摘要： 针对标准匹配追踪(MP)算法在寻找最佳原子时计算量大的问题，提出一种基于免疫匹配追踪(IA-MP)的语音稀疏分解算法。该算法采用免疫克隆优化机制搜索最佳原子，利用抗体的种群规模控制冗余字典的大小，选择实数交叉与非均匀变异方法保证字典的完备性。仿真实验结果表明，与标准MP算法和遗传匹配算法相比，IA-MP算法可明显降低匹配追踪的计算量，算法性能较稳定，利用该算法分解后的稀疏信号具有较高的重构精度。
关键词:
匹配追踪,
免疫算法,
稀疏分解,
冗余字典,
重构信号,
语音信号
Abstract: Aiming at the problem that since the standard Matching Pursuit(MP) algorithm tends to cause enormous calculation quantity when searching for the optimal atoms, this paper proposes a scheme of Matching Pursuit algorithm based on Immune Algorithm(IA-MP). This algorithm adopts immune cloning optimization mechanism to search for the optimal atoms, using the population scale of antibodies to control the size of redundant dictionary, and choosing real cross and nonsymmetrical mutation to ensure the completeness of dictionary. Simulation results show that, compared with the standard MP and Genetic Algorithm-Matching Pursuit(GA-MP) algorithms, IA-MP algorithms can obviously reduce the matching pursuit calculation quantity, and the performance of this algorithm is more stable, in addition, this algorithm has high signal reconstruction accuracy.
Key words:
Matching Pursuit(MP),
Immune Algorithm(IA),
sparse decomposition,
redundant dictionary,
reconstructing signal,
speech signal
中图分类号:
周燕, 刘韬, 尚丽. 基于免疫匹配追踪的语音稀疏分解算法[J]. 计算机工程, 2012, 38(21): 161-163,167.
ZHOU Yan, LIU Tao, CHANG Li. Speech Sparse Decomposition Algorithm Based on Immune Matching Pursuit[J]. Computer Engineering, 2012, 38(21): 161-163,167.