Development and verification of integrated photoelectric system for noncontact detection of pavement ponding and freezing

Abstract

The ponding water and ice on the road surface can cause degradation of skid resistance that seriously affects driving safety. In order to accurately detect road surface conditions, an integrated photoelectric system was developed composed of an infrared light-emitting diode and a phototransistor. Typical asphalt pavements such as AC-13, SMA-13, and OGFC-13 were selected to simulate four road surface conditions: dry, wet, ponding water, and ice. The voltage intervals for dry and wet surface conditions were proposed by the light intensity method. The functional relationships were established between the depth of ponding and the output voltage, the thickness of ice, and the output voltage, respectively. The noncontact detection system for pavement ponding and freezing was developed by the design of the hardware of real-time controller, the software of supervisory computer, and the shell of entire system. Then the system was verified in the field conditions. According to the test results, there was a good linear relationship between the ponding depth and the output voltage, while there was a quadratic parabolic relationship between the ice thickness and the output voltage. The detecting accuracy was 91.67% for the dry and wet surface conditions. The detecting errors of ponding depth for three different types of pavement surface were 0.248, 0.168, and 0.189, respectively; the detecting errors of the ice thickness for three different types of pavement surface were 0.219, 0.261, and 0.326, respectively. The errors were within a reasonable range, and the detecting accuracy of ponding depth is higher than that of ice thickness.