Research On Extension Control Of Lane Keeping/departure Assistance System

Lane keeping/departure assistance system,as a typical vehicle active safety technology,can take the corresponding measures to control the vehicle driving safely in the lane in time when the vehicle has a dangerous trend.This also reduces the traffic accidents caused by the drivers' fatigue or misoperation,and improves the safety of the vehicle.Based on the steering and braking system,the theoretical and experimental research on lane keeping/departure assistance system including TSK fuzzy extension control and extension combination control are carried out in this thesis.Firstly,the thesis introduces the background of the driving assistance system and the significance of its in-depth study.Then,the research status of lane keeping/departure assistance control is analyzed in details.According to the existing problems in control algorithm of lane keeping/departure,the key contents of this thesis were put forward.Secondly,in traditional lane keeping system,a single control algorithm was used for the global control,which often results in poor control affect in some control areas.For this problem,the extension control strategy was proposed for lane keeping control.The extension control strategy based on area division and control theory divides different control regions according to the state of the system,and adopts different control algorithms in different control areas to solve the limitation of adopting single control algorithm.Considering the problem that the boundary division of the control domain based on the artificial experience and the parameter is difficult to determine,a method based on the feature plane was proposed to divide control boundary of the domain.In order to further improve the performance of extension control,combined with TSK fuzzy control theory,the TSK fuzzy extension controller was designed to improve the performance of extension control and lane keeping assistance system by smoothly switching different control algorithms.Thirdly,aiming at the limitations of two kinds of lane departure assistance control methods based on electric power steering and differential braking,an extension combination control strategy was proposed.Based on extension control theory,the road environment information and vehicle state are fully considered in the design of the combination controller,and the controller combines electric power steering and differential braking to realize the lane departure assistance control.In order to solve the problem of human-machine coordination in the process of departure assistance,the fuzzy neural network control was applied to design the human-machine coordination controller which considers the driver torque and the lateral deviation of the vehicle.It can adjust the assistant torque of the lane departure assistance system dynamically by outputting the weight,and realize the coordinated control between the driver and the assistance system.Finally,in order to verify the effectiveness of the proposed control strategy,a co-simulation of CarSim/Simulink was carried out.At the same time,in view of the specific research problems and combining the existing conditions in the laboratory,the hardware in the loop test platform was built for test verification.The results show that lane keeping assistance system can effectively keep the vehicle running around the lane center line,and improve the lane keeping control accuracy and vehicle stability,ensuring the safety and comfort of the vehicle.Lane departure assistance system can avoid vehicle deviating from the driveway while reducing the interference between the driver and the assistance system,reducing the man-machine conflict,and having better man-machine coordination performance.