1、内容简介
matlab136-基于梯度下降和软阈值化的去噪算法
可以交流、咨询、答疑
2、内容说明
略
Diagnosing faults in large mechanical equipment poses challenges due to strong background
noise interference, wherein extracting weak fault features with periodic group-sparse property is
the most critical step for machinery intelligent maintenance. To address this problem, a periodic
group-sparse method based on a generalized minimax-concave penalty function is proposed in
this paper. This method uses periodic group sparse techniques to capture the periodic clustering
trends of fault impact signals. To further enhance the sparsity of the results and preserve the
high amplitude of the impact signals, non-convex optimization techniques are integrated. The
overall convexity of the optimization problem is maintained through the introduction of a
non-convex controllable parameter, and an appropriate optimization algorithm is derived. The
effectiveness of this method has been demonstrated through experiments with simulated signals
and mechanical fault signals.
Keywords: non-convex penal
3、仿真分析
略
4、参考论文
略