360°-Environment sensing and signal processing for an automotive pre-crash application

In this paper a new and flexible 360°-environment sensing system using short range radar sensors and active seat belt pretensioners is presented. To detect critical targets from a wide variety of closing angles, a total of only four radar sensors is used. The number of sensors can be changed, without altering the underlying sensor fusion and target detection algorithms. The challenges of pre-crash systems are presented and the multi-sensor fusion and signal processing algorithms are shown. To be able to react to highly dynamic driving maneuvers, an interacting multiple model filter (IMM) is used. A pre-crash algorithm for 360°-environment sensing is introduced. To verify the system effectiveness, a simple, yet effective, system test has been developed. Test data is gained and evaluated for different critical situations. The presented pre-crash system serves as a basis for new driver assistance and safety functions, which helps to increase overall driving safety, even in dense traffic and urban driving situations.