A Hybrid ALNS-VND-CS-LB Approach to the Heterogeneous Load Allocation Problem

The Heterogeneous Cargo Allocation Problem has great relevance in logistics
and has high combinatorial complexity due to the number of variables and res-
tritions involved. In this work, a hybrid approach based on Adaptive Large
Neighborhood Search (ALNS), Variable Neighborhood Descent (VND), Clustering Search (CS) e
Local Branching (LB). The methodology was evaluated in 12 artificial instances generated by an algo-
rhythm developed in this work, using as reference the mathematical formulation of the problem
solved by the Gurobi Optimizer solver. The results showed that the proposed approach
presented competitive performance both in solution quality and execution time,
achieving solutions equal to or superior to those obtained by the solver in 91% of the evaluated instances.