论文部分内容阅读
为了克服经典正交匹配算法获取原子集时遍历冗余字典具有较大时间开销的缺点,提出了一种基于压缩感知理论和禁忌优化算法的的稀疏故障信号特征提取方法;首先引入了压缩感知模型并描述了基于信号稀疏表示的故障诊断原理,设计了满足RIP准则以最小化l1范数为目标的稀疏信号解的求解方法,然后定义了一种基于正交匹配算法的稀疏信号重构算法,并以最小化余量为目标函数,采用改进的禁忌搜索算法在原子空间中搜索满足目标函数的最优原子集,最后,给出了基于稀疏编码和禁忌优化混合模型的故障信号提取算法;在Matlab仿真环境下对滚动轴承故障信号进行试验,仿真结果表明:文章方法能有效地对具有强噪声的故障信号进行稀疏重构,不仅具有较高的信噪比,而且具有较小的余量误差和仿真时间,与其它方法相比,具有较大的优越性。
In order to overcome the shortcomings of classical orthogonal matching algorithm traversing redundant dictionaries when acquiring atomic sets, this paper proposes a method of feature extraction of sparse fault signals based on compressed sensing theory and tabu search optimization algorithm. Firstly, compressed sensing model And describes the principle of fault diagnosis based on signal sparse representation. A solution to the sparse signal solution that satisfies the RIP criterion to minimize the l1 norm is designed. Then, a sparse signal reconstruction algorithm based on the orthogonal matching algorithm is defined. And the minimization of the margin as the objective function, the improved tabu search algorithm is used to search the atomic space for the optimal atomic set satisfying the objective function. Finally, the fault signal extraction algorithm based on sparse coding and taboo optimization hybrid model is given. Matlab simulation of rolling bearing fault signal test results show that: the article method can effectively carry out the sparse reconstruction of the fault signal with strong noise, not only has a high signal to noise ratio, but also has a smaller margin error and Simulation time, compared with other methods, has greater advantages.