Research On Decision Making And Planning Of Obstacle Avoidance For Intelligent Connected Bus At High Speed

Intelligent connected cars have huge potential in improving traffic safety and alleviating traffic pressure,which has become a research hotspot in the automotive field.Among them,the obstacle avoidance problem under high-speed conditions is of great significance to avoid major traffic accidents,which is a key part of realizing vehicle intelligence.This dissertation takes intelligent network-connected buses as the research object,and studies the decision-making planning of obstacle avoidance behavior to realize the vehicle’s safe obstacle avoidance under different road conditions.Based on the analysis of the structural characteristics and driving performance of the operating bus,the performance parameters of the vehicle were determined.Based on the parametric modeling ideas of the Truck Sim software,a vehicle dynamic model was established.Dry and wet typical structured roads are constructed.Based on the fishhook experiment and braking distance experiment,the dynamic response of the model under different road conditions is analyzed,and the dynamic equation is used to verify the reliability of the model.Carry out research on the obstacle avoidance behavior of vehicles under high-speed conditions,and divide the obstacle avoidance methods into lane change obstacle avoidance,following vehicle obstacle avoidance and parking obstacle avoidance.By analyzing the obstacle avoidance scenario,obstacles are divided into stationary obstacles,low-speed vehicles and emergency braking vehicles according to the state of motion.Utilizing the environmental perception ability of vehicle to X technology,an obstacle avoidance safety distance model that fully considers the movement state of the vehicle and the road attachment conditions is established.by analyzing the factors that affect the decision of obstacle avoidance behavior,the road adhesion coefficient,vehicle acceleration,speed and distance information are used as model inputs,and the obstacle avoidance behavior decision model is established based on the layered finite state machine theory,which selects the reasonable way in different obstacle avoidance scenarios.The feasibility of the decision model is verified by using Truck Sim and Simulink joint simulation platform,Research on trajectory planning for lane changing and obstacle avoidance and expected deceleration planning for following and stopping obstacles.The trajectory planning is divided into two parts: path and tracking speed.Based on the Bezier curve,the path planning algorithm combines the curvature of the path and the target state constraints,and uses the maximum curvature as the target function.Prevent the vehicle from rolling over and slipping due to excessive curvature at high speed.The tracked speed range is determined based on the lanechanging safety distance of the target lane and the vehicle dynamics characteristics.In order to meet the driving efficiency of high-speed conditions,the maximum safe speed is selected using the speed difference as the evaluation index.According to the braking safety distance of the vehicle and the road surface adhesion conditions,the boundary value of the braking deceleration is calculated.The expected deceleration of following the vehicle and stopping obstacle is planned by adjusting the weight ratio in different scenarios.Finally,three obstacle avoidance methods is simulated and validated by using Matlab and Truck Sim software..Combining with modern communication technology to obtain local traffic conditions,the results show that the planned lane change obstacle avoidance trajectory and braking deceleration meet the vehicle dynamics requirements,which can realize efficient and safe obstacle avoidance.