Analysis And Improvement Of Multi-Frequency Sensing Algorithm For Road Surface Coverings

With the rapid development of modern transportation industry, the impact of weather and road surface conditions on transportation has become increasingly prominent. Poor road surface conditions, especially icy roads, severely reduce the efficiency of road transportation and even threaten the safety of life and property. Thus, the intelligent road sensor system has become an important part of the transportation safety management system. Because it can not only offer the information of the severe weather timely and accurately, but also provide reliable guidance for travelling. The road surface coverings detection technology in the intelligent road sensor system plays more and more important roles in sensing icy and other extreme weather.This thesis proposes a new algorithm to solve the temperature effect problem of the multi-frequency module of the intelligent road sensor systems. The algorithm improves the systems' ability of distinguishing road surface covering types, such as dry, wet, water, ice, snow and so on. Investigation on several different common road surface coverings resolution technologies in the market are conducted at first. Then the polarization characteristics of several common road surface coverings and detection principle of the multi-frequency module are analyzed in this thesis. Based on the analysis of the detection principle and the structure of cylindrical sensitive capacitor of the multi-frequency module, a mathematical model and the theoretical capacitance formula are deduced. After analyzing the previous road surface coverings resolution algorithm, a novel one is proposed in this thesis. The algorithm proposed is based on the Lagrange Interpolation Algorithm. Two intermediate curves are established, which divide the coordinates into more rational regions to distinguish the category of the road surface coverings.Finally, the algorithms are realized and tests in experiments. The experimental results show that, by dividing rational regions, the proposed algorithm not only increases the resolution tolerance and accuracy, but also solves the temperature effects problem. As a result, the resolution performance of the multi-frequency detection module has been greatly improved and the resolution accuracy reaches 90%.