A Database Retrieval Method for the Prediction of Reduced Luminance Coefficient Tables of A Road Surface Based on Measurements in Situ

Abstract

ABSTRACT The in-situ measurements for r-tables have attracted special attention owing to their advantages, including increased speed, simplicity of use, and nondestructive road effects. This study presents a database retrieval method proposed for the construction of a full-scale r-table from an r-table database with 25 r coefficients used as input parameters. The proposed method was used in an in-situ, r-table measurement device as a data processing algorithm. To evaluate the accuracy of the proposed method, we performed a) a table-to-table comparison of Q0, S1, and r-tables constructed by different methods and b) comparisons of calculated luminance values (Lave, U0, Ul ) for a typical road-lighting scenario based on the use of different r-tables. The table-to-table comparison results indicate that following the application of the proposed method, its calculated Q0, S1, and r-table values were very close to their actual values (EQ0 = 2.1%, ES1 = 5.0%, Error = 2.3%). Moreover, the proposed method also exhibits an improved luminance response and achieves a 2.1% difference in Lave , a 3% difference in U0 , and a 1.8% difference in Ul . This study proves that the proposed database retrieval method improves the model prediction accuracy.