ID	原文	译文
53727	1) 根据聚类有效性指标值的收敛点确定信源数目;	1) determine the num ber of sources according to the convergence point of the clustering validity index;
53728	2) 根据隶属度矩阵排除离群点;	2) eliminate outliers according to the membership matrix;
53729	3) 确定K-means 的初始聚类点。	3) determine the initial clustering points of K-means.
53730	最后使用 K-means 并利用预聚类确定的信源数目及初始聚类点实现混合矩阵估计。	Finally, K-means is used to achieve mixing ma trix estimation based on the pre-determined number of sources and the initial clustering points.
53731	仿真结果表明提出的方法具有更优的混合矩阵估计性能。	The simulation results sug gest that the proposed method has superior performance of mixing matrix estimation.
53732	针对雷暴天气下伴随着多种危险气象同时存在的动态改航规划需求，提出了一种考虑多种危险气象的分层航迹动态规划策略。	A stratified path dynamic planning strategy in view of multiple weather hazards is proposed to solve the demand of dynamic rerouting planning in thunderstorm weather with a variety of Hazardous weather at the same time.
53733	首先，根据雷达回波图对改航空域进行分层规划;	Firstly, according to the radar echo map, the rerouting airspace is planned hierarchically.
53734	然后，基于改航起点和改航终点利用稀疏A* 算法结合改航约束条件对分层空域进行全局航迹规划，并利用关键点选取策略确定关键航迹点;	Secondly, based on the rerouting starting point and rerouting ending point, the sparse A* algorithm combined with the rerouting constraints is used to plan the global path in the layered airspace, and the key points are determined by using the key point selection strategy.
53735	最后，根据关键航迹点对全局航迹进行分段，并利用动态窗口法逐段进行实时航迹动态规划。	Finally, the global path is segmented by the key point, and the real-time path dynamic planning is carried out by using the dynamic win- dow method.
53736	实验结果表明所提方法具备良好的实时航迹动态规划性能，并且优于动态稀疏 A* 算法、动态窗口法和现行策略，	The experimental results show that the proposed method had good real-time dynamic path planning perform- ance, and was superior to dynamic sparse A* algorithm, dynamic window method and current strategy.