Solving truck allocation problem in sugar cane industry by genetic algorithms

In Thailand, the cost in transporting cane from farms to mill is a large portion of sugar cane production cost. In order to reduce the transportation cost, many methods have been proposed. One of the methods is to fix the number of trucks used in system. The trucks that will be used to carry sugar cane from farms to mill in the coming year crop must be registered. The registered trucks must be allocated to harvesting regions to assure that all mature cane can be carried and the mill gets enough cane. In this paper, two algorithms based on genetic algorithm have been developed to allocate trucks. The first algorithm is based on classical genetic algorithm concept, while the second algorithm is the genetic algorithm with weighted chromosome encoding (WCE). By applying the first method, a low utilization of chromosomes and many infeasible solutions have been found. In the second algorithm, the weighting system has been applied to assist in allocating trucks. By improving GA with WCE, a high utilization of chromosomes and many feasible solutions are obtained. In addition, the computation time is more efficient and the solution is also more reasonable than that obtained from the classical algorithm.