Research On Micro-doppler Signal Processing Algorithm Of Multi-rotor UAV

In recent years,with the increase of the number of multi-rotor unmanned aerial vehicles(UAVs),the safety problem caused by multi-rotor UAVs has become more and more important.Multi-rotor UAVs are always characterized by low altitude,slow speed and small size,so it is difficult to perform accurate detection using traditional energy-based signal processing methods.The rotation of the UAV rotors modulates the radar echo and generates the micro-doppler signal.Through analysis of micro-doppler signals,the parameters of the rotors can be obtained,which can achieve detection of multi-rotor UAVs.This thesis focuses on the signal processing algorithms of micro-doppler signals for multi-rotor UAV,and studies noise reduction,high time-frequency resolution representation and rotor parameter estimation algorithm of micro-doppler signal,and in addition,the design of the corresponding micro-doppler signal processing system is carried out.The main work is as follows:1.Aiming at the characteristics of low-SNR and wide frequency modulation range of the micro-doppler signals of multi-rotor UAV,the necessity of noise reduction of multi-rotor UAV’s micro-doppler signal is analyzed,and based on the theory of signal decomposition and reconstruction,an EEMD-SVD noise reduction algorithm is proposed.The algorithm combines the EEMD algorithm with the SVD algorithm,and uses the noise energy ratio in the EEMD decomposition as the threshold for IMF screening and SVD singular value screening,which effectively avoids modal aliasing and can also retain the frequency modulation characteristics of multi-rotor UAV’s micro-doppler signal.At the same time,noise reduction in the range of signal frequency modulation is realized.2.Aiming at the problem of low time-frequency resolution in the processing of microdoppler signals by traditional time-frequency analysis algorithms,based on the theory of synchrosqueezing and reassignment method,a short-time oblique synchrosqueezing transform algorithm is proposed.The algorithm introduces the second-order synchrosqueezing operator and implements the oblique synchrosqueezing transform algorithm on the basis of short-time analysis.Therefore,the algorithm has better time-frequency distribution resolution and instantaneous frequency positioning capability for multi-rotor UAV’s micro-doppler signal,and also has better anti-noise performance.3.Aiming at the need of precise detection of multi-rotor UAV,based on cadence frequency spectrum and time-domain accumulation spectrum,a multi-rotor UAV rotor parameter estimation algorithm is proposed.The algorithm uses the cadence frequency spectrum to estimate the periodicity of the time-frequency distribution to obtain rotor speed and uses the time-domain cumulative spectrum to estimate the frequency modulation bandwidth of the time-frequency distribution to obtain the maximum micro-doppler frequency.The rotor blade length can then be obtained by the maximum micro-doppler frequency and the rotor speed.4.Based on the DSP+FPGA architecture,a signal processing system is designed for multirotor UAV’s micro-doppler signal,and the proposed algorithms are mapped on the system.The system uses XC7VX485 T as the logic control core and three TMS320C6678 as the signal processing cores,and also designs power supply,clock management,data acquisition and other modules to assist in the realization of the system function.The analysis shows that the designed system meets the algorithm implementation requirements in terms of operation efficiency and logic control.