摘要
Unbalanced multi-stage logistics systems are optimized using an improved genetic algorithm based on the Prüfer number and the effective capacity coding. The improved decoding procedure uses the node capacity of the logistics system as an important factor, which influences the decoding procedure. As a result, any Prüfer number produced stochastically can be decoded to a feasible logistics pattern, which matchs the node capacities of the logistics system. With effective capacity coding, an unbalanced logistics system can be converted to a set of balanced systems. The effective capacity coding was combined with the Prefer number to construct the chromosome for the new method to search the whole solution space of the unbalanced multi-stage logistics system. Simulation results show that the new method finds a better solution with less computational time than st-GA. Although using a little more memory, the new method is still an efficient and robust method for optimizing unbalanced multi-stage logistics systems.
Unbalanced multi-stage logistics systems are optimized using an improved genetic algorithm based on the Prüfer number and the effective capacity coding. The improved decoding procedure uses the node capacity of the logistics system as an important factor, which influences the decoding procedure. As a result, any Prüfer number produced stochastically can be decoded to a feasible logistics pattern, which matchs the node capacities of the logistics system. With effective capacity coding, an unbalanced logistics system can be converted to a set of balanced systems. The effective capacity coding was combined with the Prefer number to construct the chromosome for the new method to search the whole solution space of the unbalanced multi-stage logistics system. Simulation results show that the new method finds a better solution with less computational time than st-GA. Although using a little more memory, the new method is still an efficient and robust method for optimizing unbalanced multi-stage logistics systems.