Research On Vehicle Trajectory Prediction In Autonomous Driving Environment Based On Improved C-GAN Model

Under the background of the country’s vigorous development of intelligent transportation,autonomous driving technology has become a new development direction of road transportation,promoting vehicles to realize various intelligent functions,and then assisting in solving various pain points in road transportation,which has become an inevitable trend of urban transportation development.Autonomous driving technology involves the integration of multiple module technologies such as positioning,perception,planning,and control.The planning module refers to the ability of autonomous vehicles to plan vehicle movements and complete designated driving tasks without human control,so it directly determines the safety,rationality and practicability of autonomous driving.Based on the current autonomous driving technology,this article analyzes the road environment data under the complex environment assumed in this article,uses deep learning methods to predict the trajectory of the vehicle,and provides dynamic decisions for the trajectory of the autonomous vehicle.In this paper,a trajectory model based on C-GAN network is designed.First,by predicting the trajectory of environmental vehicles,and planning the driving path of the autonomous vehicle based on the prediction result,so as to realize the safe driving of the autonomous vehicle in a complex environment.First,this article gives a brief overview of the current research on autonomous driving trajectories and discusses the technological development of autonomous vehicles in the context of the data-driven era.The vehicle trajectory prediction model based on deep learning technology uses the vehicle trajectory of the real road environment as a learning sample to train,verify and test the trajectory model,so that it can dynamically adjust the prediction results according to environmental changes in time,and finally predict the trajectory of environmental vehicles.Secondly,based on the above prediction results,this paper uses the environmental vehicle trajectory data as input parameters to design an automatic driving trajectory planning algorithm that considers the trajectory of surrounding vehicles.The trajectory planning algorithm is also based on the C-GAN model,takes the predicted trajectory of the environmental vehicle and the initial trajectory of the autonomous vehicle as input,and outputs a planned driving trajectory that conforms to the current environment.In addition,a polynomial fitting method is used to perform curve fitting on the planned trajectory points,so that the fitted driving trajectory is safe and effective,and the autonomous vehicle can smoothly complete the driving target.Finally,this paper verifies the validity of the model through error analysis and simulation calculation.Test and analyze the model based on MATLAB and Python.By comparing the prediction results of multiple trajectory prediction models,the accuracy of the model is analyzed.And build a vehicle-road environment in the MATLAB simulation environment to simulate the vehicle trajectory in real traffic scenes.And verify the effectiveness and safety of the model in this paper when dealing with various driving trajectories.Based on deep learning technology,this paper studies the trajectory prediction model and trajectory planning model of autonomous vehicles in a complex vehicle-road environment.The model simulation verification results show that the application of the model is effective and reliable in the designed environment.Through the research of autonomous driving trajectory planning in this article,it provides ideas and reference basis for the trajectory research of autonomous vehicles,and it helps the development of autonomous driving technology in the entire industry.