PVE-MCC_for_unsignalized_intersection

Aiming at the problem of the traffic efficiency of intelligent networked vehicles passing through unsignalized-intersection in the future smart cities, this paper proposed a Progressive Value-expectation Estimation Multi-agent Cooperative Control (PVE-MCC) algorithm based on reinforcement learning. The algorithm takes the intelligent networked vehicles as the research object and designed the reward function for the optimization objective from the three aspects of traffic efficiency, safety, and comfort.

Prerequisites

• Linux or macOS
• Python 3
• matlab 2017b
• CPU or NVIDIA GPU + CUDA CuDNN

python modules

• numpy==1.16.2
• opencv-contrib-python==3.4.2.16
• opencv-python==4.2.0.32
• tensorflow==1.12.0
• matplotlib==3.0.2
• scipy==1.2.1

Getting Started

Installation

• Clone this repo:

git clone [email protected]:Mingtzge/PVE-MCC_for_unsignalized_intersection.git
cd PVE-MCC_for_unsignalized_intersection

Test the pre-trained model

python main.py --exp_name baseline --mat_path arvTimeNewVeh_new_1000_12.mat  --type test  --visible --video_name test

train/test (on the "main" branch)

• Train a model:

python main.py --mat_path arvTimeNewVeh_for_train.mat --type train --exp_name train_demo

To see more intermediate results, run

tensorboard --logdir ./model_data/train_demo/log

• Test the model:

python main.py --exp_name baseline --mat_path arvTimeNewVeh_new_1000_12.mat  --type train_demo  --visible --video_name test```
Note:the visual prarameters "--visible" and "--video_name" is optional. If use the "--visible", there will be a simulation interface to show the running interface of the vehicle in real time. the "--video_name test" is used to generate a video ,named "test.avi", saved in "./result_imgs/".