基于遗传-启发算法的微动目标认知ISAR成像资源调度

ISAR Resource Scheduling of Multiple Micro-Motion Targets Based on Genetic-Heuristic Algorithm

认知逆合成孔径雷达(Inverse Synthetic Aperture Radar,ISAR)将认知雷达理论与ISAR成像技术相结合,利用目标与环境的离线先验知识及在线感知结果,提升ISAR在复杂环境下的成像能力.与传统单脉冲雷达相比较,相控阵ISAR可实现波束的快速扫描,因此具备多目标观测能力.在实际空间微动目标成像场景中,相控阵ISAR很可能存在多波束多目标情况.为实现系统效能的充分发挥,需要在系统能量、时间资源有限条件下实现成像任务有效调度.然而,面向成像任务的资源调度方法主要针对刚体目标,其约束条件及调度模型并不适用于微动目标,并且存在求解方法稳定性差、调度成功率不高等问题.此外,微动目标运动形式复杂,回波非平稳性很强,且方位缺损时难以聚焦成像.因此,迫切需要针对微动目标的特性,研究有效的ISAR资源调度和高分辨成像方法.针对上述问题,本文提出基于遗传-启发算法的微动目标认知ISAR成像资源调度方法 .该方法首先根据微动目标认知结果计算其成像所需的雷达资源及综合优先级;接着基于脉冲交错技术,在时间、能量双重约束下建立微动多目标资源调度模型.在此基础上,本文提出了基于遗传-启发算法的最优调度求解方法 .进一步,针对资源调度导致微动目标回波缺损,难以聚焦成像的问题,本文提出了基于交替方向乘子法(Alternating Direction Method of Multipliers,ADMM)的距离-瞬时多普勒(Range-Instantaneous Doppler,RID)成像方法,从而实现成功调度任务对应微动目标回波的高分辨聚焦成像.实验结果表明,相比于现有资源调度方法,本文所提方法调度结果更为稳定,调度成功率更高,并可实现回波缺损时的微动目标聚焦成像.

By including cognitive theory,cognitive inverse synthetic aperture radar(ISAR)has improved on the im⁃aging capability of ISAR for targets in complex environments through the use of offline prior knowledge and online percep⁃tion results of the target and environment.Compared with traditional monopulse radar,phased-array ISAR achieves rapid beam scanning through the control of the phase of each antenna element in the antenna array and can therefore observe mul⁃tiple targets.Multi-beam and multi-target situations are likely to exist in practical spatial micro-motion target imaging sce⁃narios of phased-array ISAR.When there are multiple spatial micro-motion targets in multiple radar beams,the success rate of imaging task scheduling is usually low,which limits the full exploitation of the ISAR effectiveness.In fully realizing system efficiency,imaging tasks need to be effectively scheduled under limited system energy and time resources.Resource scheduling methods for imaging tasks are mainly aimed at rigid targets.The basic idea is to adopt sparse aperture ISAR imaging in constructing the optimal scheduling model under the radar resource constraints and then use a heuristic algorithm to solve the model.However,the constraint conditions and scheduling models of exiting methods are not suitable for micro-motion targets,and the adopted heuristic algorithm usually has poor stability and a low scheduling success rate.In addition,the motions of micro-motion targets are usually complex and their echoes are strongly non-stationary,making it difficult to achieve focused imaging in scenarios of data corruption.There is thus an urgent need to study effective resource scheduling and high-resolution imaging methods for ISAR based on the characteristics of micro-motion targets.To tackle the above issues,this paper proposes a cognitive ISAR resource scheduling method based on a genetic heuristic algorithm for the imaging of multiple micro-motion targets.This method first calculates the radar resources and comprehensive priorities required for imaging according to the cognized characteristics of micro-motion targets.On this basis,the normalized available energy is proposed as a metric of the resource scheduling performance.Then,adopting pulse interleaving,a resource scheduling model for multiple micro-motion targets is established under the dual constraints of time and energy.Furthermore,to address the issues that the modeling struggles to provide analytical solutions and the low stability of the existing solution methods,an optimal scheduling solution method based on a genetic heuristic algorithm is proposed.The proposed method improves the scheduling success rate and the stability of the scheduling results.Furthermore,to tackle the issue that it is difficult to achieve focused imaging when the echoes of micro-motion targets are missing because of resource scheduling,a method based on the alternating direction method of multipliers is adopted to achieve range-instantaneous Doppler imaging.The high-resolution focused imaging of the micro-motion targets corresponding to the successfully scheduled imaging tasks is thus achieved.Experimental results show that compared with the existing resource scheduling method based on a heuristic algorithm,the proposed method has greater stability and a higher success rate in resource scheduling and can achieve focused imaging of the micro-motion targets in a pulse interleaving situation.