Traffic data acquisition by using GMR sensors

Intelligent Transportation Systems (ITS) aim to create a safe and efficient transportation infrastructure by utilizing computer and telecommunication technologies known as information technologies. One of the major applications of ITS is the real time traffic management and control process. The concept of sustainable transportation emerging from the traffic congestion that we encounter in modern days has raised a growing concern about traffic management and control processes.Thus, a sustainable transportation system can be achieved only by technological traffic management and control processes that measure and assess the parameters reflecting dynamics of traffic flows in real-time. In literature, a significant amount of research have been performed on measuring, monitoring, recording, and interpreting of the parameters.On the other hand, a data infrastructure enabling flexible production of long-term traffic observation information can be attained with the record of microscopic parameters obtained by this mean and macroscopic flow information derived from these parameters. Such an infrastructure can be regarded as having a databasewith indispensable importance in transportation planning activity?s perspective.Today, rapidly growing semiconductor technologies and embedded system solutions with high processing power enable small volume and low power (battery powered) systems to be developed.In this thesis, an embedded system design has been implemented with the intent of producing and recording real time traffic data depending on the facts stated above. Besides, applications, transferring data obtained through the embedded system into the database residing on the computer and implementing database management and analysis, and the interfaces accompanying these applications have been developed.Designed embedded system determines the microscopic flow information by means of processing signals produced by GMR (Giant Magneto Resistive) sensors located two-side of a circuit. The division of constant distance between sensors (15 cm) by phase difference calculated by the cross-correlation method of signals produced by these two sensors yields the speed of each vehicle. Also, the parameters of duration of the first sensor in the direction of the vehicles being occupied by vehicles (holding time) and the time between each subsequent vehicle passing over the first sensor are measured and recorded. The parameters reflecting macroscopic properties of traffic flow and vehicle lengths are calculated depending on these measurements.Test results present that vehicle speeds between 10 and 70 km/h can be calculated with an error less then ±%5 and vehicle lengths can be identified. Also data reflecting length of vehicles and time headway are given in the table.