| The intelligent vehicle is a research hot spot in the field of modern automobiles,and one of the effective ways to solve social challenge issues such as driving safety,fuel economy and environmental friendliness,and for shared ride,comfortable ride,efficient service and ideal entertainment.Intelligence is an inevitable trend in the development of automobiles.In the process of intelligent driving research and development,virtual testing technology can effectively solve the problems of long cycle,high cost,low security,and difficult feasibility in real road testing.It has become a key technology for intelligent car R&D testing.Virtual testing will also become an essential part of the evaluation certification for the future intelligent car.In the process of intelligent driving research and development,millimeter wave radar is more mature,more widely used and cheaper than laser radar.Compared with visible light camera,accuracy and stability are better.Compared with the visible light camera,the accuracy and stability are better,and the millimeter wave radar has short wavelength,wide frequency band and narrow beam,which can achieve high-precision measurement of the distance,speed and angle of the measured target.Therefore,millimeter wave radar is an indispensable sensor for intelligent cars,and the intelligent driving system based on millimeter wave radar has broad prospects.In order to effectively carry out simulation testing of intelligent car millimeter wave radar,millimeter wave radar modeling and simulation research is an urgent research topic.In the existing research results of intelligent vehicle simulation test,the model of intelligent vehicle millimeter wave radar has the disadvantages that “the simulation output results tend to be idealized,and the real radar detection mechanism and detection defects cannot be objectively reflected,and the simulation fidelity is low”.At the same time,there are difficult problems such as "the research on modeling of radar environment clutter tends to be simplified and cannot dynamically change with the space scene","the radar target RCS value is fixed,non-real-time,and cannot change dynamically with the incident direction ofthe radar".This paper takes this as the starting point for research and simulation of millimeter-wave radar based on automotive intelligent driving.Firstly,comprehensively analyze the mechanism of the detection defect problem of intelligent vehicle millimeter wave radar.The emission mechanism of radar electromagnetic waves is studied,and the detection defects caused by radar emission waveform and radar transmitting antennas and their root causes are analyzed.The propagation mechanism of radar electromagnetic waves is studied,and the detection defects and root causes caused by radar internal transmission attenuation and external environmental propagation attenuation are analyzed.The reflection mechanism of radar electromagnetic waves is studied,and the detection defects and the root causes caused by radar target RCS and radar environment clutter are analyzed.The mechanism of receiving and processing radar electromagnetic waves is studied.The detection defects and the root causes of radar receiving antennas,radar receiver characteristics and radar echo processing methods are analyzed.Secondly,this paper proposes a clutter modeling method for millimeter-wave radar environment for automobile intelligent driving.Aiming at the features of the intelligent driving traffic scene,the time-frequency statistical distribution characteristics of the features in urban traffic roads,high-speed traffic roads and rural traffic roads are analyzed.A ground clutter modeling method is proposed,and the distribution characteristics of ground objects in typical scenarios are designed.Based on the millimeter-wave radar signal processing and data processing algorithms,the application verification of ground clutter modeling method is carried out,which confirmed the effectiveness of the ground clutter modeling method.According to the weather characteristics in the intelligent driving traffic scene,according to the time-frequency statistical distribution characteristics of the weather such as rain,snow,and fog,the weather particle size,inter-particle density,wind speed,and wind direction are distinguished.A modeling method for weather clutter such as rain,snow,and fog is proposed,and typical distribution characteristics of rain,snow,and fog are designed.Based on the signal processing and data processing algorithms,the application verification of the weather clutter modeling method is carried out,which confirmed the effectiveness of the weather clutter modeling method.Aiming at the potential human interference in intelligent drivingtraffic scenarios,a modeling method of artificial active interference for intelligent vehicle millimeter wave radar is proposed,and a typical human interference scenario is designed.Based on the signal data processing algorithm,the application verification of the artificial interference modeling method is carried out,which confirmed the effectiveness of the artificial interference modeling method.At the same time,a real-time estimation method of millimeter-wave radar target RCS for intelligent driving of vehicles is designed.Aiming at the characteristics of radar targets in the intelligent driving traffic scene,factors such as target size distribution,material composition,and shape characteristics are optimized.Based on the propagation characteristics of electromagnetic waves in the millimeter wave band,and assisted by the analysis of FEKO electromagnetic simulation software,a real-time estimation process of RCS for intelligent driving millimeter wave radar targets is designed.The ARD estimation method of the target RCS is proposed,and the calculation method of the size factor,shape factor and material factor of the radar target is designed in detail.Aiming at vehicle and pedestrian targets,verification tests of RCS estimation methods of vehicle target stationary state,vehicle target motion state,pedestrian target stationary state,and pedestrian target motion state are carried out.By comparing the FEKO electromagnetic simulation software with the estimation method,the validity of the real-time estimation method for the RCS of the intelligent driving millimeter-wave radar target is verified.Thirdly,based on the radar target detection mechanism,clutter modeling method,and target RCS estimation method,a simulation model of intelligent driving millimeter wave radar is established.Based on the statistical probability theory,the probabilistic distribution characteristics of false alarms are analyzed and a false alarm modeling method for intelligent driving millimeter wave radar is proposed The modeling method of "false alarm detection defects" of intelligent vehicle millimeter wave radar is proposed.Based on the quantitative analysis of air-space occlusion between radar and target,energy-domain attenuation and frequency-domain occlusion between radar and environment,a method of false-reporting modeling for intelligent driving millimeter-wave radar is proposed.Based on a comprehensive analysis of the effects of radar internal and external factors on ranging,speedand angle measurement,a modeling method for ranging,speed,and angle errors of intelligent driving millimeter-wave radar is proposed.Finally,the application research of radar model in intelligent vehicle simulation test is carried out.The AEB algorithm is designed,the testing scene of AEB algorithm is constructed,and the AEB algorithm testing based on radar detection defect simulation is carried out.At the same time,the ACC algorithm is designed,the testing scene of ACC algorithm is constructed,and the ACC algorithm testing based on radar detection defect simulation is carried out.The testing results confirm that the simulation testing based on the radar model of this paper can effectively test the potential defects in the AEB algorithm and the ACC algorithm,and provide important technical support for the effective improvement of the algorithm. |
PDF Full Text Request