Abstract: The presence of water in the driveway is one of the key factors contributing to vehicle destabilization, and thicker water can even lead to “hydroplaning”. Therefore, real-time understanding of the thickness of accumulated water in the lane is crucial for drivers and autonomous vehicles. Based on the concept of smart tires, this paper proposes a triaxial accelerometer-based method to quantify the water accumulation in the lane. The fluid-solid coupling model between tire-water film-lane is established through ABAQUS, and the three-dimensional acceleration signal at the center of the tire liner is analyzed, and the radial direction acceleration signal is selected as the research object. The signals were acquired and processed using Butterworth low-pass filtering with a cutoff frequency of 450 Hz, aiming at removing high-frequency noise. The signal conversion from the global coordinate system to the local coordinate system is carried out, and the prediction model of lane water thickness based on BP neural network is constructed after obtaining the eigenvalue data matrix. The results of real-vehicle tests show that the triaxial accelerometer-based lane water accumulation quantification method proposed in this paper has high accuracy and good prospects for development and application.

Key words: intelligent tires, quantification of road surface humidity, triaxial accelerometer, neural network