Driver Lane Change Intention Recognition of Intelligent Vehicle Based on Long Short-Term Memory Network

Liang Tang,Hengyang Wang,Liang Li,Zhongyi Mei,Wenhao Zhang

doi:10.1109/access.2020.3011550

Abstract

Driving intention prediction is one of the key technologies for the development of advanced assisted driving systems (ADAS), which could greatly reduce traffic accidents caused by lane change and ensure driving safety. In this paper, an advanced predictive method based on Multi-LSTM (Long Short-Term Memory) is proposed to predict lane change intention effectively. First, the training data set and test set based on real road information data set NGSIM (Next Generation SIMulation) are built considering ego vehicle driving state and the influence of surrounding vehicles. Second, the Multi-LSTM-based prediction controller is constructed to learn vehicle behavior characteristics and time series relation of various states in the process of lane change. Then, the influences of prediction model structure change and data structure change on test results are verified. Finally, the verification tests based on HIL (Hardware-in-the-Loop) simulation are constructed. The results show that the proposed prediction model can accurately predict the vehicle lane change intention in highway scenarios and the maximum prediction accuracy can reach 83.75%, which is higher than that of common method SVM (Support Vector Machine).

Highlights

With the development of advanced sensor technology and artificial intelligence method, advanced assisted driving systems (ADAS) have been studied a lot in recent decades, which can effectively ensure driving safety, avoid traffic accidents, reduce energy consumption and improve ride comfort [1]–[4]
Lane change intention recognition methods can be divided into two types: driver behavior data based on prediction and vehicle trajectory data-based prediction
In this paper, a vehicle lane change intention recognition controller based on LSTM network is proposed, which can get high recognition accuracy for three lane change intensions, and show high robustness to the number of data samples

Summary

INTRODUCTION

With the development of advanced sensor technology and artificial intelligence method, ADAS have been studied a lot in recent decades, which can effectively ensure driving safety, avoid traffic accidents, reduce energy consumption and improve ride comfort [1]–[4]. Lane change intention recognition methods can be divided into two types: driver behavior data based on prediction and vehicle trajectory data-based prediction. We can analyze the driving trajectory data of vehicle to predict lane change intention. Before the large-scale application of artificial intelligence technology, the rule-based lane change intention recognition method was proposed in [20], which determines the vehicle motion maneuver by calculating lateral velocity cue and lateral position cue. [24] proposed a DBN-based LCD (Lane-changing decisions) model and LSTM-based LCI (Lane-changing implementation) model to predict LC process and testing results indicated that it had a good performance on accurately predicting lane change intention, and a meaningful conclusion is conducted that relative positions of the surrounding vehicles have an greater impact on driver decision making than relative speed. K =i−D where, xα(ti) denotes state value after filtering at ti, D is size of sliding window and is average window of intermediate data

INPUT VARIABLES

MULTI-LSTM MODEL

INFLUENCE OF NETWORK STRUCTURE AND DATA SRCTURE

EXPERIMENT VERIFICATION

Findings

CONCLUSION