论文部分内容阅读
为更好地根据用户指定的指标权重实行相应的控制策略,介绍一种以相对指标为控制目标的多目标优化模型并应用基于变权系数的遗传算法进行求解。算法每执行一次循环获得一个非劣解并对目标权重进行一次变化,从而使整个解集在Pareto前沿面运行。通过佛山市实际交叉口的例子进行了仿真试验,与传统的固定权重方法相比,机动车延误,行人延误,停车时间和路段平均速度都有不同程度的优化。
In order to carry out the corresponding control strategy according to user-specified index weight, a multi-objective optimization model based on relative index is introduced and the genetic algorithm based on variable weight coefficient is used to solve the problem. The algorithm obtains a non-inferior solution for each cycle of execution and makes a change to the target weight, so that the entire solution set runs on Pareto frontier. Through the example of Foshan actual intersection, the simulation experiment is carried out. Compared with the traditional fixed weight method, vehicle delay, pedestrian delay, parking time and the average speed of road sections are all improved to some extent.