城市交通控制智能优化配时及仿真

城市交通系统是一个随机性很强的、复杂的巨型系统。开发了双层反馈神经网络，将其应用于城市交通控制信号配时优化，并开发了应用于优化计算的能量函数和车辆平均延误计算式。以广东江门市某平面交叉路口为对象进行了仿真计算，结果表明：与传统的配时方法相比，采用所开发的双层反馈神经网络进行智能优化配时，交叉路口车辆的平均延误可以平均减少18．2％，可以大大提高路口的通行效率。开发的双层反馈神经网络也可以应用于其他对象的优化，具有较强的推广应用价值。

Urban Traffic Signal Timing Optimization and Simulation Based on Artificial Intelligence

Urban traffic system is a complicated and huge system with strong randomicity. A double-layer Artificial Neural Network with feedback was developed, and was effectively used to tackle the optimization of urban traffic signal timing. Then an energy function and an equation on the average delay per vehicle for optimal computation were developed. Simulation research was carried out based on the intersection in Jiangmen city of Guangdong province in China, which indicates that urban traffic signal timing optimization based on the double-layer feedback network could reduce 18.2% of the average delay per vehicle in intersection than that based on the conventional timing means. The double-layer feedback Artificial Neural Network could also been used in other fields.