A clonal selection algorithm using the rolling planning and an extended memory cell for the inventory routing problem

We consider the inventory replenishment problem and the vehicle routing problem simultaneously in the vending machine operation. This problem is known as the inventory routing problem. We design a memory cell in the clonal selection algorithm. The memory cell stores the best solution of previous solved problem and uses an initial solution for next problem. In general, the other clonal selection algorithm used memory cell for reserving the best solution in current problem. Experiments are performed for testing efficiency of the memory cell in demand uncertainty. Experiment result shows that the solution quality of our algorithm is similar to general clonal selection algorithm and the calculations time is reduced by 20% when the demand uncertainty is less than 30%.