New method for the calibration of a road condition sensor

Abstract

Purpose This study aims to improve the calibration accquracy of the road condition sensor. A road condition sensor is widely used to detect water or ice deposits on the road to assess associated driving risks. Its quantitative calibration is central to the thickness measurement accuracy and conventionally performed using the single fitting method-based calibration method. One existing limitation is that the distribution characteristics of calibration data of different road conditions are diversified, which leads to the fitting error can not be minimized when using the conventional calibration method. Thus, the multiple fitting methods-based calibration method is developed to realize an optimal calibration for the road condition sensor. Design/methodology/approach A fitting method assignment for the calibration data of different road conditions was introduced to realize an optimal combination for fitting method and calibration data. In the experiments, the calibration methods were tested in the absence of measurement errors, then tested with calibration data, and finally, in the covering thickness measurement. Findings The comparison results indicate that compared with the conventional calibration method, the multiple fitting methods-based calibration method cuts the fitting error in the quantitative calibration by 13.3% and contributes to reducing the thickness measurement error by 8.11% for different road conditions. Originality/value The multiple fitting methods-based calibration method has been successfully applied for quantitative calibration and shown to reduce calibration errors. The comparison between different calibration methods demonstrates the superior performance of the new calibration method.