Research On Active Emergency Collision Avoidance Technology And Occupants Protection Performance For Intelligent Vehicles

In recent years,advanced driver assistance system(ADAS)has been rapidly developed and widely applied,and it has become the focus of global research Especially,autonomous emergency braking(AEB)systems have been applied to vehicles that can not only give warning to drivers,but also autonomously implement emergency braking.Due to AEB systems work in emergency and danger driving situation,therefore,AEB systems have a great influence on the driver's normal driving judgment,riding comfort and collision avoidance safety However,there are still many flaws in the performance of current AEB systems,and they do not have the function of lane change to collision avoidance Therefore,this paper proposes an active emergency collision avoidance system with autonomous emergency braking and autonomous emergency steering lane change functions,hoping improving the active safety technology of vehicles and solving the problem of casualties caused by road traffic accidentsThis paper focuses on the key technologies of improving and optimizing the comprehensive performance for intelligent vehicle's active emergency collision avoidance system(including the accuracy and reliability of target detection,the timeliness of the intervention timing,ride comfort,vehicle stability and safety during collision avoidance,etc.)and the technology of effectively restraining and protecting the occupants.The main research contents and innovations of this thesis are as follows:1.An active emergency collision avoidance system with driving mode decision mechanism was proposed,this system has the function of collision avoidance in vertical and horizontal directions,which improves the environmental adaptability of collision avoidance for intelligent vehicle2.The detection method of multi-sensors fusion was studied.The effective targets detected by the millimeter wave radar were selected by setting thresholds of reflection cross-sectional area,signal-to-noise ratio,lateral distance,relative speed and target detection life cycle.When using a visual camera to detect a target object,a vehicle classifier was trained by combining the Adaboost machine learning algorithm with Haar-like rectangular features to detect the target,the detected image and the size of the detection window were optimized to improve detection efficiency.In addition,the compression tracking algorithm was used to enhance the environmental adaptability and reduce the false detection rate and missed detection rate for the perception systemFinally,the data information perceived by the millimeter wave radar and the visual camera was fused in spatial scale and frequency scale.The accuracy and reliability of the perception system for detecting targets were verified by some real world vehicle experiments3.The control strategy of hierarchical warning and hierarchical braking were studied and the key performance parameters were optimized.Three levels of warning function were designed to extend the early warning time and improve the effectiveness of the early warning.A road surface automatic recognition system was established to provide real-time road adhesion coefficient for AEB systems,and hierarchical braking strategy was decided based on road adhesion coefficients.The algorithm of TTC value in time scale and the braking distance algorithm reflecting the characteristics of the brake were used to calculate the intervention time of AEB systems.Based on the above control strategy and performance parameters,a hierarchical controller of AEB systems was designed for planning the intervention timing and braking deceleration,and controlling the actual deceleration of the vehicle to accurately follow the desired deceleration.The feasibility and effectiveness of the designed hierarchical controller were verified by simulation experiments under various working conditions,and the safety and stability of the vehicle during autonomous emergency braking were also verified by simulation experiments4.AELC system is constructed to make intelligent driving vehicles adapt to more emergency and risk driving scenarios.Firstly,the lane change path is planned based on the geometric curve of the fifth degree polynomial;Then,the lane change performance is analyzed under different vehicle speeds and different lane change time;Further,the lane change path is optimized under the constraints of lane change safety,riding comfort,etc.Finally,the front wheel steering angle of the intelligent vehicle is controlled by using model predictive control(MPC)method to track the optimized lane change path.Furthermore,a joint simulation model is established based on Simulink module and Carsim,and the experimental results demonstrate that the intelligent driving vehicle can automatically complete path planning and trajectory tracking for intelligent lane change driving operation5.To reduce the injure of occupants during emergency collision avoidance,an active pre-tensioned seat belt was integrated with the emergency collision avoidance system to protect occupants.A simulation model of emergency collision avoidance was established using Madymo software.By comparing and studying the protection effects of three different kinds of seat belts on the occupants,the simulated results showed that the active pre-tensioned seat belt can provide a better protection effect for occupants during the emergency collision avoidance process.In addition,the performance parameters such as the preloading speed and preloading force of the reversible pretensioned safety belt are studied through bench experiments and volunteer real vehicle experiments.6.A set of test equipment for active collision avoidance system is developed based on the above theoretical research,simulation experiments,and the test requirements specified in China New Vehicle Assessment Program(C-NCAP)of 2018 version.In addition,some field experiments were carried out by using this developed test equipment to test the performance of AEB systemsIn summary,this thesis has carried out in-depth analysis and research on the key technologies for improving and optimizing the performance of the proposed active emergency collision avoidance system in many aspacts and study on the restraint protection performance for occupants during a collision avoidance process.In addition,a set of test equipment is developed and used to implemented field experiments for performance evaluation.