快速移不变稀疏分类算法在线识别汽油机故障-计算机工程与应用.PDFVIP

230 2018 ，54（11） Computer Engineering and Applications 计算机工程与应用 快速移不变稀疏分类算法在线识别汽油机故障 用 1，2 1 张晓焱 ，刘 永 应 ZHANG Xiaoyan1，2 , LIU Yong1 与 1. 南京理工大学 计算机科学与工程学院，南京 210094 程 2. 南京交通职业技术学院，南京 211188 g r 1.School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094, China 工 o 2.Nanjing Vocational Institute of Transport Technology, Nanjing 211188, China . 机 j a ZHANG Xiaoyan, LIU Yong. On-line faults recognition algorithm for gasoline engine using sparse representation 算 e classification based on fast shift invariant sparse coding. Computer Engineering and Applications, 2018, 54（11）：230-235. c 计 . Abstract ：According to the low on-line efficiency of the shift invariant sparse coding, a fast shift invariant sparse coding w algorithm which has high on-line efficiency is proposed in this paper. Then the on-line faults recognition can be realized w combining the sparse representation classification. The algorithm transforms the shift invariant optimization problem from w the time domain to the frequency domain firstly. The sparse coefficient and classification dictionary are solved quickly using the feature sign and Lagrange dual method, which can reduce the computational time complexity without the loss of classification accuracy. And the real-time performance for fault recognition will be promoted. The experimental results of engine show that, while the engine works in