小波分析在机车优化粘着控制中的应用

Applications of the wavelet analysis theory to optimized locomotive adhesion control

分析了机车粘着控制中粘着利用机理,提出了空转趋势识别方法,并采用小波变换这一工具对含有随机干扰信号的瞬时轮周加速度信号进行了多分辨率分析,即把轮周加速度信号分解为多个层次的细节部分和平滑部分。轮周加速度信号中的噪声被分解到了各层的细节部分,由剩余的平滑部构成轮周加速度发展趋势曲线,并对其微分。然后根据其加速度微分值进行空转趋势识别。最后,给出了仿真结果表明:基于空转趋势识别方法的模糊粘着控制能充分利用粘着力,避免空转发生;小波分析方法的应用有效避免空转趋势误识别,提高了粘着控制稳定性。

This paper analyses locomotive adhesion control mechanism and presents a new method of identifying slip tendency. A wavelet transform method to process wheel acceleration signal containing random noise is proposed. The original wheel acceleration signal is decomposed into approximation components and detail components at several levels. This decomposition process is called multiresolution analysis. After the decomposition, the noise in the wheel acceleration signal is separated from the approximation components and stored in the detail components. From the approximation components, the wheel acceleration tendency is estimated and differentiated. Based on this differential value, the wheel slip is identified. The simulation results show the application of wavelet transform paves a new way to effectively avoid wrong slip identification, enhance control stability and fully utilize adhesive force. 