consider the planning of a product family consisting of N items under non-stationary stochastic demand and finite capacity under a rolling schedule regime. We model the demand process and forecasting process in a generic way, allowing for modelling a wide range of practical situations. Given the demand and forecasting process we propose two conceptually different approaches to decide on the periodic order release decisions for each item. The first approach is based on detailed forecasting and detailed inventory balance equations up to some planning horizon. The resulting model is solved by LP and yields the required order release decisions. The first period decisions are implemented. The second approach is based on aggregate forecasting and aggregate inventory balance equations. The resulting aggregate problem is solved yielding the aggregate plan for the first period. The first period order release decisions are determined by allocating the aggregate plan among the different items. We propose three different allocation rules.