LMS

  在一种变步长LMS 算法的基础上，引进动量因式，提出了一种新的改进LMS 的算法。新算法整体性能优 于变步长LMS 算法以及LMS 算法。通过理论分析，比较了新的算法和变步长LMS 算法以及LMS 算法的收敛性 和稳态性，提出了一种设想以提高新算法的稳态性。仿真试验证明了新算法的优越性以及设想的在仿真条件下的正 确性。(In a variable step size LMS algorithm, based on the introduction of momentum factor, a new improvement of LMS algorithms. The overall performance of the new algorithm is superior to variable step size LMS algorithm and LMS algorithm. Through theoretical analysis, comparing the new algorithm and variable step size LMS algorithm and LMS algorithm convergence and steady-state, and proposes a new algorithm is envisaged to enhance the steady-state nature. Simulation tests proved the superiority of the new algorithm, as well as envisaged under the conditions in the simulation is correct.)

一种改进的LMS算法.pdf