Research On Vehicle Passenger Detection Technology Based On Millimeter Wave Radar

At present,with the rapidly increasing of the number of automobiles and more complex usage scenarios of automobiles,various safety accidents appear in large numbers.With the continuous development of automobile intelligence,vehicle occupant detection technology has become a hot spot for further protection of passengers.The technology of detecting the occupants inside the cars has risen to the same important position as detecting the environment outside the car.Meanwhile,major automakers and sensor manufacturers are making great efforts in this field.This paper studies the detection technology of the car members based on 77 GHz millimeter wave radar.First,the design of vehicle occupant detection system based on FMCW radar is introduced,including the system framework,work flow of FMCW radar and the method of solving target information.Based on the detection scene in the car,the related parameters of the millimeter wave radar system are set.Secondly,aiming at the recognition problem of millimeter wave radar in ultrashort distance dense targets in the vehicle,the radar signal processing framework,DOA and CFAR are studied.After analysis,the signal processing framework based on the "distance-angle" is chosen.According to the consideration of comprehensive performance and calculation amount,minimum variance spectrum estimation is chosen to DOA.In the field of CFAR detection,OS-CFAR can effectively suppress the "masking effect" of dense multi-targets.Then,the realization of the vehicle occupant detection system based on millimeter wave radar is introduced.A 77 GHz millimeter-wave radar sensor is designed based on AWR1642,and the hardware circuit and antenna design are optimized for the in-vehicle environment.A set of radar signal processing flow is designed,and the implementation of signal processing on the chip is introduced in detail.Finally,the vehicle occupant detection system based on the millimeter-wave radar was tested in the indoor and in-vehicle environment respectively,and the results showed that the system could accurately detect the occupancy of passengers in the area.