Research On DOA Estimation Technology Of Vehicle-Mounted Millimeter Wave Radar Based On MIMO Array

Millimeter-wave radar has become a key component of automotive advanced driver assistance systems due to its ability to work all day and all weather,as well as its advantages of low cost and small size.For vehicle-mounted millimeter-wave radar,its core function is the precise identification and positioning of targets,which relies on the ability to estimate angles with high precision and high resolution.Array optimization design and super-resolution DOA estimation algorithm are important contents for improving the angle estimation performance of vehicle-mounted millimeter-wave radar.This thesis has carried out the corresponding array optimization algorithm research,super-resolution DOA estimation algorithm research and other work.The details are as follows:First,the sparse array optimization algorithm of vehicle-mounted millimeter wave MIMO radar based on fireworks algorithm is studied.Aiming at the problem that the original fireworks algorithm cannot be directly applied to discrete array optimization,the method of explosion and mutation to generate sparks and the distance representation of individual fireworks are improved,and a discrete fireworks algorithm suitable for solving discrete array optimization problems is proposed.The simulation results show that,compared with the classical genetic algorithm and particle swarm optimization algorithm,this algorithm can search for a better array arrangement,the corresponding pattern can achieve lower side lobe level,and the convergence speed is faster.Discrete fireworks algorithm is a superior optimization algorithm for vehicle-mounted millimeter wave MIMO radar array.Second,the DOA estimation algorithm based on iterative adaptive method is studied.Vehicle-mounted millimeter wave radar has a high requirement for real-time algorithm.Aiming at the problem that the iterative adaptive method needs to scan the whole space in each iteration,which results in too much calculation,an adaptive selection strategy is introduced.In each iteration process,only the signal estimation is performed on the angle region of interest where the target may exist,in order to improve the computational efficiency of the algorithm.The simulation results show that the improved algorithm is basically the same as the original algorithm in terms of super-resolution performance,but the computational efficiency is greatly improved.The improved algorithm is more suitable for vehicle-mounted scenes which require high real-time performance.Third,a DOA estimation algorithm based on compressive sensing theory is studied.A high precision DOA estimation algorithm is needed for vehicle-mounted millimeter wave radar.Aiming at the low precision of DOA estimation of OMP algorithm,RELAX algorithm is introduced to further correct its DOA estimation result.Simulation results show that the improved algorithm can effectively improve the DOA estimation accuracy of the original OMP algorithm.Aiming at the problem of excessive calculation of the SBL algorithm,consider reducing the number of grid divisions,which increases the inherent angle estimation error for off-grid targets,so the RELAX algorithm is introduced to further correct its DOA estimation results.Simulation results show that the improved algorithm is slightly lower than the original algorithm in super-resolution performance,but the computational efficiency has been significantly improved.The improved algorithm is more suitable for vehicle-mounted scenes which require high real-time performance.Fourth,the AWR2243 cascade radar developed by Texas Instruments was used to collect the measured data in the field.The experimental results demonstrate the super resolution performance and computational efficiency of the proposed algorithm.