ID 原文 译文
20005 针对阵元间距大于信号波长引起阵列方向图出现栅瓣的问题，该文提出一种基于粒子群优化(PSO)算法的宽带真延时方向图栅瓣抑制方法。 A grating lobe suppression method of wideband real time delay pattern based on Particle Swarm Optimization(PSO) algorithm is proposed to solve the problem of grating lobe arise from inter-element is larger than wavelength.
20006 该方法首先定义了基于宽带真延时的阵列能量方向图，其次构造了以阵列能量方向图的最高副瓣电平作为适应度函数，最后利用粒子群优化算法优化阵元分布来实现对阵列方向图栅瓣的进一步抑制。 Firstly, the array energy pattern based on wideband real time delay is defined. Then, a fitnessfunction is constructed with maximum sidelobe level of the array energy pattern. Finally, the grating lobe isfurther suppressed by optimizing the elements position distribution using Particle Swarm Optimization (PSO)algorithm.
20007 仿真结果表明：相比于单独使用粒子群算法和单独使用宽带真延时方法，该方法对方向图栅瓣的抑制性能更加有效， The simulation results show that the proposed grating lobes suppression method is more effective than individually using the particle swarm optimization method or the wideband real time delay method.
20008 在此基础上，该文还研究了阵元个数、平均阵元间距、信号时宽和信号中心频率等因素对方法抑制栅瓣性能的影响。 Furthermore, the influence of the element space, the element number, the time width and the center frequency of signal on the performance of grating lobe suppression are studied.
20009 随着机器学习被广泛的应用，其安全脆弱性问题也突显出来。 As machine learning is widely applied to various domains, its security vulnerability is also highlighted.
20010 该文提出一种基于粒子群优化(PSO)的对抗样本生成算法，揭示支持向量机(SVM)可能存在的安全隐患。 A PSO (Particle Swarm Optimization) based adversarial example generation algorithm is proposed to reveal the potential security risks of Support Vector Machine (SVM).
20011 主要采用的攻击策略是篡改测试样本，生成对抗样本，达到欺骗SVM分类器，使其性能失效的目的。 The adversarial examples, generated by slightly crafting the legitimate samples, can mislead SVM classifier to give wrong classification results.
20012 为此，结合SVM在高维特征空间的线性可分的特点，采用PSO方法寻找攻击显著性特征，再利用均分方法逆映射回原始输入空间，构建对抗样本。 Using the linear separable property of SVM in high-dimensional feature space, PSO is used to find the salient features, and then the average method is used to map back to the original input space to construct the adversarial example.
20013 该方法充分利用了特征空间上线性模型上易寻优的特点，同时又利用了原始输入空间篡改数据的可解释性优点，使原本难解的优化问题得到实现。 This method makes full use of the easily finding salient features of linear models in the feature space, and the interpretable advantages of the original input space.
20014 该文对2个公开数据集进行实验，实验结果表明，该方法通过不超过7%的小扰动量生成的对抗样本均能使SVM分类器失效，由此证明了SVM存在明显的安全脆弱性。 Experimental results show that the proposed method can fool SVM classifier by using the adversarial example generated by less than 7 % small perturbation, thus proving that SVM has obvious security vulnerability.