论文部分内容阅读
为适应交叉口多态特征交通流,采用期望最大化算法,选取稠密性混合Gamma分布函数,准确拟合交叉口到达车头时距,统计交叉口进口到达交通量概率.以两相位全感应控制交叉口为例,根据交通量统计概率和信号相位状态获得每相位最大排队长度,并由此提出交叉口感应控制延误模型.在混合Gamma分布拟合精度不低于95%条件下的计算实例表明,基于混合Gamma分布的感应控制延误与实际结果更为吻合.
In order to adapt to the multi-modal traffic flow at the intersection, the expectation maximization algorithm is adopted to select the dense mixed Gamma distribution function to accurately fit the arrival time at the intersection and to estimate the traffic arrival probability at the intersection. As an example, the maximum queuing length per phase is obtained according to the statistic probability of traffic volume and the signal phase state, and the delay model of intersection sensing control is proposed.With the fitting accuracy of mixed Gamma distribution not less than 95% The delay in inductive control based on the mixed Gamma distribution is more in line with the actual result.