摘要
在车辆调度建模时,由于突发性和不规律性,无法形成稳定的可预测状态。传统的车辆调度网络中,缺少必要的预报机制,导致调度存在较大的滞后性,调度时间过长。提出改进经纬格任务的遗传算法,利用经纬格性能体系结构的预报机制,改良初始种群生成方式,提高遗传算法衍生率减少迭代次数,减少运算提高运行速度。对车辆调度任务提出了适应度函数,满足改进算法对调度任务的适应度。实验结果表明,提出的改进优化经纬格任务遗传算法提高了调度的性能,更优于传统调度策略。
In traditional vehicle scheduling network, the vehicle scheduling model cannot form a stable predictable state due to the sudden and irregular nature, leading to the fact that the scheduling has large lag, and the scheduling time is too long. A genetic algorithm based on improved Lat-Lon Grid task was proposed in this paper. The prediction mechanism of Lat-Lon Grid ＇s performance system structure was used to improve the generation mode of initial population, increase the derivative ratio of genetic algorithm, and reduce the number of iterations and the operation speed. Fitness function was proposed for vehicle scheduling task, to meet the fitness of improved algorithm for scheduling task. Experimental results show that the proposed genetic algorithm can increase the performance of scheduling, which is more superior to traditional scheduling strategy.
出处
《计算机仿真》
CSCD
北大核心
2015年第11期203-206,367,共5页
Computer Simulation
关键词
经纬格任务
遗传算法
车辆调度
Lat-Lon Grid task
Genetic algorithm
Vehicle scheduling
