CAJ | 학술논문

针对传统调度模型预见性不强的弱点，提出一个基于支持向量机（ SVM）的公交车辆到达枢纽时间的预测模型，基于该模型构建以所有乘客节约时间最大为目标的调度模型，动态协调公交车辆从枢纽的发车时间，并基于遗传算法对该模型进行求解。最后，我们以大连市沙河口火车站枢纽为实例，对该模型和算法的可行性进行了检验，结果显示，本文提出的调度方法优于传统调度策略。
침대전통조도모형예견성불강적약점，제출일개기우지지향량궤（ SVM）적공교차량도체추뉴시간적예측모형，기우해모형구건이소유승객절약시간최대위목표적조도모형，동태협조공교차량종추뉴적발차시간，병기우유전산법대해모형진행구해。최후，아문이대련시사하구화차참추뉴위실례，대해모형화산법적가행성진행료검험，결과현시，본문제출적조도방법우우전통조도책략。
Due to the poor prediction of traditional scheduling model, a SVM-based prediction model is put for-ward to forecast the arrival time of public transport vehicles at the transfer station.Then a scheduling model aimed at reducing the total waiting time of passengers is constructed to dynamically coordinate the departure time at the transfer station.Genetic Algorithm is applied to solve the dynamic scheduling problem in this paper. Finally, this paper verifies the feasibility of the model and algorithm with the data of Shahekou station in Dalian city.And the results show that the scheduling method this paper proposed is superior to the traditional scheduling method.