快速约束极值子空间跟踪算法

Fast Constrained Extremum Subspace Tracking Algorithm

子空间跟踪算法是许多工程计算问题的核心.Valizadeh等人基于约束优化问题提出了一种快速投影逼近子空间跟踪算法.在指出快速投影逼近子空间跟踪算法不足的同时,应用矩阵求逆引理给出了一种新的快速子空间跟踪算法——快速约束极值子空间跟踪算法.仿真实验表明,所提算法是收敛且稳定的,其性能优于或相当于几种常见的快速子空间跟踪算法.

Subspace tracking algorithm is the core of the problem in many engineering calculation.Valizadeh et al.have proposed a fast constrained projection approximation subspace tracking（CPAST） algorithm based on the constrained optimization problem.Based on noting the disadvantages of the CPAST algorithm,a new fast subspace tracking algorithm,namely fast constrained extremum subspace tracking algorithm is proposed by applying the matrix inversion lemma.Simulation results show that the proposed algorithm is convergence and stableness,and its performance is better than or equal to some well-know fast subspace tracking algorithms.