Abstract: Taking the cross for example，on the basis of the relationship between exit traffic flow，entrance trafic flow and turning flow．the objective function Was fixed through the minimum Shill of absolute value of the diferences between the meas ured value of entrance trafic flow as well as the sum of computed value of entrance trafic flow and the differences between the meassured value of exit trafic flow and the sum of computed value of exit trafic flow．By considering the range of turning flow as constraint conditions，a multiple objectives programming model was built up to solve the turning flow problem．Then， taking use of the parallel computing of genetic algorithm，a method to solve the turning flow problem Was put forward by using multiple objectives programming model based on genetic algorithm．Finally，the effectiveness of this method Was illustrated by an example．

Key words: turning flow, Cross, multiple objectives programming, genetic algorithm

摘要： 以十字交叉口为例，在分析交叉口进、出口流量与转移流量之间关系的基础上，以进口交通流量观测值与来自 各个出口并进入该进口的估计转移流量之差的绝对值和出口交通流量观测值与该出口到达各个进口的估计转移流 量之差的绝对值之和最小为目标函数，转移流量的实际取值范围作为约束条件，建立了求解转移流量的多目标规划 模型．利用遗传算法并行运算的特性，提出了基于遗传算法求解转移流量的多目标规划方法．最后通过～个算冽给 出了应用该方法求解转移流量的多目标规划模型的求解效果．

关键词: 转移流量, 十字交叉口, 多目标规划, 遗传算法.