A Graph-Based Approach to the Topological Optimization of Cycling Networks for the Improvement of Safety and Comfort of Cyclists

This work aims to improve perceived safety and comfort of cyclists by proposing a topological optimization of existing cycling networks. We assign weights to a six-category system for bike lanes based on their segregation from motorized vehicles using the well-developed CycleRAP tool. Each bike lane is weighted based on its category and topological features. Graph theory metrics are then applied to analyze the core topological characteristics of cycling networks across various French municipalities. These metrics form the basis for estimating and predicting cyclists’ perceived safety and comfort levels, as reported in local surveys. Building on this relationship, we formulate a topology optimization problem aimed at maximizing predicted safety and comfort within budgetary constraints. To tackle this complex problem, we introduce a topological optimization algorithm and compare its performance with existing algorithms to ensure reliability. This approach integrates graph theory with real-world indicators, providing a comprehensive quantitative framework to support decision-making in urban planning and resource allocation.