Vehicle-mounted millimeter wave radar non-direct-view front vehicle detection method

Abstract

The invention discloses a vehicle-mounted millimeter wave radar non-direct-view front vehicle detection method, which is applied to the technical field of non-direct-view target detection and positioning and solves the problem that a non-direct-view front vehicle target detection scheme is lacked in the prior art, and comprises the following steps of: detecting and positioning a non-direct-view front vehicle target in a road environment by utilizing a millimeter wave radar; the method comprises the following steps: firstly, according to a propagation phenomenon of electromagnetic waves, analyzing a multi-path signal propagation path in a front vehicle detection scene on a road; then establishing an echo signal model based on echo time delay of a ground primary reflection path; and finally, target constant false alarm detection and a multi-channel phase comparison positioning algorithm are applied to the echo signal to obtain a positioning result of a front vehicle target. According to the scheme, the electromagnetic propagation path for detecting the front vehicle target can be truly determined, and the function of detecting and positioning the front vehicle target is effectively achieved.

Description

technical field [0001] The invention belongs to the technical field of non-direct sight target detection and positioning, and in particular relates to a front vehicle target detection and positioning technology in a road non-direct sight scene. Background technique [0002] In road scenarios, traffic accidents caused by vehicle collisions will cause a large amount of property damage and casualties. A worst-case scenario is multiple vehicle collisions. Currently, on-board collision avoidance systems (CASs) improve vehicle safety by detecting and predicting the motion characteristics of vehicles ahead in the Line of Sight (LOS) area. However, the existing collision avoidance system cannot detect the motion state of Non-Line of Sight (NLOS) vehicles, and cannot avoid the occurrence of multiple vehicle collisions. Normally, detecting non-line-of-sight vehicles is a challenging task, especially in complex road scenes. Therefore, the detection of non-line-of-sight vehicles has ...

AI Technical Summary

Problems solved by technology

In 2014, scholars from the Swedish Defense Research Agency applied X-band radar to detect NLOS moving vehicles behind corners by using multi-path echoes of the primary reflection path of the wall (T.Johansson, A.Orbom, A.Sume, et al.Radar measurements of moving objects around corners in a realistic scene[C].in Radar Sensor Technology XVIII,2014,vol.9077.), the actual measurement experiment confirmed that the non-direct sight vehicle can be detected after the corner, but the experimental results are mainly used in the urban construction environment , and is limited by the multiple attenuation of the reflection path of the building wall, so the range of detecting non-direct sight vehicles is very limited

This method needs to set the best reflective surface for non-line-of-sight vehicle detection. This research can only be applied to specific intersections, and cannot bring universal help to the detection of non-line-of-sight vehicles.

In 2019, scholars from the University of Hong Kong specifically studied the problem of non-direct sight front vehicle detection in road environments (Z. Zhang, S.W. Ko, R. Wang and K. Huang. Millimeter-Wave Multi-Point Vehicular Positioning for Autonomous Driving[C] .2019IEEE Global Communications Conference, Waikoloa, USA, 2019, pp.1-6.), the article uses the NLOS front vehicle target as the transmitter, and sets up a cooperative receiving signal vehicle, and then uses the side car reflection path to detect the front vehicle target For multi-scattering point positioning, this method can obtain the position of the vehicle in front, but it requires a one-way clock synchronization between the vehicle receiving the signal and the transmission signal of the vehicle in front of NLOS, and the article does not rely on actual measurement experiments to verify the effectiveness of the method. In summary The method described is not suitable for non-cooperative detection of the vehicle in front of the real road environment

Images