tisljaricleo / fuzzy-highway-bottleneck-python

Python-based code for estimation of highway bottleneck probability using speed transition matrices.

fuzzy-highway-bottleneck-python 🚛

Python-based code for estimation of highway bottleneck probability using speed transition matrices.
Full paper with detailed explanation will be available soon!

Info

Bottlenecks on the highways are caused by two types of congestion: 1) recurrent ones, which happens in the rush-hours due to a large number of commuters, and 2) non-recurrent ones that happen after some unexpected event like a traffic accident or sudden breaks. Bottlenecks are formally defined as a decrease in the capacity of the highway due to a congestion event.
This code implements the fuzzy method for estimation of the highway bottleneck probability occurrence using a novel traffic data model called Speed Transition Matrix (STM). The method is based on a Python package for fuzzy inference systems simpful and implemented class for STM computation.
The data is proveded in the file TrafData.pkl. It contains the routes of the vehicles on a highway simulated in the SUMO simulator.

Requirements

1. Install Python (3.8 recommended) Download link.
2. Install required packages from requirenments.txt using virtual environment.

More info

Medium article https://towardsdatascience.com/using-fuzzy-logic-for-road-traffic-congestion-index-estimation-b649f8ddede1.

Cite as

Text:
Tišljarić, Leo, Filip Vrbanić, Edouard Ivanjko, and Tonči Carić. 2022. "Motorway Bottleneck Probability Estimation in Connected Vehicles Environment Using Speed Transition Matrices" Sensors 22, no. 7: 2807. https://doi.org/10.3390/s22072807

.bib:
@Article{s22072807, AUTHOR = {Tišljarić, Leo and Vrbanić, Filip and Ivanjko, Edouard and Carić, Tonči}, TITLE = {Motorway Bottleneck Probability Estimation in Connected Vehicles Environment Using Speed Transition Matrices}, JOURNAL = {Sensors}, VOLUME = {22}, YEAR = {2022}, NUMBER = {7}, ARTICLE-NUMBER = {2807}, URL = {https://www.mdpi.com/1424-8220/22/7/2807}, PubMedID = {35408421}, ISSN = {1424-8220}, DOI = {10.3390/s22072807} }

Contact and connect

Leo Tisljaric

fuzzy-highway-bottleneck-python's Issues

Plot results

• evaluation results

Create plots for evaluation

Create plots in grount_truth package:

• exact speed
• binary speed (if critical) [0/1]
• exact density
• binary density
• bottleneck evaluated (using speed and density) [0/1]
  Bottleneck is identified if speed \le than critical speed and density \ge than critical density.

Dynamic links lengths

Add functionality for computing bottleneck prob. index for different lengths of the highway links.

Code optimization

Add constants to config.py file

Add all constraints from main files regarding:

• critical density
• critical speed
• ...

Add new data

Add highway traffic dataset divided into 5 min intervals.

Transitions cleaning

Transitions dictionary in FuzzySTM.py:

• remove transitions with a difference of 100 (vehicle skipped one segment)
• change dict keys from "from_id" + "to" + "dest_id" into (from_id, to_id)
• sort by from_id

Combine Fuzzy method with evaluation

• move ground_truth/main.py to ground_truth/init to make it importable