Study On Speed Planning And Energy Recovery Strategy Of Pure Electric Vehicle Platoon Considering Lane Change Of Adjacent Front Vehicles

In the context of Io V,autonomous driving platoons may encounter interference from human-driven vehicles changing lanes.Such interference can lead to emergency reactions,including excessive speed reduction and emergency braking,which can compromise the platoon’s driving safety and efficiency.Consequently,this paper proposes a novel approach to address these issues by introducing a purely electric platoon speed planning and braking energy recovery strategy.This strategy leverages vehicle-vehicle communication to obtain the driving information of the vehicle in front,predicts its lane change intention and trajectory,and plans the platoon deceleration based on the predicted results to avoid excessive platoon deceleration and ensure driving safety.Furthermore,we establish a braking energy recovery strategy to reduce the platoon’s energy consumption while maintaining driving performance.Overall,our research offers a new modelling paradigm for intelligent networked platoon control in mixed traffic environments.To obtain driver lane change data that adheres to China’s traffic regulations,a real vehicle collection platform was established and ten drivers of diverse gender and age were recruited.Four distinct lane change working conditions(namely,left lane change,right lane change,straight line driving,and abandoned lane change)were designed to capture driver vehicle data under various scenarios.The collected data were filtered and normalized to create a dataset of lane change behavior.Subsequently,we employed a random forest algorithm to construct a lane change behavior recognition model,which was then evaluated based on metrics such as accuracy,recall,precision,and F1 score.A vehicle platform was designed to collect location information of vehicles during lowand high-speed lane change operations.The acquired data were processed through Kalman filtering and trajectory coordinate transformation,resulting in a dataset of lane change trajectories.A lane change trajectory prediction model was then developed by leveraging Bidirectional Long-Short Term Memory(Bi-LSTM)and calibrated via control variables.The performance of the Bi-LSTM model was compared with that of a traditional LSTM model in terms of trajectory prediction accuracy.A homogeneous platoon speed planning and energy recovery strategy is proposed that incorporates the recognition of preceding vehicles’ lane changing behavior and prediction of lane changing trajectories.This approach involves generating an ST diagram based on the predicted trajectory data of the preceding vehicle,which is then used to dynamically plan the initial speed profile for the pilot car.The main objective of this process is to minimize acceleration and its change rate while avoiding collisions with the preceding vehicle using a quadratic planning algorithm to optimize the speed profile.Additionally,a following vehicle speed following model is established to complete the overall speed planning of the platoon system.To further enhance energy efficiency,this paper introduces a braking energy recovery strategy that optimizes the proportion of electric braking force at the front axle relative to the total demand braking force.To verify the platoon control strategy proposed in this paper,a joint simulation platform based on Car Sim/Simulink was built to establish simulation conditions for both low-speed and high-speed lane changes of the preceding vehicle.The results show that the intelligent network platoon,under the control strategy of this paper,can effectively avoid excessive speed reduction when faced with interference from neighboring preceding vehicles during low-speed and high-speed lane changes.Additionally,it can reduce vehicle energy consumption through braking energy recovery by utilizing advanced planning techniques for deceleration.