| With economic development and social prosperity,almost each family can afford a car which is an indispensable means of transportation for people’s daily travel.However,with the increase in number of cars,the number of traffic accidents has also increased year by year,causing property losses and casualties.And a considerable proportion of traffic accidents is caused by the fact that there is a blind spot in the rearview mirror of the car,while the driver fails to make judgments about blind spot scenes correctly.To solve this problem,with the help of the Blind Spot Detection(BSD)system,the driver can be aware of the situation in the blind spot at any time,then make a reasonable judgment,which makes driving comfortable and safe.Therefore,it is of great significance to study the BSD system.Considering that the BSD system based on millimeter-wave radar has the characteristics of large detection range,high resolution,and strong anti-interference ability,and it still has a good working condition in bad weather such as rain and snow and low visibility environments,this thesis designs and improves a BSD system algorithm using 77 GHz millimeter-wave radar based on the linear frequency modulated continuous Sawtooth wave,combined with the classic algorithms of target distance measurement,speed measurement and angle measurement.Also,the feasibility of the algorithm is verified by using the DCA1000 EVM and IWR1642 embedded platforms.This thesis mainly carried out the following work:1.The structure and function of the BSD system are studied,and the overall structure of the BSD system,the range of blind spots in different situations and the signal processing flow are explained in combination with national standards and related documents.Besides,the parametric requirements of the BSD system are given.2.Introduce the classic ranging and speed measuring algorithm and the constant false alarm algorithm,select the 2D-FFT ranging and speed measuring algorithm and the OS-CFAR target detection algorithm for the BSD system,and design an improved OSCFAR algorithm with low complexity which reduces the amount of computation while ensuring the target detection rate.3.To solve the problem of low angular resolution of the classic angle-dimension FFT algorithm,this thesis uses the Unitary-IMUSIC algorithm to estimate the angle and improve the angular resolution based on the classic MUSIC algorithm.Besides,the thesis proposes an angle estimation matching algorithm based on the Unitary-IMUSIC algorithm to complete the matching of target distance and speed information,and the performance of the algorithm is verified by simulation experiments.4.Collect data based on DCA1000 EVM and IWR1642 embedded platform and then conduct actual measurement and verification.Also,this thesis designs a simple test environment,to verify the feasibility of the proposed BSD system algorithm including distance measurement,speed measurement,angle measurement and target detection algorithms,and designs a GUI interface to display the detection results. |
