基于小波时频图和量子遗传算法的车轮扁疤检测方法

A Wheel Flat Scar Detection Method Based on Wavelet Time-frequency Diagram and Quantum Genetic Algorithm

针对地铁车轮扁疤在线诊断困难问题,提出了一种基于小波时频图、灰度投影法和量子遗传算法的智能在线车轮扁疤检测方法。首先对轮轨间的垂向振动信号进行小波时频分析,得到小波时频图,然后结合灰度投影法从时频图中提取车轮扁疤的特征;最后采用量子遗传算法对支持向量机进行优化以提高SVM的分类准确率。通过一个工程实例验证了该方法在识别车轮扁疤时的有效性。结果表明,该方法能够准确且快速地检测出对车轮扁疤,且所使用的分类模型的性能优于现有模型。

To solve the difficulty in online diagnosis of wheel flat scars of metro,an intelligent online diagnosis method of out-of-round wheel based on wavelet time-frequency diagram,gray-projection method and quantum genetic algorithm is proposed.Firstly,the wavelet time-frequency analysis is carried out on the vertical vibration signal between the wheel and rail,and the wave-let time-frequency diagram is obtained.Then,the features of wheel flat scars are extracted from the time-frequency map combined with the grayscale projection method.Finally,the quantum genetic algorithm is used to optimize the support vector machine to im-prove the classification accuracy of SVM.An engineering example verifies the effectiveness of the method in identifying wheel flat scars.The results show that the method can accurately and quickly detect the wheel flat scars,and the performance of the used clas-sification model is better than that of the existing models.