Multi-agent systems for energy efficient train and train station interaction modelling

Energy-efficient train control and timetabling are typical research problems in the field of rail transit. The multi-objective optimization of rail transit network under interference is always a challenge due to the excessive demand for computing power, which makes it difficult to realize multi-train optimization when the system is disturbed during practical operation process. In the present paper, a multi-agent system (MAS) is proposed to establish an interactive mechanism to reduce the energy consumption in rail transit field. In this system, every train and station is considered as an agent with certain autonomy. Through cooperation and negotiation, each train can adjust local time schedule according to specific situations. In these processes, several factors (such as passenger flow and energy consumption) are comprehensively considered to generate optimized solutions. The result of the case study shows that the energy efficiency is improved when a station has a high passenger density and longer boarding time is needed.