• Smart City: A Traffic Signal Control System for Reducing the Effects of Traffic Congestion in Urban Environments

      Hardy, James (University of Derby, 2019-06-10)
      This thesis addresses the detrimental effects of road traffic congestion in the Smart City environment. Urban congestion is a recognisable problem that affects much of the world’s population through delays and pollution although the delays are not an entirely modern phenomena. The progressive increase in urbanisation and the numbers of powered road vehicles have led to an increasing need to control traffic in order to maintain flows and avoid gridlock situations. Signalised methods typically control flows through reduction, frequently increasing delays, holding traffic within the urban area and increasing local pollution. The current levels of vehicular congestion may relate to an increase in traffic volumes of 300% over 50 years while traffic control methods based on delaying moving traffic have changed very little. Mobility and Socio-economics indicate that the number of active road vehicles will increase or at least remain at the same levels in the foreseeable future and as a result congestion will continue to be a problem. The Smart City concept is intended to improve the urban environment through the application of advanced technology. Within the context of road transportation, the urban area consists of a wide variety of low to moderate speed transportation systems ranging from pedestrians to heavy goods vehicles. Urban roadways have a large number of junctions where the transport systems and flows interact presenting additional and more complex challenges as compared to high speed dual carriageways and motorways. Congestion is a function of population density while car ownership is an indicator of affluence; road congestion can therefore be seen as an indicator of local economic and social prosperity. Congestion cannot be resolved while there is a social benefit to urbanisation, high density living and a materialistic population. Recognising that congestion cannot be resolved, this research proposes a method to reduce the undesirable consequences and side effects of traffic congestion such as transit delays, inefficient fuel use and chemical pollution without adversely affecting the social and economic benefits. Existing traffic signal systems manage traffic flows based on traffic arrivals, prediction and traffic census models. Flow modification is accomplished by introducing delays through signal transition in order to prioritise a conflicting direction. It is incorrectly assumed that traffic will always be able to move and therefore signal changes will always have an effect. Signal transitions result in lost time at the junction. Existing Urban Traffic Control systems have limited capability as they are unable to adapt immediately to unexpected conditions, have a finite response, cannot modify stationary flow and may introduce needless losses through inefficient transition. This research proposes and develops Available Forward Road Capacity (AFRC), an algorithm with the ability to detect the onset of congestion, actively promote clearance, prevent unnecessary losses due to ineffective transitions and can influence other AFRC equipped junctions to ensure the most efficient use of unoccupied road capacity. AFRC is an additional function that can be applied to existing traffic controllers, becoming active only during congestion conditions; as a result it cannot increase congestion above current levels. By reducing the duration of congestion periods, AFRC reduces delays, improves the efficiency of fuel use and reduces pollution. AFRC is a scalable, multi-junction generalised solution which is able to manage traffic from multiple directions without prior tuning; it can detect and actively resolve problems with stationary traffic. AFRC is evaluated using a commercial traffic simulation system and is shown to resolve inbound and outbound congestion in less time than Vehicle Actuated and Fully Timed systems when simulating both morning and evening rush-hours.