Design and analysis of segmented sensor for road traffic classification

Individual vehicle information, specially, vehicle classification data play a key role in road traffic measurement. In inductive loop and piezoelectric sensor systems, traffic data such as number of axles and the distance between axles are used for vehicle classification. However, classification errors occur in distinguishing class 3 (long bed pickup truck) and class 5 (small truck) vehicles that are moving at highway speeds. In order to overcome these classification errors and improve the durability, a new extruded polymer segmented sensor approach is developed.;Segmented sensor has an additional ability to provide an estimate of the tire width to discriminate between different tire types and classify vehicles accurately. The extruded polymer conductors used in the existing prototype segmented sensor have a very high resistance that was found to limit the frequency response of the sensor. Segmented sensor with limited frequency response could not always properly identify the closed segments for vehicle moving at highway speeds. Several analysis and simulations were conducted to improve the performance of the extruded segmented sensor. In this thesis, previous efforts conducted in designing interface electronics of segmented sensor and continued efforts on designing an accurate simulation model for extruded polymer segmented sensor with its detection electronics are discussed.;Contribution to the current research work on extruded segmented sensor include validating Dr. Bourner's 3x3 segmented sensor model, applying the 3x3 segmented sensor approach to existing extruded polymer segmented sensor and comparing the results with measured responses of existing 16-segmented extruded polymer sensor to develop an accurate simulation model. The cross-sectional dimensions are then increased to get an improved frequency response. Further 16-segmented extruded model is extended to a 49-segmented simulation model based on the foot print results provided by Dr. Ren Moses. Also the final model with increased cross-section is simulated to attain the required bandwidth needed to design a new prototype of extruded segmented sensor that can classify vehicles moving at highway speeds.