ID 原文 译文
49096 末段反导作战火力任务分配建模是一个复杂的不确定多约束问题建模,首先建立了末段双层反战术弹道导弹火力-目标匹配模型,其次对传统粒子群优化算法(particle swarm optimization,PSO)进行改进给出了一种吸引子PSO(attractor PSO,APSO),APSO引入吸引子,在保持群体多样性的基础上,将粒子聚集在最优值附近,增加相应区域的粒子密度。 Terminal antimissile fighting fire modeling is a complex task assignment constraint problem modeling uncertainty, end of the first established the double anti tactical ballistic missile firepower - target matching model, next to the traditional particle swarm optimization algorithm, particle swarm optimization, PSO) to improve a is provided by PSO (attractor PSO and APSO), attractor, introduced APSO, on the basis to maintain the diversity of the population, the particles gathered near the optimal value, increase the corresponding area of the particle density.
49097 其中,为了方便问题求解,将火力-目标匹配优化任务进行分解,转化成多个子时间段,再用APSO对多个子时间段进行求解。 Among them, in order to facilitate problem solving, match the fire - target optimization task decomposition, into a more child time, reoccupy APSO for solving more child time.
49098 仿真实例表明,APSO有更加优良的收敛精度尤其是收敛速度,满足了反TBM作战火力任务分配的高时效性要求。 Simulation examples show that APSO has more excellent especially the convergence speed, convergence precision can meet the capabilitie s fighting fire high timeliness requirements of task assignment.
49099 研究了以景象匹配制导为目的的曲线弹道合成孔径雷达(synthetic aperture radar,SAR)快视成像问题。 Studied for the purpose of scene matching guidance curve ballistic synthetic aperture radar (synthetic aperture radar, SAR) quick view imaging problems.
49100 首先建立了曲线弹道SAR回波信号模型,分析了多普勒历程和距离徙动特点,提出了等效正侧视成像的概念,并分析了等效条件,大大减小了距离徙动校正难度和回波信号距离向与方位向的耦合,然后推导了曲线弹道SAR回波信号的二维频谱表达式,在此基础上提出了结合距离-多普勒算法和频谱分析(spectral analysis,SPECAN)的快视成像方法。 Ballistic model of SAR echo signal curve was established by the analysis of the doppler course and distance migration characteristics, puts forward the concept of equivalent is side-view imaging, and analyzed the equivalent conditions, greatly reduced the difficulty of range migration correction and echo signal to the distance and azimuth to coupling, and then the curve is deduced two-dimensional spectrum expressions of ballistic SAR echo signal, on the basis of this puts forward the range - doppler algorithm and frequency spectrum analysis, spectral analysis, SPECAN) fast imaging method.
49101 所提算法使用高效的SPECAN方法进行方位压缩,能够完成曲线弹道SAR部分孔径数据的精确相干处理,没有因为孔径的非直线而增加成像算法的复杂性。 Proposed algorithm using efficient methods of SPECAN azimuth compression, able to complete ballistic curve precise coherent processing of SAR part of aperture, not because of the aperture of the straight line and increase the complexity of imaging algorithm.
49102 已有演化元胞遗传算法中的演化规则多从元胞自动机中直接引入,未在状态演化中考虑个体间适应值的差异。 Existing evolutionary cellular genetic algorithm of evolution rules directly from the cellular automata is introduced, did not consider between individuals in the state evolution fitness differences.
49103 根据密度制约关系提出一种新的演化元胞遗传算法来处理动态优化问题,在考虑个体适应值优劣与局部种群密度的前提下,通过密度制约与种内竞争实现个体在元胞空间内的生死演化,并建立种群规模增长模型控制元胞空间内存活个体规模。 According to the density restriction relationship presents a new evolutionary cellular genetic algorithm to deal with dynamic optimization problem, considering the individual fitness and the local population density under the premise of individual by density and intraspecific competition in cellular space within the evolution of life and death, and establish the population size scale control cellular space for individual growth model.
49104 选取不同强度、复杂度的动态优化问题对算法性能进行验证,结果表明新算法具有良好的处理动态优化问题的能力。 Selection of different intensity, complexity and dynamic optimization problem validate the performance of algorithm, the results show that the new algorithm has good ability to deal with dynamic optimization problems.
49105 针对K-奇异值分解(sigular value decomposition,SVD)算法存在的问题,结合结构聚类和字典学习,提出了一种基于非局部正则化稀疏表示的图像去噪算法。 For K - singular value decomposition (sigular value decomposition, SVD) algorithm existing problems, combining with clustering structure and dictionary learning, is proposed based on a nonlocal regularization sparse representation of image denoising algorithms.