Traffic Signal Control System

Definition and System Positioning

The traffic signal control system forms a vital part of urban traffic management infrastructure. Its primary objective is to rationally allocate road space and time through the co-ordinated control of traffic signals at junctions, minimising vehicle stops and travel delays to maximise the overall operational efficiency of the road network. By monitoring traffic volume in real time and dynamically adjusting signal timing plans, the system ensures optimal traffic flow whilst safeguarding road safety.

System Architecture and Composition

A traffic signal control system comprises four core functional modules: vehicle detectors, junction signal controllers, communication transmission systems, and a centralised networked control system. Vehicle detectors capture traffic data—such as volume, vehicle speed, and lane occupancy—from all junction approaches, providing the necessary data inputs for signal optimisation. Serving as the field execution unit, the junction signal controller can execute standalone local optimisation based on on-site detection data, carry out co-ordinated control commands issued by the central system, or interface with adjacent junction controllers to achieve regional co-ordination. The communication transmission system handles data exchange between detectors and controllers, as well as between controllers and the traffic command centre. Meanwhile, the central control system oversees comprehensive functions, including surveillance, data management, timing plan design and library administration, and overall system maintenance.

In terms of control hierarchy, the system typically adopts a three-tier architecture—junction level, regional level, and central level—establishing a cohesive operational framework that scales from single-junction control to regional co-ordination and overall network-wide optimisation.

Classification of Control Methods

Traffic signal control systems are broadly categorised by their control strategy into fixed-time control and adaptive control. Fixed-time control relies on historical traffic data for offline timing optimisation, storing pre-configured signal plans in the controller for activation across different time slots, making it best suited for road networks with predictable and stable traffic patterns. Adaptive control leverages real-time traffic data captured by roadside detectors to perform online calculation and timing optimisation, dynamically responding to stochastic fluctuations in traffic flow.

By control architecture, systems are categorised into centralised and hierarchical control modes. Centralised control relies on one or more central servers to execute unified scheduling across all signal controllers in a designated zone; while conceptually straightforward, it places high processing demands on the central system. Hierarchical control splits the architecture into upper and lower tiers: the upper tier handles macro-level co-ordination and optimisation, while the lower tier executes and fine-tunes timing at individual junctions, working in tandem to deliver granular traffic management from a network to a local scale.

Modern Evolution: From Fixed Timing to AI-Driven Adaptive Control

Traditional signal control systems predominantly utilise time-of-day fixed timing plans, running on predetermined cycle times and green splits that struggle to cope with real-time traffic fluctuations. In recent years, adaptive control systems powered by artificial intelligence (AI) and real-time sensing technologies have rapidly emerged as the industry mainstream.

Taking a pilot project in Jianye District, Nanjing, as an example, traffic authorities fitted 36 junctions with radar and video detectors, enabling precise detection of traffic flow variations within a 200-metre radius of each junction. Concurrently, signal control algorithms were upgraded to shift from isolated junction management to multi-junction regional co-ordination. Equipped with 14 built-in algorithmic models and dozens of scenario presets, the system automatically adapts its timing strategies to cater to scenarios such as school runs, morning peak hours, and late-night traffic. Official data revealed an 18.8% reduction in the morning peak congestion index across the area.

In Quanzhou, Fujian, AI-enabled smart traffic lights have pioneered a shift from "cars waiting for lights" to "lights responding to cars". Utilising an all-visual sensing module, the system monitors traffic volume and tailback queues across all approaches in real time, executing timing adjustments in mere seconds; once side-road traffic has cleared, the green signal terminates immediately to grant extended green time to the main road. Driven by real-time demand, signal cycle lengths are no longer rigid, with each phase and cycle generated dynamically based on instantaneous traffic conditions.

Technical Limitations

The efficacy and real-time responsiveness of traffic signal control systems hinge heavily upon sensor deployment density, data precision, communication network reliability, and the computing power of the central platform. In the event of hardware failure, communication loss, or extreme weather conditions, the system must fall back to a failsafe fixed-cycle mode. Furthermore, safeguarding non-motorised road users—including accommodating pedestrian crossing phases and non-motorised vehicle flow—introduces critical constraints to the design and execution of control strategies.