A multiobjective approach for optimization of the multihead weighing process

Packaging technology can be of strategic importance to a company, as it can be a key to competitive advantage in the modern food industry. A multihead weighing process is a packaging process based on the sum of weights from several individual hoppers. The final quality of the packaged product in a modern automated food packing system should fulfill mo objectives -the quality of the packaging process itself and the sensory qualiry of the packaged food. The first one is often related to the proximity of the total weight to a nominal value of sale and to the variability of the packaging process, and the second one can be achieved by minimizing the total residence time of the food in the packing system. In this paper, we jointly address both objectives through multi objective programming. More precisely, a biobjective algorithm is developed, in which a subset of hoppers is selected from the set of available ones at each packing operation, in such a way that the relative importance of both aforementioned objectives is dynamically managed and adjusted. We conduct numerical experiments to examine the quality of the solutions being produced.