论文部分内容阅读
针对高速列车同时发生两种故障的问题,提出了一种基于短时傅里叶变换(Short-Time Fourier Transform,STFT)和时频比的高速列车转向架混合故障盲分离算法。首先对观测数据进行STFT线性时频变换,再运用混合信号的时频比(Time-Frequency Ratio of Mixtures,TIFROM)估计单源域,得到分离矩阵;然后对源信号和分离后的估计信号分别提取矢量特征,用于盲分离准确度评价;最后用支持向量机(Support Vector Machine,SVM)识别盲分离的正确率。
In order to solve the problem that two kinds of faults occur simultaneously in a high-speed train, a hybrid high-speed train bogie blind separation algorithm based on Short-Time Fourier Transform (STFT) and time-frequency ratio is proposed. Firstly, the STFT linear time-frequency transform is applied to the observed data, and then the single-source domain is estimated by using the time-frequency ratio of mixtures (TIFROM) of the mixed signal to obtain a separation matrix. Then, the source signal and the separated estimated signal are respectively extracted Vector feature for the accuracy of blind separation evaluation; Finally, support vector machine (Support Vector Machine, SVM) to identify the accuracy of blind separation.