On The Active Safety Collision Avoidance Control System For Electirc Vehicle

With the rapid development of new energy vehicle, active safety issues forelectric vehicle (EV) are attracting more and more attention from companies andresearch institutions, which have become the most promising project in the automibleindustry.Vehicle collision avoidance system began to study in the1960s, led by developedcountries, such as Germany and America. The original target of collision avoidancewarning system was to supervise the highway conditions and assist the drivers toavoid collision. While EV research started in the1990s as a result of energy crisis andenvironmental crisis. Active safety collision avoidance system for EV is a newengineering, which is designed to guarantee drivers’ safety and improve vehicleenergy efficiency. Currently, there are still a lot of unresolved issues.In such circumstances, supported by national EV project"electric vehicle activesafety technology research", this paper focuses on vehicle safty distance model inlongitudinal braking and lateral lane-changing and corresponding controllers design,and then verifies the correctness of the model for collision avoidance systems withMatlab/Simulink simulation and analysis. This paper mainly includes four aspects:(1) According to overall structure of EV, models are built for vehicle’s maincomponents, such as vehicle model, tire model, motor model. A seven DOFs vehiclemodel is chosen in view of vehicle longitudinal movement, lateral movement, yawmotion and four wheels rotary motion respectively. Typical Pacejka model is chosenas tire model. Moreover PMSM motor model is used as motor mode for EV. Finallyan experiment is carried out in Matlab/Simulink.(2) Lane-changing and longitudinal braking safety distance models are builtaccording to the process of collision avoidance. For longitudinal braking process,three kinds of safe distance expressions are obtained, which are distinguished from theleading vehicle’s status, such as static condition, uniform motion condition andemergency braking condition. For lateral lane-changing process, safe distance modelsare given according to constant velocity and accelerated movements in longitudinaldirection. Finally safety evaluation indexes are set up as traffic safety criterion.(3) Collision avoidance controllers are designed in accordance with controlrequirements. First, system function is defined to lay the foundation for controllers design. And then longitudinal controller is established by optimal control theory,which can provide upper-level decision-making for low-level control implementation.Finally optimal feedback matrix for lane-changing is obtained by LQR theory;deflection angle tracing is controlled by feed-forward compensation strategy based oninput back-off, which can ensure that system can accurately follow a given desiredyaw angle.(4) Simulation system is built in Matlab/Simulink and tested in dSPACEexperimental platform. Two methods are analyzed to verify the validity of thecollision avoidance control system.In a word, collision avoidance system studied in this paper can effectivelyachieve vehicle safety requirements and verify the validity of control strategy used inthis paper, which lay a foundation for further deeper research on EV active safety.