Fuzzy Group Model of Traffic Flow

Simplicity and accuracy of models are crucial criteria of their construction in general and in road traffic systems especially. The quality of models is indeed a trade off between them because to complicate models are computational inefficient and can not be used in on line control and simulation of large systems. These goals fulfil the group model of traffic flow (Kaczmarek 1986), in which vehicle moving dependently are grouped in one object, so at the level of abstraction, models are enough detailed for control and simulation purposes. There are many non-linear relationships in traffic flow process, so the low of superposition does not be used. Application of aggregated input data of traffic parameters in models, as their mean values may produce substantial errors in output data as performance indexes. The aim of this work is to present a fuzzy version of the group model of traffic flow lack of the disadvantage. The paper consists of short presentation of group model, definition of fuzzy vehicle group and its fuzzy transformations.