A heuristic algorithm for master planning that satisfies multiple objectives

This study proposes a heuristic algorithm, called the multi-objective master planning algorithm (MOMPA), to solve master planning (MP) problems for a supply chain network with multiple finished products. MOMPA has three objectives: to minimize delay penalties, to minimize use of outsourcing capacity, and to minimize the costs of materials, production, processing, transportation, and inventory holding-all while respecting the capacity limitations and the demand deadlines of all those involved in a given supply chain network. MOMPA plans each demand, one by one, without backtracking, and sorts those demands using a sorting mechanism that is part of the algorithm. For each demand, the minimum production cost tree is determined within the limits of the time bucket for the demand deadline. The maximum available capacity of this tree is then computed for the ''no delay'' case. Following this calculation, the delay-or-not criterion is evaluated to determine whether or not further delay is necessary. MOMPA compares the results of these two procedures and allocates the appropriate capacities to the demand for all the nodes on the selected tree. If the minimum cost production tree has no available capacity, MOMPA adjusts the network and looks for a new tree. With complexity and computational analysis, MOMPA is shown to be very efficient in solving MP problems, sometimes generating the same optimal solution as the LP model.