A Research Of The FMCW Automobile Anti-collision Radar System Based On DSP

The automotive anti-collision radar system is an important part in intelligent transportation.The use of such system can effectively reduce the incidence of traffic accidents and it is one of the effective methods to improve road traffic safety problems.So the research on the automotive anti-collision radar system has a wide development prospect and important practical significance.Taking CW with frequency modulation in Millimeter wave band as a research object,this thesis gives due importance to the study on the ranging algorithm of signal processing module and the whole software construction and implementation,moreover,their experimental verifications have been done on the DSP platform.The main contents and the innovation points of the thesis are as follows:Firstly,the thesis summarizes the research background and significance in the field of automotive anti-collision.It also summarizes the researches at home and abroad and some of the existing problems.It determines the radar working system and target detection methods.Furthermore,it introduces the efficient automotive anti-collision mathematical model,and the effectiveness of which has been theoretically analyzed.Besides this,the working principle of the automotive anti-collision radar system has been deeply analyzed.Those successfully provide theoretical basis for the research and implementation of signal processing algorithms.Secondly,the frequency estimation is an important content of the ranging function of the automotive anti-collision radar system.In order to improve the ranging accuracy of the radar system,based on the introduction of Rife frequency estimation algorithm,the thesis analyzed the root-mean-square error of the Rife algorithm that was in different frequency range.Besides,the frequency spectrum shift was done in regular area of the beat signal,which ensured that the estimating frequency is in the centre of the sampling frequency area and proposed the high precision distance measurement method based on IRife algorithm.The efficiency of the method has been proved through experiments.Experimental results show that the measuring performance of the improved method has greatly improved.Thirdly,in order to reduce the ratio of false alarm of the system and improve the accuracy of the target recognition,the thesis theoretically analyzed the working principle of detection method and false alarm probability of several frequently used Constant False Alarming Rate,CFAR.Furthermore,it also determined the CFAR as the technological basis of the system and verified the effectiveness of the proposed algorithm.Fourthly,the thesis completed the modual of signal process.It first introduced the characteristics,advantages and the project development platform of the DSP chips that was used.And then,it studied the realization process of software system,which included the software construction and the initial of the implementation process.What's more,it analyzed the working principle of each device in the system and configured parameters that were needed in the project.Then on the experimental development platform it realized two kinds of ranging methods in this system,and four groups of different distance intervals are selected to do the experiments and analysis.It verified the improved methods are better than the previous ones.