针对集装箱堆场多场桥调度问题,构建调度模型,提出一种混合演化策略算法。采用基于实数的四维个体编码方法,设计了基于三点交叉互换的重组算子以及基于两点互换的变异算子,并采用三种不同的局部搜索策略来优化算法的性能。通过算例分析证明混合演化策略算法在优化多场桥调度问题时的有效性。在三种不同的局部搜索策略中,基于互换操作的局部搜索策略要优于其他两种,能明显改善演化策略算法的优化性能。最后,通过一组对比试验对局部搜索的次数进行了分析。
We built the scheduling model and proposed a hybrid evolutionary strategy algorithm for the problem of container yard cranes scheduling. We adopted the real number-based four-dimension individual coding method,and designed the recombination operator,which is based on three-point crossover interchange,and the mutation operator which is based on two-point swap,as well as used three different local search strategies to optimise the performance of the algorithm. Through example analysis we proved the effectiveness of hybrid evolutionary strategy algorithm in optimising yard cranes scheduling problem. Among three different local search strategies,the local search strategy based on swap operation was better than the other two,it could obviously improve the optimisation performance of evolutionary strategy algorithm. At last,we analysed the local search times through a series of contrast tests.