| ID | 原文 | 译文 |
| 55047 | 在源信号个数未知条件下,提出一种基于改进K-均值聚类的欠定混合矩阵盲估计方法。 | Under the condition of unknown number of source signals, proposes a k-means clustering are based on the improved underdetermined mixing matrix blind estimation method. |
| 55048 | 该方法首先计算观测信号在单位半超球面上投影点的密度参数,然后去掉低密度投影点,并从高密度投影点中选取初始聚类中心,最后对剩余投影点进行聚类,根据Davies-Bouldin指标估计源信号个数,并估计出混合矩阵。 | This method first calculates the observation signal in the projection point on the unit of half a hypersphere density parameters, and then remove the low density projection point, and from the high density projection point selecting initial cluster centers, the last remaining projection points for clustering, according to the number of Davies - Bouldin index to estimate the source signals, and estimate the mixing matrix. |
| 55049 | 仿真结果表明,该方法的复杂度低,其运行时间仅为拉普拉斯势函数法的1% | The simulation results show that the method of low complexity, its running time is only 1% of the Laplace's potential function method |
| 55050 | 针对机群编队分组问题,提出了一种加权双质心支持向量聚类算法。 | In view of the fleet formation grouping problem, puts forward a double mass center weighted support vector clustering algorithm. |
| 55051 | 所提算法在支持向量训练时引入最大熵原理,快速求解Lagrange乘子;针对样本特征对聚类结果的贡献不同,在聚类标识过程中,引入加权密度质心,提出了加权双质心聚类标识,并在典型数据集上验证了所提算法的有效性。 | The proposed algorithm is the maximum entropy principle, was developed for the training of support vector fast solving though laser multiplier;Contribution to the clustering results based on the sample is different, in the process of clustering identification, introducing weighted centroid density, puts forward the weighted centroid clustering identification, and in a typical data sets verify the effectiveness of the proposed algorithm. |
| 55052 | 通过对机群编队分组模型的描述,建立了机群聚类时一个目标点需要的特征集,完成了编队分组的仿真实验。 | Through the description of fleet formation group model, set up a fleet of clustering feature set, a target needs when completed the formation of grouping simulation experiment. |
| 55053 | 仿真结果表明了所提算法能够针对应用的具体样本集实行快速聚类分析,并保证聚类结果的有效性。 | The simulation results show that the proposed algorithm can according to the specific sample set of application for fast clustering analysis, and ensure the effectiveness of the clustering results. |
| 55054 | 针对装备活动中危险耦合传导关系描述不清楚、分析不具体的问题,构建考虑任意分布的危险耦合传导图形评审技术(graph evaluation and review technique,GERT)改进模型。 | For equipment in dangerous coupled conduction relationship description is not clear, concrete problems, don't build considering the risk of any distribution coupled conduction graphics review technology (graph evaluation and review technique, GERT) to improve the model. |
| 55055 | 提出了系统危险度、危险耦元重要度、危险路径隶属度等扩展分析参数,进一步刻画装备活动中的微观危险信息;采用系统耦合理论和极大熵方法(maximum entropy,ME)配置GERT模型变量,并设计量子和声算法(quantum harmony search algorithm,QHS)求解危险度概率密度极大熵模型。 | Proposed system risk, the dangerous coupling yuan importance and dangerous path to membership extension analysis parameters, further depict the micro dangerous equipment activity information;Using the system coupling theory and maximum entropy method (maximum entropy, ME) configuration GERT model variables, and design a quantum harmonic algorithm (quantum harmony search algorithm, QHS) to solve the risk probability density maximum entropy model. |
| 55056 | 最后,以飞机大表速低空俯冲科目为例进行分析。 | Finally, in the plane dived large table speed low, subjects, for example for analysis. |