Details:

The PhD thesis deals with the problem of estimation of queue lengths in city traffic network. There are several types of traffic networks like historical centres, peripherals, arterial roads, etc. with specific characteristics. The traffic control needs the current traffic information regardless of type of roads. The quantities ideally describe the real situation on the road and suit the traffic control, is queue length, time delay, number of stops, etc. The target of the work is to get the possible best estimation of unmeasured traffic information of queue length. The queue length cannot be computed from basic traffic quantities directly, it has to be estimated from traffic model and from available data using a proper estimation method. The key problem, either for estimation or control, is to specify the sufficiently exact model of intersection or micro-region. New approach of estimation allows us to improve the traffic model, because the nonlinear estimation can work with nonlinear traffic model. The PhD thesis suggests hierarchic estimation, when we need stable as possible estimation . It allows us to use estimation of queue length of traffic control in real traffic systems.