高速公路交通流建模及匝道控制研究

【作者】 韦彦秀；

【导师】 梁新荣；

【作者基本信息】 五邑大学 ， 交通信息工程及控制， 2008， 硕士

【摘要】 准确的交通流模型不仅有利于理解车辆的行驶行为,而且对分析交通状况,规划交通网络和实现交通优化控制策略都有十分重要的作用。入口匝道控制通过调节进入高速公路的车辆数目可以使高速公路交通流运行在最佳状态,从而提高主线的通行能力,避免拥挤和阻塞。本文详细探讨了高速公路交通流建模及入口匝道控制。全文的主要工作如下:1.提出用小波变换消除交通噪声和干扰信号,用动态回归神经网络建立交通流模型。分析了高速公路宏观交通流模型,阐述了小波消噪方法,并用Elman回归神经网络对交通流动态建模,采用一种改进的算法得到神经网络的权值,最后对一条5路段,1个入口匝道和1个出口匝道的高速公路进行仿真。为了比较,还用BP神经网络和RBF神经网络对交通流建模。结果表明,Elman回归神经网络模型训练步数最少、误差最小、泛化能力最好。2.采用非线性反馈方法对高速公路进行入口匝道控制。建立了高速公路交通流动态模型,在此基础上,结合先进的PID控制设计了三种非线性反馈匝道控制器:免疫控制器、单神经元自适应控制器以及模糊免疫控制器,并分别用MATLAB对它们进行了系统仿真,仿真结果表明,非线性反馈方法具有良好的控制效果。3.采用递阶结构和CMAC进行高速公路多匝道控制。建立了反映高速公路交通流动态变化的宏观模型,研究了CMAC与PID复合控制算法。递阶结构分为两层:协调控制层负责计算各路段的期望密度轨线,直接控制层采用CMAC与PID复合控制决定各匝道的调节率,最后用MATLAB进行系统仿真。结果表明,与模糊逻辑控制相比较,复合控制具有更好的动态性能,更快的响应速度,该方法能有效地消除交通拥挤,实现车辆在高速公路上高效、安全地运行。更多还原

【Abstract】 Accurate model of traffic flow is important not only for better understanding of the collective behaviour of vehicles, but also for analyzing flow conditions, planning traffic networks, and designing efficient control strategies. On-ramp control can make traffic flow move in optimum conditions by regulating the number of vehicles entering a freeway entrance point, thus improve the passing capability of the mainline, and avoid traffic jams and congestion. Freeway traffic flow modeling and on-ramp metering have been presented and discussed in detail. The main contributions can be stated as follows.1. The wavelet transform is used to eliminate traffic noise and disturbance, and the traffic flow model is built based on a recurrent neural network in this paper. First, the freeway macroscopic traffic flow model is analyzed. Then the noise elimination method of wavelet transform is formulated, and Elman recurrent network is used for traffic flow modeling. The weights of the Elman network are obtained with an improved algorithm. Finally, a freeway with five segments, an on-ramp and an off-ramp is simulated. BP and RBF neural networks are chosen in contrast to the Elman network. The results show that the Elman network has the fewest training epochs, the smallest error and the best generalization ability.2. Nonlinear feedback methods are proposed for freeway on-ramp metering. The freeway traffic flow dynamic model is built. Based on the model and in conjunction with advanced PID control, three nonlinear feedback ramp controllers are designed: artificial immune controller, single neuron network self-adaptation controller, and fuzzy-immune controller. The above controllers are simulated seperately in MATLAB software. Simulation results show that nonlinear feedback is effective to the on-ramp control.3. The hierarchical structure and CMAC (Cerebellar Model Articulation Controller) are used for freeway ramp metering. The macroscopic model to describe the evolution of freeway traffic flow is first established. The algorithm of the composition controller of PID and CMAC is then studied. There are two layers in this control architecture: the coordinated control layer is responsible for computing the desired state of each segment, and the direct control layer is in charge of the ramp metering rate via the composition controller of PID and CMAC. Finally, the control system is simulated in MATLAB software and fuzzy logic control is also chosen in contrast to the composition control. The result shows that the composition controller improves evidently on the aspects of response speed and dynamic performance. The method can effectively eliminate traffic jams, andmake vehicles travel more efficiently and safely.更多还原