ID 原文 译文
9584 仿真结果表明,所提两种方法在有效窃密的同时具有较低的计算复杂度,另外,当窃听天线数目越多时,窃密效果越好。 ‭The simulation results show that the proposed two methods at the same time of effective espionage has lower computational complexity, in addition, when the more hacking antenna number, involving the better the results.
9585 针对合成孔径雷达(synthetic aperture radar,SAR)分布式场景回波仿真计算量巨大的问题,提出了一种结合改进的同心圆算法与图形处理器(graphics processing unit,GPU)技术的高效SAR回波仿真方法。 For synthetic aperture radar (synthetic aperture radar, SAR) problem of echo distributed scene simulation computation, an improved algorithm of concentric circles with GPU (graphics processing unit, the GPU) technology efficient SAR echo simulation method.
9586 首先,针对常规同心圆算法精度较低造成的图像信噪比低的问题,提出了一种改进的同心圆算法。 First, in view of the low precision conventional concentric algorithm caused the problem of low signal-to-noise ratio of images, an improved algorithm of concentric circles.
9587 其次,为了充分发挥GPU处理核之间的并行优势,对该算法的GPU并行处理进行了深度优化,进一步提升了仿真速度。 Secondly, in order to give full play to the GPU handle parallel between nuclear advantage, on the depth of the GPU parallel processing of the algorithm optimization, further improve the simulation speed.
9588 具体方法是,根据并行度的高低设计核函数,确定了先采用"线程外推"实现部分目标回波的同心圆累加,再用"归约相加"实现所有目标回波的累加。 Specific methods, according to the degree of parallel design of nuclear function, confirmed the first part of the target echo is realized by using "thread extrapolation" concentric rings of accumulation, reoccupy reduction "addition" to realize all the target echo of accumulation.
9589 最后,与常规GPU方法进行了实验对比,验证了所提方法的精确性和高效性。 Finally, compared with conventional GPU method has carried on the experiment, the accuracy and effectiveness of the proposed method was verified.
9590 针对不可分解函数求解问题,基于合作式协同进化(cooperative co-evolutionary,CC)框架,发展一种双系统协同进化算法。 For non biodegradable function to solve the problem, based on cooperative co-evolutionary (cooperative co - evolutionary, CC) framework, development of a dual system coevolutionary algorithm.
9591 该算法给出一种双系统A,B的CC框架新结构形式及其相应的协调机制,以增加算法的多样性和收敛性; ‭The algorithm presents A dual system A, B CC new frame structure form and its corresponding coordination mechanism, to increase the diversity and convergence of the algorithm;
9592 ‭给出双系统A,B各自求解的两种算法,例如差异进化、改进粒子群算法选择原则和匹配方式,使该两种算法具互补性,并且与双系统A,B各自角色相匹配,目的是提高基于CC框架双系统算法的计算性能。 ‭Dual system is given A, B respectively to solve the two kinds of algorithms, such as differences in evolution, improved particle swarm algorithm selection principle and the matching way, so that the two algorithms with complementary, and with A double system, B match their respective roles, the purpose is to improve the framework based on CC dual system computing performance of the algorithm.
9593 经不可分解函数集(维数D=1 000)测试表明,本文算法计算性能(计算精度和标准差)与其他3种典型算法相比,对于其中某些函数求解占优,总体上4种算法对函数集的求解各有所长,具有互补性。 The indecomposable function sets (dimension D = 1, 000) tests show that this algorithm in computing performance (accuracy and standard deviation), compared with other three kinds of typical algorithms for some of the function to solve the dominant, the four kinds of algorithm of function sets overall strengths and complementary.