Traffic signal optimisation in disrupted networks, to improve resilience and sustainability

Abstract

Transport has a critical role in economic development; an efficient transport system can enable economic growth and enhance social well-being. Road networks, as a part of a transport system, are among the most important lifeline systems. Urban road networks experience serious congestion because of infrequent major disruptions. Due to these disruptions the traffic system performance is reduced and the travel time and emissions in a road network are increased. This paper describes a method for optimising traffic signal settings (i.e. green times and offsets) to assist drivers to avoid partial or complete blockages, to minimise the travel time or carbon dioxide emissions in the case of disruptions in road networks. This method involves applying the Cross-Entropy optimisation method, along with a semi-dynamic approach, to optimise traffic signal control in disrupted road networks. This includes investigating two objective functions: minimising the travel time or minimising carbon dioxide emissions. To achieve this, short-term blockages are tested with different durations (a degradation for 4 minutes up to full hour) and level of capacity reductions (i.e. 0%, 25%, 50%, 75%, and 100%). The results for minimising the travel time, in the Cambridge (UK) network, show that applying signal optimisation reduces the travel time by almost 6% and carbon dioxide emissions by almost 8% for a complete capacity reduction at the most congested node, compared to not optimising the signal settings.