Abstract: The Orthogonal Matching Pursuit（OMP） algorithm which is based on the idea of greedy iteration is one of the compressed sensing signal reconstruction algorithm. To reduce the OMP computational complexity, quantum particle swarm（QPSO） algorithm which is more powerful in searching for global optimal solution is applied to optimize the matching course of orthogonal matching pursuit algorithm （QPSO-OMP）. According to OMP algorithm features the atomic weight secondary matching is introduced to improve the reconstruction accuracy of QPSO-OMP. Simulation results show that the orthogonal matching pursuit algorithm with secondary matching based on the quantum particle swarm algorithm performs better than that based on the particle swarm algorithm in accurate reconstruction probability while with low complexity.

Key words: compressed sensing, over-complete dictionary, orthogonal matching pursuit algorithm, quantum particle swarm algorithm

摘要： 正交匹配追踪算法（OMP）是一种基于贪婪迭代思想的算法，是压缩感知中信号重构方法之一。为了降低OMP算法的计算复杂度，采用一种全局寻优能力较强的量子粒子群算法（QPSO）优化OMP算法中的匹配过程（QPSO-OMP）；针对OMP算法特点，引入原子分量二次匹配，进一步提高QPSO-OMP算法重构精度。仿真结果表明，所提出的基于QPSO算法的二次匹配OMP算法复杂度低，精确重构概率高于基于粒子群算法的正交匹配追踪算法。

关键词: 压缩感知, 冗余字典, 正交匹配追踪算法, 量子粒子群算法