Konferenční příspěvek

3rd International Workshop on Data - Algorithms - Decision Making

ÚTIA

Praha

2007

Ing. Pavla Pecherková (ÚTIA - Oddělení adaptivních systémů)

traffic control, state space model

Intelligent traffic control is one possible way how to improve capacity of light controlled network. Basic principle of the traffic-flow control is based on minimization of total queue lengths in a micro-region [6, 4]. Measurement of the queue lengths is difficult, so it is necessary to estimate them. The traffic micro-region can be described by a linear state space model (SSM) [6] where state is just formed by queue lengths. The SSM includes some unknown parameters which need to be estimated together with the states. Such the estimation requires the use of non-linear estimation techniques [2, 5, 1, 3]. The traffic model is based on traffic flow conservation principle: the queue at time k is equal to the sum of the previous queue at time k − 1 and the arrived cars, given by the intensity and the sample period, minus the number of cars which pass through the intersection. The number of cars depends on the existence of queue.