中文摘要：

信号稀疏分解是获取非平稳信号本质属性的有力工具.原子库的紧致性和最佳原子的搜索方式是分解算法的关键问题.在给出稀疏分解算法基本模型的基础上,阐述了匹配追踪和基追踪算法的基本原理,着重分析了两类算法的融合思想和智能优化在信号稀疏分解中的应用.最后指出了不同稀疏分解算法的异同与发展方向.

英文摘要：

Signal sparse decomposition is a powerful tool to obtain essential attribute of non-stationary signal analysis. The compactness of atomic library and search mode for best atom is the key problem of decomposition algorithm. In this paper, the basic mathematic model about sparse decomposition algorithm is shown firstly, the fundamental principles of matching pursuit and basis pursuit algorithms are elaborated. Then the fusion of these two kinds of algorithm and the application of in telligent optimization are discussed in more details. Finally the similarities and differences of sparse decomposition algorithm are reviewed and give tendency on it.