Research On Key Technology Of Millimeter-Wave Radar For Intelligent Drone

With the development of intelligent technologies,drones are used in military and civilian fields due to their advantages of low cost and high mobility.In the numerous research of drone technology,autonomous safe landing is the key link to guarantee the successful completion of mission,and has received extensive attention from scholars.In order to realize the autonomous safe landing of drones,an effective solution is to use sensors to detect the landing environment.With the advantages of all-weather,low-cost,small size,and high precision,the millimeter radar can provide a powerful guarantee for the safe landing of drones.Therefore,it is of great significance to conduct in-depth research on millimeter-wave radar technology.In this paper,the basic theoretical knowledge of LFMCW radar signal processing is introduced in the first place.It also analyzes several typical algorithms to improve the precision of the radar,including zero padding method,ZFFT,Chirp-Z,and the FFT-DTFT transformation method.At last,the ZFFT algorithm is chosen for its fitness of the hardware platform by the algorithms simulation,as well as the contrast analysis of required storage space and the amount of calculation.Based on the implementation of signal processing algorithm by millimeter wave radar sensor,the measurement function about distance and speed is realized on the platform.In the meanwhile,the ZFFT algorithm is achieved as well.Experiments are carried out to test then.Firstly,the collected data is transmitted by the first 1Kpoint fast Fourier transform(FFT),16K-point FFT transform and ZFFT transform processing,and is made the online comparison by using ZFFT algorithm.The processed results show that the precision is greatly improved by using the ZFFT after the transformation,which can be accurate to the millimeter level.Finally,a set of suitable parameters suitable for radar platform are designed and the maximum unambiguous distance exceeds 40 meters.At last,this paper researches the identification of landing surface of drone.Considering the phase of LFMCW radar about intermediate-frequency signal characteristics which is sensitive to small distance with the difference of diffuse reflection and mirror reflection about echo intensity,the research analyzes the land surface feasibility by using the above two characteristics.Moreover,the Support Vector Machine(Support Vector Machine,SVM)is utilized to train and identify the experimental data and to distinguish.The test results show that the recognition rate is in a position to more than 90% in the use of the above two characteristics.