Automobile rollover warning algorithm based on SVM (Support Vector Machine) model

Abstract

Theinvention discloses an automobile rollover warning algorithm based on an SVM (Support Vector Machine) model, belonging to the field of traffic security. The algorithm comprises the following stepsof: a, selecting an automobile model for design of a rollover warning controller, and creating a driver-automobile-road environment model; b, determining a rollover risk evaluation index; and c, building an automobile rollover risk prediction experience model by using an SVM algorithm, determining an explicit function of an automobile rollover limit state function with a deterministic variable anda random variable output by the driver-automobile-road environment model as inputs of the automobile rollover risk prediction experience model, and visibly dividing automobile state samples into a rollover insecurity area and a rollover security area by using a hypersurface. A probability method considering uncertainty is applied to automobile dynamic rollover warning algorithm design, an explicit function for automobile rollover security limit and gradient thereof is provided, external interference of system strong nonlinearity and uncertainty is lowered, and the automobile dynamic rolloverwarning accuracy is improved effectively.

Description

technical field [0001] The invention relates to the field of traffic safety, in particular to an algorithm for early warning of vehicle rollover, which is used for research on early warning of vehicle rollover. Background technique [0002] Traffic accidents are an important form of serious loss of life and property, and vehicle rollover accidents are one of the most serious traffic accidents. According to statistics, vehicle rollover has become the most frequent accident after frontal collision. In 2014, although the car rollover accident rate in France was only 4.7%, 14% of the people were killed. At the same time, it is very easy to damage roads, bridges and other public facilities after rollover, and even cause serious environmental pollution. Therefore, the research on early warning and control of vehicle rollover has received sufficient attention worldwide. The research on vehicle rollover warning and control can be classified into two categories: active and passive ...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved in the present invention is to develop a vehicle rollover warning algorithm based on the SVM empirical model considering uncertainty, and apply the probability method considering uncertainty to the design of the vehicle dynamic rollover warning algorithm, considering the driver- The uncertainty of the vehicle-road system and the real variability of parameters provide an explicit function of the vehicle rollover safety limit and its gradient. Non-linear and uncertain external disturbances can effectively improve the accuracy of vehicle dynamic rollover warning

Images