摘要
通过分析快速蚂蚁算法的原理和易陷入局部最优的缺点,提出了将贪婪算法和快速蚂蚁算法相结合的混合算法求解物流车辆路径问题。混合算法在最优值未改进次数超过限定次数时,自动调用贪婪算法来寻找一个局部最优解,并调整相应路径上信息素的量。为保证解的多样性,对贪婪算法本身使用随机选择第一个客户的方法进行了调整。用计算实例比较并分析了快速蚂蚁算法、混合算法及其他算法应用到车辆路径问题上的结果,说明了贪婪算法使混合算法跳出局部最优的过程以及混合算法的不足之处。
Through analyzing the principles of fast ant algorithm and its shortage of easily trapping into the local optimization value, a hybrid algorithm combining the fast ant algorithm with greedy algorithm was put forward to solve the vehicle routing problem. When optimization value was not improved for certain times, the improved greedy algorithm was invoked to search a local optimization value, and the pheromones on the corresponding routes were adjusted. To ensure the solution diversity, the greedy algorithm was adjusted through randomly choosing the first customer. The results of fast ant algorithm, hybrid algorithm and other algorithms applied to the vehicle routing problem were compared and analyzed by the case study. The process that how the greedy algorithm enables the hybrid algorithm avoid converging to the local optimization value and the shortages of the hybrid algorithm were presented.
出处
《工业工程与管理》
2007年第4期15-19,共5页
Industrial Engineering and Management
关键词
快速蚂蚁算法
车辆路径问题
贪婪算法
fast ant algorithm
vehicle routing problem
greedy algorithm
