Selection of range and azimuth angle parameters for a forward looking collision warning radar sensor

This paper develops an algorithm for specifying range and azimuth angular coverage of a radar sensor. The algorithm combines geometric and accident data analysis of straight and curved roads with worst-case horizontal curvature. A bounded area for the selected collision warning algorithms (stopping distance and closing rate) is defined for a particular set of vehicle decelerations, initial speeds and delay time. These parameters are then varied by a Monte-Carlo technique. The bounded area is then integrated with the accident data for generating the number of potential accidents per specified sensor range. Results based on computer simulations show that the stopping distance algorithm is better than the closing rate algorithm in giving warning to the driver. The sensor´s range and the azimuth angular coverage is specified, assuming range gating on curved roads is implemented