Due to the poor robustness of classification accuracy for traditional sparse representation classification algorithms in a low-dimensional space,a novel kernel-mapping sparse representation classification algorithm was proposed.Kernel-mapping was used to project samples in a low-dimensional space to a high dimensional one,thus the linear separability among samples was improved.On this basis,the sparse solutions of the samples in the high-dimensional space were obtained using a sparse representation classification algorithm.The simulation results of bearing failure data showed that the proposed algorithm has a better robustness of kernel parameters and improves the classification accuracy significantly.%针对传统稀疏表示分类算法在低维空间分类精度难以保证问题,提出基于核映射的稀疏表示分类算法,并获得低维样本在高维空间坐标,样本间线性可分度得以改善;在此基础上,利用稀疏表示分类算法获得样本在高维空间的稀疏解.经滚动轴承故障分类实验验证,新算法对核参数具有较高的鲁棒性;可明显提高分类精度.
展开▼