Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Parking of Connected Automated Vehicles: Vehicle Control, Parking Assignment, and Multi-agent Simulation

Xu Shen, Yongkeun Choi, Alex Wong, Francesco Borrelli, Scott Moura, Soomin Woo

TL;DR

The paper tackles parking efficiency for connected automated vehicle fleets by developing a hierarchical, multi-level framework that spans low-level vehicle control, mid-level collision avoidance, and high-level fleet-wide spot assignment. It introduces a comprehensive multi-vehicle simulator that integrates A*-based cruising, eight parking maneuvers via kinematic modeling, Stanley and MPC path-following, and a decentralized collision-avoidance scheme, all driven by real-world data from the Dragon Lake Parking dataset. A set of parking-spot assignment strategies, including a data-driven neural-network approach trained on DLP data, are evaluated under empty and pre-occupied lot scenarios, showing that automation can reduce total parking time by up to 43.8% compared to human baselines. The work provides a reusable simulation framework and concrete benchmarks for fleet parking, with implications for real-world deployment, mixed autonomy environments, and potential integration with charging infrastructure and energy management.

Abstract

This paper introduces a comprehensive approach to optimize parking efficiency for connected and Automated vehicle (CAVs) fleets. We present a multi-vehicle parking simulator, equipped with hierarchical path planning and collision avoidance capabilities for individual CAVs. The simulator is designed to capture the key decision-making processes in parking, from low-level vehicle control to high-level parking assignment, and it enables the effective assessment of parking strategies for large fleets of ground vehicles. We formulate and compare different strategic parking spot assignments to minimize a collective cost. While the proposed framework is designed to optimize various objective functions, we choose the total parking time for the experiment, as it directly reflects the congestion level and the cost associated with the parking efficiency. We validate the effectiveness of the proposed strategies through an empirical evaluation against a dataset of real-world parking lot dynamics, realizing a substantial reduction in parking time by up to 43.8%. This improvement is attributed to the synergistic benefits of driving automation, the utilization of shared infrastructure state data, the exclusion of pedestrian traffic, and the real-time computation of optimal parking spot allocation.

Parking of Connected Automated Vehicles: Vehicle Control, Parking Assignment, and Multi-agent Simulation

TL;DR

The paper tackles parking efficiency for connected automated vehicle fleets by developing a hierarchical, multi-level framework that spans low-level vehicle control, mid-level collision avoidance, and high-level fleet-wide spot assignment. It introduces a comprehensive multi-vehicle simulator that integrates A*-based cruising, eight parking maneuvers via kinematic modeling, Stanley and MPC path-following, and a decentralized collision-avoidance scheme, all driven by real-world data from the Dragon Lake Parking dataset. A set of parking-spot assignment strategies, including a data-driven neural-network approach trained on DLP data, are evaluated under empty and pre-occupied lot scenarios, showing that automation can reduce total parking time by up to 43.8% compared to human baselines. The work provides a reusable simulation framework and concrete benchmarks for fleet parking, with implications for real-world deployment, mixed autonomy environments, and potential integration with charging infrastructure and energy management.

Abstract

This paper introduces a comprehensive approach to optimize parking efficiency for connected and Automated vehicle (CAVs) fleets. We present a multi-vehicle parking simulator, equipped with hierarchical path planning and collision avoidance capabilities for individual CAVs. The simulator is designed to capture the key decision-making processes in parking, from low-level vehicle control to high-level parking assignment, and it enables the effective assessment of parking strategies for large fleets of ground vehicles. We formulate and compare different strategic parking spot assignments to minimize a collective cost. While the proposed framework is designed to optimize various objective functions, we choose the total parking time for the experiment, as it directly reflects the congestion level and the cost associated with the parking efficiency. We validate the effectiveness of the proposed strategies through an empirical evaluation against a dataset of real-world parking lot dynamics, realizing a substantial reduction in parking time by up to 43.8%. This improvement is attributed to the synergistic benefits of driving automation, the utilization of shared infrastructure state data, the exclusion of pedestrian traffic, and the real-time computation of optimal parking spot allocation.
Paper Structure (36 sections, 10 equations, 16 figures, 2 tables, 3 algorithms)

This paper contains 36 sections, 10 equations, 16 figures, 2 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Background
  3. Contribution
  4. Research Overview
  5. Methodology
  6. Problem Definition
  7. Assumptions
  8. Data
  9. Decentralized Planning and Control Design
  10. Path Planning
  11. Cruising
  12. Maneuvering
  13. Path Following
  14. Stanley Controller
  15. Model Predictive Control
  16. ...and 21 more sections

Figures (16)

  • Figure 1: Overview of the Simulator Framework and Control Strategies: The numbers indicate the sections in this manuscript that elaborate the design.
  • Figure 2: Dragon Lake Parking (DLP) Dataset
  • Figure 3: Path planning for cruising. In the driving lanes, the square markers indicate the starting location and the cross markers indicate the end location.
  • Figure 4: Parking maneuver trajectories
  • Figure 5: A sample trajectory of "left-north-up" maneuver and corresponding state-input profiles.
  • ...and 11 more figures