Dynamic Space Allocation Based on Internal Demand for Optimizing Release of Shared Parking

Area of Science:

• Urban planning and intelligent transportation systems.
• Artificial intelligence and machine learning applications.
• Sustainable urban development and resource management.

Background:

• Rapid urbanization leads to increased vehicle numbers, causing traffic congestion, accidents, and pollution.
• Existing transportation infrastructure expansion is insufficient to solve urban mobility challenges.
• Parking scarcity is a significant issue in densely populated urban areas.

Purpose of the Study:

• To propose and evaluate a shared parking system for urban environments.
• To enable the public use of 'open-to-specific-groups' parking lots without compromising designated user access.
• To generate potential new income streams for parking management and owners.

Main Methods:

• Utilized an Internet of Things (IoT) system for data collection and preprocessing.
• Employed recurrent neural network (RNN) algorithms, specifically many-to-many Long Short-Term Memory (LSTM), for demand prediction.
• Evaluated the system using artificial and real-world datasets and simulated various parking release scenarios.

Main Results:

• The many-to-many LSTM RNN model demonstrated the highest prediction accuracy among tested algorithms.
• Simulations confirmed the shared parking system effectively provides public parking access without overriding personal usage.
• The system successfully balances public access with the needs of designated users.

Conclusions:

• The proposed shared parking system offers a viable solution to urban parking challenges.
• AI-driven demand prediction is crucial for optimizing shared parking resources.
• This system can enhance urban mobility, reduce congestion, and create economic benefits.