Research On ABS Wheel Speed Processing And Estimation Of Absolute Vehicle Speed

A huge convenience was brought by the car to people, but the safety problem can not be ignored, so a series of technical problems needs to solve during the car’s design. The anti-lock braking system which is closely related to the automobile safety performance is a focus of the current research. Without the ABS system, the drift, direction failure and other dangerous situations are prone to happen during the braking; the ABS system reliability is an indicator of automotive safety performance, so more and more attention was paid by the manufacturers and the consumers to the reliability of the ABS system.The wheel speed signal processing and reference speed estimation which is related to the reliability and the accuracy of the of automobile ABS control system, are the key issues of the automobile ABS control system. The focus of this article is to study the method of ABS wheel speed signal processing and the reference speed estimation. Firstly, the overview of the development of the ABS, the structure as well as the working principle of ABS were introduced, meanwhile the significance of the wheel speed signal processing and the reference speed estimation of the ABS control system were made clear. Secondly, the detailed description of wavelet theory was given, including the continuous wavelet transform, discrete wavelet transform, dyadic wavelet transform, Mallat algorithm and several wavelet denoising methods. On this basis, the wavelet transform denoising methods were introduced and the theoretical principles and the steps of the modulus maxima denoising method were given. The result of this method in the MATLAB platform shows that the wavelet transform modulus maxima filtering method can effectively remove the noise from the ABS wheel speed signal and improve the signal-to-noise ratio. Finally, the data fusion concept, its advantages, theory, and the Kalman filtering data fusion method was introduced. Traditional reference speed estimation methods were also analyzed, including the slope method, maximum wheel speed method, complex method. The result of data fusion in the MATLAB platform was compared to the result of the traditional methods. The comparison told us that the Kalman filter based on data fusion got a satisfied result.