[Paper Review] Optimal Control of Connected Automated Vehicles at Urban Traffic Intersections: A Feasibility Enforcement Analysis

Yue Zhang, Christos G. Cassandras+1 more|arXiv (Cornell University)|Aug 15, 2016

Traffic control and management15 references3 citations

TL;DR

This paper proposes a feasibility enforcement zone (FEZ) to ensure collision-free, optimal control of connected automated vehicles (CAVs) at urban intersections by pre-adjusting vehicle speed and arrival time before entering the control zone. The FEZ guarantees that each CAV enters the control zone with initial conditions within a feasible region, enabling decentralized optimal control that minimizes fuel use and maximizes throughput while enforcing hard safety constraints.

ABSTRACT

Earlier work has established a decentralized optimal control framework for coordinating online a continuous flow of connected automated vehicles (CAVs) entering a control zone and crossing two adjacent intersections in an urban area. A solution, when it exists, allows the vehicles to cross the intersections without the use of traffic lights, without creating congestion on the connecting road, and under the hard safety constraint of collision avoidance. We establish the conditions under which such solutions exist and show that they can be enforced through an appropriately designed feasibility enforcement zone that precedes the control zone. The proposed solution and overall control architecture are illustrated through simulation.

Motivation & Objective

• To address the challenge of ensuring feasible initial conditions for optimal control of CAVs entering a control zone at urban intersections.
• To identify the conditions under which optimal control solutions exist, particularly regarding speed and arrival time.
• To design a Feasibility Enforcement Zone (FEZ) that proactively adjusts CAVs' speed and entry time to meet feasibility requirements before entering the control zone.
• To enable decentralized, real-time coordination of CAVs across two adjacent intersections without traffic lights, minimizing fuel consumption and avoiding collisions.
• To establish a framework that supports scalable, safe, and efficient intersection management using only vehicle-to-vehicle and vehicle-to-infrastructure communication.

Proposed method

• The paper defines a feasible region in the speed–arrival time space where optimal control solutions exist, based on hard safety constraints including rear-end collision avoidance.
• It introduces a Feasibility Enforcement Zone (FEZ) preceding the control zone, where CAVs are controlled to reach a feasible initial state before entering the control zone.
• The FEZ design uses a feedback control law that adjusts vehicle speed based on real-time information from preceding vehicles, ensuring entry into the control zone within the feasible region.
• The method leverages a decentralized optimal control framework where each CAV solves its own optimal control problem using local information and safety constraints.
• Theoretical analysis proves that if the FEZ is properly designed, all CAVs entering the control zone will have feasible initial conditions, enabling collision-free, optimal operation.
• Simulation results validate the FEZ’s effectiveness by showing that CAVs with the FEZ maintain safe inter-vehicle distances, while those without the FEZ violate safety constraints.

Experimental results

Research questions

• RQ1 Under what conditions on speed and arrival time does a feasible solution exist for the optimal control problem of a CAV entering the control zone at an intersection?
• RQ2 How can a CAV be guaranteed to enter the control zone with initial conditions that satisfy the feasibility constraints of the optimal control problem?
• RQ3 What is the optimal design of a Feasibility Enforcement Zone (FEZ) that ensures all CAVs enter the control zone within the feasible region without centralized coordination?
• RQ4 How does the inclusion of an FEZ affect the safety, fuel efficiency, and throughput of CAVs at urban intersections compared to systems without such a zone?
• RQ5 Can the proposed framework support diverse CAV types and dynamic traffic conditions while maintaining safety and optimality?

Key findings

• The feasible region for CAV initial conditions (arrival time and speed) is fully characterized in terms of information available before entering the control zone.
• The FEZ is proven to be effective in enforcing feasible initial conditions for all CAVs entering the control zone, ensuring that optimal control solutions remain feasible.
• Simulation results show that without the FEZ, 6 out of the first 20 CAVs violate the minimum safe distance constraint (s_i(t) < 10 m), while all CAVs with the FEZ maintain safe spacing.
• The FEZ length is analytically derived as 44 meters under the given system parameters, ensuring that all CAVs can be adjusted into the feasible region.
• CAVs that enter the control zone with already feasible initial conditions (e.g., first vehicles in a lane) require no adjustment, demonstrating the efficiency of the approach.
• The framework enables decentralized, real-time coordination of CAVs across two adjacent intersections, achieving fuel-efficient, congestion-free, and collision-free operation.