Font Size: a A A

Humanoid Decision-making And Path-planning For Lane-changing Of Unmanned Ground Vehicle On Highway

Posted on:2020-06-02Degree:MasterType:Thesis
Country:ChinaCandidate:F LiuFull Text:PDF
GTID:2392330575477755Subject:Control engineering
Abstract/Summary:PDF Full Text Request
Increasingly complex traffic environments lead to frequent traffic accidents,and intelligent driving technology is considered to be one of the effective ways to solve current traffic problems.With the rapid development of intelligent vehicle technology,decisionmaking and path-planning has become a new hot spot in intelligent vehicle technology research.Intelligent vehicles become new executives and decision makers,and drivers no longer have the right to execute and make decisions.In order to improve the ride comfort of the intelligent vehicle on the basis of satisfying safe driving,it is hoped that the intelligent vehicle has the ability to make humanoid decision-making and path-planning of the lane-changing.Therefore,when the intelligent vehicle is used as the executive and decision maker,how to realize the decision-making and path-planning for lane-changing of the intelligent vehicle like the driver is the problem of this paper.This paper is supported by the National Natural Science Foundation of China project three(No.61790563)'Human-machine Cooperation Mechanism and Switching Control under Extreme Operating Conditions',National Natural Science Foundation International(Regional)Cooperation and Exchange Key Project(No.61520106008)'Safety-Oriented Energy Efficiency Moving Horizon Optimization for Electrified Vehicles' and National Natural Science Foundation of China(No.U1664263)'Dynamics modeling and cooperative control method of driver-automation copiloted intelligent vehicles'.This paper studies the lane-changing decision-making and path-planning for highway.The main research contents are as follows:In view of the key points of research,this paper first understands and analyzes the driver characteristics,decision-making and path planning of intelligent vehicles,and proposes some improvement strategies for the shortcomings and problems existing in existing algorithms.This paper analyzes and understands the driving styles of different types of drivers through driving simulation experiments,and extracts reasonable feature data to characterize various types of drivers.Then,this paper analyzes different lanechanging decisions and different path planning behavior characteristics for different types of drivers,and proposes a new lane-changing decision-making and path planning method—humanoid decision-making and path-planning for lane-changing.For the problem of humanoid decision-making and path-planning for lane-changing,this paper first designs the driving simulation experiment,obtains the driver driving data,and analyzes,extracts and processes the initial driving data to obtain a data set that can characterize the driver's driving characteristics.The data set is divided into two parts:the training set and the test set.In order to accurately identify the driver's driving characteristics,it is also necessary to use the data of the training set combined with the k-means clustering algorithm to design the driver characteristic classifier and the neural network to design the driver characteristic recognizer.Finally,the paper verifies the validity of the driver characteristic classifier and the driver characteristic identifier using the verification set data.Considering that different types of drivers will have different lane-changing decisionmaking behaviors,this paper designs a kind of humanoid decision-making and pathplanning for lane-changing method for highway scenes.Firstly,considering the different types of drivers' different expectations for the driving speed of the vehicle and the safety distance between the vehicles,this paper designs a lane-changing possibility recognizer based on the speed expectation,the gap expectation and the end expectation.Then,considering that different types of drivers have different requirements on the longitudinal safety distance and the lateral safety distance during the lane change of the vehicle,this paper designs a lane-changing gap selector.The decision module generates a lane change request signal when there is both a lane change possibility and a reasonable lane change gap.Finally,the paper verifies the practicability and reliability of the humanoid decisionmaking and path-planning for lane-changing method for different scenarios.After obtaining the lane change request signal,and also facing the highway scene,this paper designs different driving paths based on the artificial potential field method for different types of drivers.Firstly,considering the different types of drivers' fear of the distance between the two vehicles,this paper designs a repulsive field and a gravitational field with different repulsion coefficients and gravitational coefficients.Then for this research,this paper improves the artificial potential field method and forms a new segmented multi-objective artificial potential field method to guide the intelligent vehicle to the designated lane change position,guide the intelligent vehicle to complete the lane change,and guide the intelligent vehicle to move forward.Finally,this paper proves the feasibility and security of intelligent vehicle path by comparing intelligent vehicle path and weekly vehicle path,and realizes humanoid path planning.
Keywords/Search Tags:Driving simulation experiment, Driver characteristics, Neural network, Lane-changing decision-making, Path planning
PDF Full Text Request
Related items