Research On Decision & Control Of Front Wheel Angle Of Intelligent Vehicle Lane Following Control

With the maturity of advanced driver assistance systems,many universities and enterprises have gradually shifted their efforts to the research and development of high-level autonomous driving.Currently,the research on autonomous vehicles are mainly focused on the following three aspects: environmental perception,motion planning and decision making,and vehicle control.In this context,this paper proposes a research on the decision and control of the front wheel angle of intelligent vehicle lane following control.Based on the understanding of the domestic and foreign research status of lane following control,this paper compares the different lane following control algorithms and finally start the research of the front wheel angle decision and control algorithm of lane following control based on the "Preview-Follow" theory.Firstly,the vehicle lateral dynamics model is established in the case of vertical and horizontal dynamic decoupling,and the feedforward control law of the front wheel rotation decision algorithm of lane following control is derived under the assumption of constant lateral acceleration.In addition,in order to improve the robustness of the system and avoid the tracking error caused by the model mismatch,this paper introduces the lateral acceleration error feedback module,which constitutes the form of “Feedforward & Feedback”.In order to realize the precise control of the front wheel angle with the torque interface of EPS,the C-EPS dynamics model is established based on the research of the response characteristics of C-EPS,and designed the front wheel angle control algorithm through combining the traditional "three closed loop" position control of the motor with C-EPS dynamics model.In order to verify the effectiveness of the proposed algorithm,this paper builds the MIL and HIL simulation platform based on SCANeR software,and builds the model of decision and control algorithm of the front wheel angle of intelligent vehicle lane following control in Simulink.Finally,the effectiveness of the algorithm is verified by MIL and HIL simulation test.Considering that the decision and control algorithm of the front wheel angle of intelligent vehicle lane following control proposed in this paper is designed for L2 and L2 above autonomous vehicles,and the autonomous vehicles above L2 level should have the ability to change lanes autonomously.Therefore,this paper also did some research on the effectiveness of the proposed algorithm in the autonomous lane change scenario.Firstly,this paper proposes a decision-making algorithm based on TTC by the analyzing the traffic accidents that may occur in lane change.In addition,this paper proposes five evaluation price indicators for the lane change path planning algorithm,and compares the different path planning algorithms with the proposed indicators.Finally,the vehicle routing path is planned based on the fifth-degree polynomial.In order to verify the effectiveness of the algorithm in the lane change scenario,this paper builds a lane change path planning algorithm model in Simulink and integrates it with the model of decision and control algorithm of the front wheel angle of intelligent vehicle lane following control,and the effectiveness of the proposed algorithm in the lane change scenario is verified by MIL and HIL simulation.In order to furtherly verify the effectiveness of the designed algorithm on the real vehicle,this paper upgrades and modifies the original platform vehicle.In addition,the data acquisition system CAN_Insight and BootLoader software were optimized and improved.Finally,the actual vehicle test was carried out and verified that the vehicle can be accurately controlled along the lane center with the algorithm,and the maximum error does not exceed 25 cm.