一种面向时间分配问题的群智能劳动分工新方法

A new approach to labor division in swarm intelligence for time allocation problem

本文以给不同信号相位的车辆分配绿灯时间的交通信号配时问题为代表,将群智能劳动分工应用到时间分配问题的求解中,提出一种新颖的蜂群劳动分工算法(bee swarm labor division algorithm, BSLDA)。首先从时间分配的视角对交通信号配时问题进行分析,然后将激发抑制原理引入 BSLDA,为每个信号相位定义了激发剂和抑制剂,并设计了增加绿灯时间、减少绿灯时间和保持绿灯时间 3种行为。在群智能劳动分工激发抑制原理作用下,BSLDA中的每个信号相位都能根据环境变化选择恰当的行为完成时间分配。最后采用真实的交通流数据进行仿真实验,结果表明本文方法适于求解不确定环境下的交通信号配时问题。

In this paper,we use the labor division in swarm intelligence to solve the time allocation problem represented by the traffic signal timing problem of allocating green light time to signal phases,and propose a new bee swarm labor division algorithm (BSLDA).The traffic signal timing problem is analyzed from the perspective of time allocation,and the activator-inhibitor mechanism under labor division is introduced in BSLDA.For each signal phase,BSLDA defines one activator,one inhibitor and three behaviors (i.e.,increasing green light time,reducing green light time and keeping green light time).With the activator-inhibitor mechanism,each signal phase in BSLDA could choose appropriate beha-vior according to environmental change to achieve the time allocation.The real traffic flow data is used in the simula-tion experiment,and the results show that the proposed approach is effective and suitable for the dynamic traffic signal timing problem in uncertain environment.