| As the number of car owners continues to grow and the frequency of traffic accidents follows,intelligent assistance systems for car offer a solution to this problem.Active collision avoidance as one of the active vehicle safety technologies has a clear effect to prevent or slow down collision accidents,so it receives increasing attention.To improve the effectiveness of active collision avoidance systems,this paper examines the issue of active collision avoidance due to car braking or track change.First of all,to address the safety problems posed by the road adhesion conditions to avoid vehicle collages,an estimator of the road adhesion coefficient was designed based on the kalman filtering algorithm without traces.By analyzing the dynamic vehicle model,reasonable simplification to establish the dynamic vehicle model at three degrees of freedom,and to attach the estimated number of condition observations to establish the Doguff tire model,calculated the normalized tire force.The trace free kalman filtering algorithm was then studied according to its characteristics and an estimator of the pavement adhesion coefficient was designed.Finally,to validate the designed algorithm,a simulation analysis was carried out.The results show that the algorithm meets performance requirements.Second,assessed the risk of longitudinal transverse collision to complete the design of the evasive decision system.Only one evaluation method has certain limitations.A risk assessment method for collision time margins was developed and a safety distance model was added taking into account the pavement adhesion factor.The active evasive course of the turn was planned by a fifth degree polynomial,the safe distance of the turn was calculated,and the lateral acceleration during the turn was selected based on the restraining limits.Third,in the collision avoidance control system,controllers were designed separately using longitudinal cross-section to reduce the difficulty of movement control.First,in the design of the longitudinal collision avoidance control system,longitudinal acceleration was used as the interface and hierarchical control was chosen.The upper controller used a linear quadratic controller(LQR)to control and resolved the desired longitudinal acceleration.The lower controller used the inverse model of vehicle dynamics to achieve the corresponding opening of the main brake cylinder gas/pressure lever,at the same time,the PID feedback controller was designed for the drive braking system and the gas/brake switching logic was designed with the vehicle rolling curve as compensation.Secondly,in the field of lateral controllers,the dynamic vehicle model was further simplified: a linear vehicle model with two degrees of freedom combined with a lateral error model under the Frenet coordinate system was used as a predictive model.Finally,joint simulations of the transverse and longitudinal controllers were conducted and their effectiveness was examined separately.Finally based on Carsim/Simulink co-simulation.Designed different simulation working conditions,the collision avoidance control system according to the decision design,chose different collision avoidance modes,the results show that this active avoidance system can accurately complete the collision avoidance decision,control,to realize the function of reducing the collision accident. |
