| The function of adaptive cruise control system has been greatly extended since the development of the conventional cruise control system from the basic function of cruise control speed under a specific road scene and a specific speed range to the current stop&go function under more complex road scene and full speed range.Its function gets more and more perfect,as well as its application road scene is more and more complicated.In order to adapt to the development of adaptive cruise control system,this paper studies the forward key target recognition algorithm based on millimeter-wave radar sensor in the multi-target scene from the following aspects.Firstly,the rise background and development history of adaptive cruise control system is introduced.And then,the significance of target recognition algorithm and what problem it need to solve in adaptive cruise control system is analyzed.Some typical target recognition algorithm research based on millimeter wave radar at home and abroad is respectively introduced in detail,and algorithm based on other sensors is also introduced briefly.Secondly,the basic coordinate system and the system coordinate update rules is established,and the coordinate transformation formula related to the trajectory fitting and the coordinate transformation formula from radar coordinates to coordinate transformation formula are derived.Then,the working principle of radar and some information about ESR radar are introduced.Thirdly,a 3-DOF dynamics model,including lateral and longitudinal speed and yaw rate,is established in Matlab/Simulink environment,through the co-simulation of Carsim and Matlab/Simulink,the accuracy of the established model was verified.Based on the 3-DOF model,and by application of unscented kalman filter algorithm,the state estimator for accurately obtaining the host vehicle lateral velocity,longitudinal velocity and yaw rate is developed.The accuracy and validity of the estimator are verified by simulation.Fourthly,The UKF algorithm is respectively applied to the CA model and the CTRV model to establish the target vehicle motion state estimator.By establishing the relative motion scene in the Carsim environment and making co-simulation of Carsim and Matlab/Simulink software,the accuracy of the UKF state estimator estimator is validated.Lastly,An effective target recognition algorithm based on moving target spatial sequence fitting and time series fitting is proposed.In the algorithm,support vector machine theory is applied to radar target classification,and a trend selection strategy is proposed to screen the same direction moving targets.Finally,the feasibility of the algorithm is verified based on real vehicle test radar data.The results obtained in this paper are of great theoretical and practical significance for the further development and improvement of the ACC system. |
PDF Full Text Request