Bin Packing Problem with Compatible Categories Utilizing an Enhanced Variable Neighborhood Search Algorithm

The reduction of packing cost, computational time and increase in the utilization of resources are desirable in the fields of logistics and production systems. The research study is focused on the enhancement of the variable neighborhood search (VNS) metaheuristic algorithm for Bin Packing Problem with Compatible Categories (BPCC). The Searching and Matching Theories as a basis in the modification of changing the random perturbation of the VNS for BPCC algorithm. Significantly, the enhancement improved the searching and matching time of items and categories and effectively packed non-conflicting items in the least number of bins in a reduced CPU time. An average of 86.181 % reduction in the computational complexity of the algorithm in term of CPU time and an average of 0.046% required number of bins was obtained based on the original VNS for BPCC. This result was achieved by replacing the initial random choice of category of items with the selection of a least compatible category of items ranging from 500 to 5000 instances in a bin to be removed, compared, and repacked in the VNS’ matching procedure.