基于NS-BML模型的记忆密度在交通信号灯控制系统中的研究

交通信号灯管理与控制直接影响着交通网络中的运行效率。NS-BML模型广泛应用于交通信号灯控制系统仿真，针对目前NS-BML模型中只考虑现在瞬时密度而忽略历史密度的问题，提出记忆密度策略，从长时记忆密度策略和短时记忆密度策略两个角度来分析该策略对曼哈顿式网络的影响，通过对时间离散化，求解短时记忆密度的最优比例因子。仿真结果表明，通过采用所提出的短时记忆密度策略可以有效提高系统的运行效率，同时保证计算机处理的速度，交通网络的平均速度和到达率分别同比增长8.51%和9.28%，说明了该策略的有效性。

Memory density research in traffic signal control system based on NS-BML model.

Traffic signal management and control directly affect the efficiency of traffic network. The NS-BML model is widely used in the simulation of traffic signal control system. For the problem of the current NS-BML model only considered the present instantaneous density but neglected the history density, this paper proposed the memory density strategy. This strategy analyzed the influence of the proposed strategy to the efficiency of Manhattan network from two perspectives, the long memory density strategy and the short memory density strategy. It used time-discretization obtained the optimal proportion factor of the short-term memory density. The simulation results show that the efficiency of the system can be improved effectively by using the proposed short memory density strategy. At the same time, it guaranteed the efficiency of computer processing. The average speed and the rate of arrival of the system are increased by 8.51% and 9.28% respectively, indicating the effectiveness of the proposed strategy.