Research On The Localization Method Of Multiple Radiation Source Based On Sparse UAV Array

Unmanned aerial vehicles(UAVs),due to their high concealment,strong maneuverability,and low susceptibility to ground obstacles,have high application value in both military and civilian fields.The use of UAVs as airborne localization platforms provides novel solutions for high-precision radiation source sensing and localization.Among them,UAV array treat a single UAV as an array element,with the characteristics of agile deployment and flexible localization,gradually attracting the attention of researchers.However,existing methods for multiple radiation source localization based on UAV array mostly adopt uniform linear array(ULA),where the distances between adjacent array elements are fixed and equal,which leads to following drawbacks: 1.The cost of expanding the array aperture and degree of freedom(DOF)is high.2.Dense array configurations will increase the mutual coupling effect between array elements and are detrimental to the security of UAV array systems.Therefore,this dissertation starts from the perspective of the factors that affect the performance of multi-radiation source localization,and uses sparse array technology to compensate for the shortcomings of uniform UAV arrays.This research covers the design of sparse array,estimation of UAV position errors,and final reach localization of off-grid multiple radiation sources.The main works of the paper are as follows:(1)In the existing research on sparse array design for improving direction-finding and localization performance based on swarm intelligence algorithms,weight assignment method is commonly adopted to transform the multi-objective optimization problem into a single-objective optimization problem.However,it is difficult to select appropriate weights in practice,and it is uncertain whether the optimization objectives conflict when designing more complex optimization objectives.To address the issue,a method for designing of sparse array based on adaptive NSGA-II is proposed using the concept of multi-objective optimization.Meanwhile,this dissertation takes into account the security of the UAV array system.Based on the two traditional optimization objectives of DOF and the Cramer-Rao bound(CRB)of parameter estimation,an optimization function is designed to further reduce the number of adjacent UAVs(i.e.,array elements)in consecutive positions by minimizing the number of continuous array elements.Simulation results demonstrate that the proposed sparse array effectively increases the DOF of array,reduces the number of continuous array elements in the physical array,and improves the direction-finding accuracy of multiple radiation sources while ensuring the security of the UAV array.(2)Due to perturbations such as airflow disturbances or errors in the self-positioning system,the actual position of a UAV may deviate from its ideal position,resulting in decreased localization accuracy for multiple radiation sources.To mitigate this issue,a two-dimensional position error estimation method of array elements based on gradient descent method is proposed,which is suitable for sparse arrays.Firstly,a direction of arrival(DOA)estimation model with two-dimensional position error is established for sparse arrays using the atomic norm(AN)framework.Then,the gradient descent method is employed to estimate the two-dimensional position errors of the elements in difference array.And a transformation matrix is defined to obtain the true errors of physical array elements from the estimation above,leveraging the relationship between the difference array and the physical array.Subsequently,the errors estimation of physical array are utilized to rectify the actual position of the UAVs,thereby attaining more precise estimations of DOAs.Simulation results indicate that the proposed approach can efficiently estimate the two-dimensional position error of elements in a sparse array,which helps enhance the precision of localization for multiple radiation sources.(3)In this dissertation,we proposed a multiple radiation sources localization system based on sparse UAV arrays.However,for the same radiation source in space,the division between the far-field source and near-field sources may differ for UAV arrays at different positions,making a single model of near-field source or far-field source is no longer applicable.To address this issue,we discretized the localization space and constructed a universal model for far-field and near-field sources using distance and angle parameters.Meanwhile,on the basis of the above model,to mitigate grid mismatch problem in off-grid compressive sensing,an off-grid model based on a joint distance-angle dictionary is designed to further improve localization accuracy.Finally,the off-grid model is solved under the framework of sparse Bayesian learning(SBL).The shared sparsity of observation data from multiple UAV arrays at different spatial locations can be effectively utilized to jointly adjust the spatial grids,thereby achieving the localization of multiple radiation sources.Simulation results demonstrate the effectiveness of the proposed approach,indicating that it can attain high-precision localization of off-grid multiple radiation sources.