ID 原文 译文
51577 针对多波段激光干涉图像中的频谱数据混叠问题，文章提出了一种基于数据挖掘技术的自相关提取分类算法。 Aiming at the problem of spectral data aliasing in multi-band laser interference images, this paper proposes an autocorrelation extraction classification algorithm based on associated feature information.
51578 首先，采用自相关计算模式完成数据属性分类; First, the auto-correlation calculation mode is used to complete the data attribute classification.
51579 然后，通过属性隶属度函数作为判据设计了基于特征数据挖掘的自回归平均模型; Then, an autoregressive average model based on feature data mining is designed by using attribute membership function as a criterion.
51580 最后，采用时序衰减权值系数循环迭代的方式完成数据挖掘。 Finally, the data mining is completed by cyclic iteration of time-decay weight coefficients.
51581 实验对混合激光光源进行测试，并与传统均值分类算法和最小二乘法分类算法作了对比。 In the experiment, a mixed laser light source is tested. The algorithm is compared with traditional mean classification algorithm and least square method classification algorithm.
51582 仿真结果表明，该算法的分类平均误差为0.110，优于另外两种算法(0. 191和0.167)， Experimental results show that the classification error of the algorithm is 0. 110, which is better than the other two algorithms(0. 191 and 0. 167).
51583 并且在相同数据总量的条件下该算法的耗时最少。 And under the condition of the same total amount of data, the proposed algorithm takes the least time.
51584 由此可见，该算法对多波段激光干涉图像中混叠数据的有效分离具有很好的改善作用，在混叠数据分类应用中具有一定的实用价值。 It can be seen that this algorithm has a very good effect on the effective separation of aliasing data in multi-band laser interference images. It has certain practical value in the application of aliasing data classification.
51585 针对电数据中心网络(DCN)延迟高和吞吐量低等问题，文章提出了两种无源光DCN架构，一种针对小型DCN，一种针对大型DCN。 Aiming at the problem of high latency and low throughput in electric Data Center Network(DCN), two kinds of passive optical DCN architecture are proposed, one for small DCN and the other for large DCN.
51586 两种架构均采用了阵列波导光栅路由器(AWGR)、耦合器和解复用器等无源光器件，并针对每种规模的架构给出了波长分配和分组传输方法。 Using passive optical devices such as Arrayed Waveguide Grating Router(AWGR), coupler and demultiplexer, wavelength assignment and packet transmission algorithms are proposed for each scale architecture.