混合粒子群算法在ETV调度优化中的应用

为了提高机场货运区(Elevating Transfer Vehicle,ETV)转运效率,建立以最小化任务集调度时间为优化目标的调度模型,提出一种混合的粒子群算法对ETV调度问题求解。算法对加速因子采取动态的自适应调整策略;采用混沌序列替代标准粒子群中的随机数;建立平均粒距、适应度方差和汉明距离相结合的早熟判断机制并采用混沌算子扰动微粒的位置来跳出局部最优。通过实例验证和遗传算法、模拟退火等经典的优化算法以及非线性学习因子粒子群、混沌粒子群等改进的粒子群算法相比,该算法在ETV调度最优序列的求解中收敛速度快,全局寻优能力强,稳定性好;和传统的链式调度算法相比,平均调度任务时间减少了15.6%,较好地解决了ETV转运效率低的问题。

APPLICATION OF HYBRID PARTICLE SWARM OPTIMIZATION IN ETV SCHEDULING OPTIMIZATION

In order to improve the ETV transshipment efficiency of the airport freight area,we established a scheduling model with the minimum task set scheduling time as the optimization goal and proposed a hybrid particle swarm scheduling algorithm.The algorithm adopted a dynamic adaptive adjustment strategy for the acceleration factor;the chaotic sequence was used to replace the random number in the standard particle group;the premature judgment mechanism combining the average grain distance,fitness variance and Hamming distance was established and the position of the particle was updated by chaotic disturbance to come out of local optimum.The results show that the improved particle swarm scheduling algorithm has a fast convergence speed and strong global optimization ability with good stability in solving the optimal sequence of ETV scheduling by compared with the standard particle swarm,nonlinear learning factor particle swarm,chaotic particle swarm algorithms.Compared with the traditional chain scheduling algorithm,the average scheduling task time of the improved particle swarm scheduling algorithm is reduced by 15.6%,which solves the problem of low ETV transshipment efficiency better.