Research on Road Covers Detection System With Multifrequency Excitation

In light of serious traffic accidents observed in recent years and the complexities associated with road detection environments, this project tackles the challenges of accurately measuring each road covering. Aiming at the limitations of the single-frequency detection method in the identification of icy and snowy road surfaces, the dielectric loss and relaxation phenomena existing in capacitors and coverings were studied based on the principle of dielectric polarization and the series equivalent model of capacitors, and then a capacitive detection method based on multi-frequency excitation was proposed. By analyzing the relationship between the dielectric characteristics of the covering and the frequency response, the equivalent model of the sensitive capacitor was constructed, the hardware circuit for multi-frequency impedance detection was rigorously designed, and the equivalent parameters were extracted in combination with the vector analysis method. A single-pole double-throw weighted resistor switching scheme was proposed to solve the problem of parasitic capacitance interference. The 10~100 kHz multi-frequency detection system was implemented based on the STM32 platform, and the coverage classification was realized in combination with the CNN model. The measured results indicate that in the environment of -30 to 60°C, the recognition accuracy rates of dryness, water accumulation, ice formation and snow accumulation of the system reach 97%, 99%, 95% and 86% respectively. Overall, the performance of this system meets the practical application requirement for road cover detection.