Pavement Spectrum Measurement And Performance Index Analysis Based On Vehicle Acceleration

Pavement roughness is an important index of grade evaluation,construction and acceptance of pavement, which directly reflects the comfort, safety and the use state of road surface, it also provides an important reference for the repair, and maintenance of pavement. So how to measure the pavement spectrum quickly and analyze its performance index accurately has become a practical subject. With the rapid development of embedded technology and sensor technology, and the maturity of vehicle suspension theory, the research on collecting pavement spectrum data by non-contact detection method has gradually become the main trend. In this paper, a system which is realized road pavement spectrum measurement by obtaining vehicle body acceleration without additional sensor is designed, and it is also used to analyze and process the performance index of pavement spectrum data.In this paper, the feasibility of the system is proved by theoretical deduction. We firstly focus on the vehicle suspension model of the system by the modeling software, ADAMS; And then in order to realize the data acquisition function, a lower machine is designed and developed, its core controller is ARM Cortex-M4 MK60DN512ZVLQ10 which produced by Freescale company. Then, a upper machine of the system is designed and developed based on the platform of C#.net language, which controlling the lower machine, analyzing and processing the collected data; In the end, the experiment of vehicle is tested, and analyze the collected data by using the optimization deep learning algorithm to realize pavement grade evaluation. The test results show that the design of the system which realized the road spectrum soft measurement by using vehicle body acceleration is feasible and reasonable. Compared with the current situation that domestic road test equipment is imported, and the price is expensive, the system has good practicability and innovation. At the same time, it is feasible and efficient to use the optimization deep learning algorithm in the classification of road surface roughness.The main completed contents of this article include:Firstly, introduced the theoretical deduction of the system;Secondly, designed the lower machine of the system;Thirdly, designed the upper machine of data acquisition system based on C#.NET platform;Finally, analyzed the results of original data collected by the test through the optimization deep learning algorithm, and analyzed the performance index of the pavement spectrum, at last realized the 3D reconstruction of pavement.