Rule-Based Approach to Real-Time Distributed Adaptive Signal Control

Most solutions for the problem of adaptive signal control involve optimization of an objective function. This paper presents a departure from this approach and describes a rule-based algorithm for effective distributed adaptive signal control of traffic networks. The control strategy, which has been termed GASCAP (Generalized Adaptive Signal Control Algorithm Project), uses queue estimates and a set of rules to determine the signal state at each intersection in the network. The GASCAP queue estimation algorithm uses upstream detector information to estimate the number of vehicles approaching an intersection and the vehicles in queue. The queue estimation algorithm is relatively insensitive to the extreme variations that may exist for cycle-to-cycle turning percentages because it uses the concept of a “partial” vehicle. The queue estimation logic and the control logic used to determine the signal state reside locally at each intersection and can be implemented as a distributed system. The signal control logic consists of a set of rules for uncongested control and an algorithm that creates a fixed time plan for congested control. The occupancy of the upstream detectors on opposing approaches are used to determine if a node is experiencing congestion. This control strategy has been tested with the TSIS/CORSIM simulation tool for three arterial traffic networks. The traffic conditions and network geometries for these arterials provide a wide range of challenges for any adaptive control strategy. Results from simulation indicate that GASCAP significantly reduced the delay and increased the throughput for each of these networks.