[Paper Review] Decentralised Cooperative Collision Avoidance with Reference-Free Model Predictive Control and Desired Versus Planned Trajectories

C. E. Wartnaby, Daniele Bellan|arXiv (Cornell University)|Apr 15, 2019

Traffic control and management4 citations

TL;DR

This paper proposes a decentralized cooperative collision avoidance framework using reference-free model predictive control (MPC) with desired versus planned trajectories. Vehicles compute individual desired trajectories based on non-cooperating agents and adjust to avoid others' desired paths, enabling leaderless, robust cooperation without explicit coordination, achieving effective collision avoidance in diverse scenarios with linear computational cost scaling.

ABSTRACT

Connected and automated vehicles provide a new opportunity for highly advanced collision avoidance, in which several cars cooperate to reach an optimal overall outcome, that no single car acting in isolation could achieve. For example, one car may automatically swerve to allow another to avoid an obstacle. However, this requires solving the challenging problem of deciding what joint trajectories an ad-hoc group of cooperating vehicles should follow, with no obvious leader known in advance. To avoid the complexities of agreeing what plan to follow in an ever-evolving situation, a protocol requiring no leader and no explicit inter-vehicle agreement is desirable, which nevertheless yields cooperative, robust behaviour. One method is demonstrated here, in simulation. This uses the notion of "desired" versus "planned" trajectories, allowing vehicles to influence each other for mutual benefit, without requiring a leader or explicit agreement protocol. Essentially the desired trajectory is that which the vehicle would choose if other cooperating vehicles were not present, avoiding the predicted paths of non-cooperating actors. The planned trajectory additionally accounts for the planned trajectories of other cooperating vehicles, giving the safest currently available path. Both trajectories are broadcast. As each vehicle attempts to (weakly) avoid the desired trajectories of other vehicles, cooperative behaviour emerges. A simple form of model predictive control is used. The cost function penalises predicted collisions, accounting for severity. There is a weak preference for maintaining the current road lane, but no explicit reference trajectory. This decentralised planning and simple optimisation scheme results in effective handling of a wide range of collision scenarios, with no hard limit to the number of cooperating vehicles. The computing cost is linear in the number of vehicles.

Motivation & Objective

• To enable cooperative collision avoidance among connected and automated vehicles without requiring a leader or explicit inter-vehicle agreement.
• To address the challenge of decentralized trajectory planning in dynamic, ad-hoc groups of vehicles with no central coordination.
• To develop a scalable solution that maintains safety and robustness while minimizing computational overhead.
• To enable vehicles to influence each other’s behavior through trajectory broadcasting, promoting mutual safety without explicit communication of plans.

Proposed method

• Each vehicle computes a 'desired' trajectory based on avoiding predicted paths of non-cooperating agents, using a reference-free MPC formulation.
• A 'planned' trajectory is generated by incorporating avoidance of other vehicles' desired trajectories, ensuring cooperative safety.
• Both desired and planned trajectories are broadcast to neighboring vehicles for mutual awareness.
• The MPC cost function penalizes predicted collisions with severity weighting, favoring safer path choices.
• A weak preference for lane-keeping is included, but no explicit reference trajectory is used.
• The scheme operates in a decentralized manner, with computation cost scaling linearly with the number of vehicles.

Experimental results

Research questions

• RQ1 How can cooperative collision avoidance be achieved in a decentralized manner without a designated leader?
• RQ2 What role do desired versus planned trajectories play in enabling implicit cooperation without explicit coordination?
• RQ3 Can a reference-free MPC approach maintain safety and scalability in dynamic multi-vehicle scenarios?
• RQ4 How does the absence of a reference trajectory affect the robustness and performance of collision avoidance?
• RQ5 What is the computational scalability of the proposed method with increasing numbers of cooperating vehicles?

Key findings

• The proposed method achieves effective collision avoidance across a wide range of complex scenarios without requiring a leader or explicit coordination.
• Cooperative behavior emerges naturally through mutual avoidance of other vehicles' desired trajectories, without direct plan exchange.
• The system maintains robust performance with linear computational cost scaling relative to the number of vehicles.
• The absence of a reference trajectory simplifies the control design while preserving safety and stability.
• Simulation results demonstrate the method's ability to handle dynamic, unpredictable environments with multiple interacting agents.