Abstract: The hierarchical control structure of multi-lane traffic flow is divided into the coordinated control layer，segment control layer and vehicle control layer.In the paper an optimal control model is proposed for the segment control layer.The objective of the model is to maximize the total traffic flow volume and the average velocity in a segment，and the decision variables are the rates of vehicles’ number in different lanes on the same segment.The number of decision variables in this model is greatly reduced compared with that of traditional linear model.The genetic algorithm，which can overcome the disadvantage of some optimal algorithm that may obtain local optimal solution，is employed to solve the nonlinear program model.The simulation examples demonstrate the effectiveness of the model and its solution algorithm.

摘要： 多车道高速公路交通流的递阶控制结构分为协调控制层、路段控制层和车辆控制层，文中提出了路段控制层的优化控制模型。该模型以路段上交通流量与平均速度最大化为控制目标，以该路段上车辆在各车道的分配比例为决策变量。决策变量的个数比传统线性模型大为减少。为避免一般优化算法可能得到局部最优解的缺点，采用遗传算法求解所提出的非线性规划模型。仿真算例验证了模型及求解算法的有效性。