MADM-based smart parking guidance algorithm

Area of Science:

• Intelligent Transportation Systems
• Operations Research

Background:

• Smart parking systems face challenges in guiding drivers to optimal facilities based on multiple criteria.
• Driver preferences for parking vary, including walk duration, cost, and ease of finding a space.

Purpose of the Study:

• To develop and evaluate a smart parking guidance algorithm using Multiple Attribute Decision Making (MADM).
• To incorporate driver preferences and parking facility attributes into a decision-making framework.

Main Methods:

• Utilized MADM theory to create a parking guidance algorithm considering walk duration, parking fee, and vacant spaces.
• Employed queueing theory to estimate the expected number of vacant parking spaces.
• Proposed a novel Markov Chain-based attribute for parking availability.

Main Results:

• The MADM-based algorithm effectively guides drivers to suitable parking resources aligned with their preferences.
• The proposed Markov Chain-based availability attribute is superior to arrival rate-based methods for representing parking space availability.
• The algorithm demonstrated superior performance compared to a blind search approach in various scenarios.

Conclusions:

• The developed MADM-based smart parking guidance algorithm is effective for optimizing parking choices.
• The novel Markov Chain-based availability metric enhances the accuracy of parking availability assessment in smart parking systems.