Automobile’s Vertical Anti-Collision Information Detection And Warning Strategy Research

With the rapid development of urban traffic and vehicle ownership, the accident rate rise quickly, traffic safety crisis is getting more and more serious, and thus the automotive active anti-collision system attracts more attention. In this context, study on the information detection and warning strategy has been conducted, the study covers with the research of information acquisition and de-noising, the safety distance model and the warning strategy, etc.The traffic information of advanced warning system and driving operations of different drive tendency are analyzed in this paper. Appropriate sensors, such as speed sensors, radar and angle sensors are selected, designed corresponding signal collection circuits, built traffic information detection system, provided data supports for the following safety car model based on driving operation.It is evitable to be interfered by a large amount of noise in the process of signal gathering and transmission, in order to improve the accuracy of detected information and safety distance model, the original signal needed de-noising process before modeling. The wavelet transform has better ability to signal processing and partial characteristic both in the time domain and the frequency range, is employed in signal de-noise. Maximum energy measure, maximum energy-to-Shannon entropy ratio and other indexs are evaluated for appropriate base wavelet selection. A new thresholding function is proposed based on optimization parameter selections, and tested definite de-noising effect on MATLAB platform.Driving operation tendency is judged base on BP model with collected de-noising sensor information. On the basis of drivers’actual driving braking process, multi-grade warning (the preset driving, critical risk warning and danger alarming) safety distance model established, in addition to the vehicle and environment factors, the key point is the driver’s behavior parameters. Finally the safety distance model based on driving operation is constructed, which adopted most roads.Alarming unit, different alarming levels and warning strategy for different driving requirement are designed. Vehicle braking system dynamics model, safety distance calculation model are set up on the MATLAB platform, analyzed several typical riding conditions such as static targets, low-speed targets and deceleration targets. The results showed that the warning strategy and models are rationality and availability.