The Design And Realization Of Intelligent Driving Test System And Key Algorithm

In recent years, with the development of technology, control platform of embedded system and product is also in constant progress. Android smartphone is more and more popular. The products of Android platform which is easy to use and has beautiful interface, fast and comfortable touch experience features will become the mainstream. Industry and driving test industry are pay more and more attention on the Android platform, Android platform control become a current research hot spot, however, the Android system often can’t satisfy the demand of real-time control, at the same time, the development of various peripherals driven for Android platform is difficult.In the above case, this paper studies and designs a kind of high-precision and intelligent vehicle driving test system based on GNSS difference locating, which can keep up with fashionable platform, and meet the real-time requirements of control. this paper studies the current mainstream CPU of control platform, selected S5PV210 and STM32F103ZET6 CPU as experiment test. For the research of this intelligent driving test system, this paper accomplished the following research work:Investigated the new platform architecture for the system, three-dimension model localization algorithm, vehicle pressing line detection algorithm. The hardware and software were designed and implemented, and it is applied to a practical driving training project.According to the deficiencies of the current driving test system platform, an kind of improvement scheme of the platform was given. Make full use of rich peripherals and interfaces of STM32F103ZET6 to design a kind of lower computer with strong extensibility, high real-time performance, cost-effective. Using the multi task processing and high speed characteristic of S5PV210, the Android system is designed, and the software of Android is designed.A kind of localization algorithm for three-dimensional model of the vehicle body based on differential positioning technology was given. According to the analysis of requirements for the actual driving test project, vehicle pressing line detection algorithm was given. By the analysis of the experiment results, it is verified that the improved algorithm is more efficient.