Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Modeling Electric Vehicle Car-Following Behavior: Classical vs Machine Learning Approach

Md. Shihab Uddin, Md Nazmus Shakib, Rahul Bhadani

TL;DR

This paper tackles EV car-following modeling in mixed autonomy traffic by benchmarking classical physics-based models (IDM, OVM, OVRV, CACC) against a Random Forest regression approach using a real-world EV dataset. Classical models are calibrated via RMSE between simulated and observed spacings, while the RF models predict acceleration and spacing from features such as spacing, speed, and gap type. The RF approach achieves markedly lower RMSEs across gap settings ($RMSE$ values of $0.0046$ for medium, $0.0016$ for long, and $0.0025$ for extra-long gaps) compared to the best physics-based performance ($RMSE=2.67$ for long gaps with CACC), highlighting the data-driven model’s superior accuracy. The findings underscore the potential of ML for accurate EV car-following simulations and mixed-autonomy traffic analysis, with implications for AV systems and TOD/traffic management; code and data are available at the provided GitHub repository.

Abstract

The increasing adoption of electric vehicles (EVs) necessitates an understanding of their driving behavior to enhance traffic safety and develop smart driving systems. This study compares classical and machine learning models for EV car following behavior. Classical models include the Intelligent Driver Model (IDM), Optimum Velocity Model (OVM), Optimal Velocity Relative Velocity (OVRV), and a simplified CACC model, while the machine learning approach employs a Random Forest Regressor. Using a real world dataset of an EV following an internal combustion engine (ICE) vehicle under varied driving conditions, we calibrated classical model parameters by minimizing the RMSE between predictions and real data. The Random Forest model predicts acceleration using spacing, speed, and gap type as inputs. Results demonstrate the Random Forest's superior accuracy, achieving RMSEs of 0.0046 (medium gap), 0.0016 (long gap), and 0.0025 (extra long gap). Among physics based models, CACC performed best, with an RMSE of 2.67 for long gaps. These findings highlight the machine learning model's performance across all scenarios. Such models are valuable for simulating EV behavior and analyzing mixed autonomy traffic dynamics in EV integrated environments.

Modeling Electric Vehicle Car-Following Behavior: Classical vs Machine Learning Approach

TL;DR

This paper tackles EV car-following modeling in mixed autonomy traffic by benchmarking classical physics-based models (IDM, OVM, OVRV, CACC) against a Random Forest regression approach using a real-world EV dataset. Classical models are calibrated via RMSE between simulated and observed spacings, while the RF models predict acceleration and spacing from features such as spacing, speed, and gap type. The RF approach achieves markedly lower RMSEs across gap settings ( values of for medium, for long, and for extra-long gaps) compared to the best physics-based performance ( for long gaps with CACC), highlighting the data-driven model’s superior accuracy. The findings underscore the potential of ML for accurate EV car-following simulations and mixed-autonomy traffic analysis, with implications for AV systems and TOD/traffic management; code and data are available at the provided GitHub repository.

Abstract

The increasing adoption of electric vehicles (EVs) necessitates an understanding of their driving behavior to enhance traffic safety and develop smart driving systems. This study compares classical and machine learning models for EV car following behavior. Classical models include the Intelligent Driver Model (IDM), Optimum Velocity Model (OVM), Optimal Velocity Relative Velocity (OVRV), and a simplified CACC model, while the machine learning approach employs a Random Forest Regressor. Using a real world dataset of an EV following an internal combustion engine (ICE) vehicle under varied driving conditions, we calibrated classical model parameters by minimizing the RMSE between predictions and real data. The Random Forest model predicts acceleration using spacing, speed, and gap type as inputs. Results demonstrate the Random Forest's superior accuracy, achieving RMSEs of 0.0046 (medium gap), 0.0016 (long gap), and 0.0025 (extra long gap). Among physics based models, CACC performed best, with an RMSE of 2.67 for long gaps. These findings highlight the machine learning model's performance across all scenarios. Such models are valuable for simulating EV behavior and analyzing mixed autonomy traffic dynamics in EV integrated environments.

Paper Structure

This paper contains 18 sections, 8 equations, 8 figures, 1 table.

Table of Contents

  1. Introduction
  2. Dataset
  3. Classicial Models
  4. Intelligent Driver Model (IDM)
  5. Optimal Velocity Model (OVM)
  6. Optimal Velocity Relative Velocity Model (OVRV)
  7. Cooperative Adaptive Cruise Control (CACC)
  8. Model Calibration
  9. Calibration Objective
  10. Optimization Strategy
  11. Machine Learning Models
  12. Acceleration Prediction Model
  13. Space Prediction Model
  14. Evaluation and Comparison
  15. Residual Analysis of Machine Learning Models
  16. ...and 3 more sections

Figures (8)

  • Figure 1: Medium Gap
  • Figure 2: Long Gap
  • Figure 3: Extra-Long Gap
  • Figure 5:
  • Figure 6:
  • ...and 3 more figures