道路拥挤收费被认为是城市交通管理和控制的一个有效方法,国内外不少学者提出了模型和相应的计算方法。文章根据弹性需求下的拥挤收费模型,利用BP神经网络算法模拟路段流量和路段收费之间的对应关系,并在神经网络的训练过程中引入遗传算法,加速神经网络的全局收敛。通过训练后的神经网络制定收费,可以使得各路段流量基本达到最优路段流量的要求,并通过实例证明了该算法的有效性。
Congestion Pricing is an effective way to manage and control the urban traffic. Many toll model and algorithm had been proposed both at home and abroad.In this paper, neural network is used to simulate the function of road toll and flow, so as to solve the congestion pricing model which takes elastic demand into account. The neural network is trained via genetic algorithm and its convergence is accelerated. The toll which is got from the trained neural network can make the road flow close to the: optimal flow.The algorithm is proved effective from an example.