AMoD

---

SUMO ride-hailling parking model simulation

This repository implements a comprehensive traffic simulation system focusing on roadside parking behavior and emission analysis in Beijing. The system integrates SUMO with advanced optimization algorithms for fleet management and ride-hailing dispatch.

---

• project
  • config
  • paths
  • simulation_setup
  • od_demand_generator
  • fleet_optimizer
  • interfaces
  • dispatch_manager
  • ridehail_coordinator
  • parking_manager
  • detector
  • emissions

---

Main Modules Overview

config

Defines simulation parameters including area boundaries, vehicle densities, parking ratios, and emission factors. Supports both local and city-wide simulation modes.

paths

Manages file paths, detects SUMO installation locations, and organizes output directories. Ensures cross-platform compatibility for SUMO tools.

simulation_setup

After downloading the road network data based on OSM, convert it into a SUMO network and generate vehicle routes. Use randomTrips.py to generate initial trip data.

od_demand_generator

Implements Origin-Destination (OD) matrix-based demand generation, incorporating Beijing-specific travel patterns. Creates zone-based traffic analysis areas and generates realistic trip distributions.

fleet_optimizer

Optimizes vehicle shareability based on MIT research. Constructs shareability networks and solves minimum path cover problems to determine the optimal fleet size.

interfaces

Provides modular interfaces for demand generation, fleet optimization, dispatch optimization, and parking decisions. Implements algorithms based on the "Nature" paper for fleet optimization and real-time dispatching.

dispatch_manager

A comprehensive ride-hailing dispatch system with driver-order matching algorithms. Calculates distances using the OSMNX formula, manages driver preferences, and maintains rejection lists.

ridehail_coordinator

Coordinates order generation, vehicle dispatch, and parking decisions post-service. Integrates with the main simulation loop for real-time ride-hailing operations.

parking_manager

Estimates roadside parking density based on road types and urban characteristics. Implements parking spot distribution algorithms tailored for different road categories.

detector

Performs real-time detection and management of parking events during simulation. Handles parking position validation, traffic impact analysis, and emission calculations.

emissions

Calculates vehicle emissions based on the HBEFA4 model, supporting gasoline, diesel, and electric vehicles. Implements driving mode detection and CO₂-equivalent calculations.

traci_manager

Manages SUMO TraCI connections with automatic retry mechanisms and error handling.