Scalable Multi-modal Model Predictive Control via Duality-based Interaction Predictions

2024-02-02

2024 IEEE Intelligent Vehicles Symposium (IV)

AI Summary

To address the high computational burden and redundant constraints in model predictive control (MPC) arising from multi-vehicle interaction modeling in complex multimodal traffic scenarios, this paper proposes a hierarchical interaction-aware MPC framework. Methodologically: (1) we design RAID-Net—a neural architecture integrating attention mechanisms with RNNs—to learn critical interaction patterns via Lagrangian duality theory; (2) we formulate a constraint-reduced stochastic MPC, pioneering the incorporation of dual learning into multi-agent interaction modeling to dynamically prune irrelevant collision-avoidance constraints. Our key contribution lies in achieving interaction-driven constraint compression and efficient cooperative optimization. Evaluated in high-dynamic intersection simulations, the proposed method accelerates motion planning by 12× compared to conventional multimodal MPC approaches, enabling real-time closed-loop control while significantly improving prediction accuracy and computational efficiency.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

We propose a hierarchical architecture designed for scalable real-time Model Predictive Control (MPC) in complex, multi-modal traffic scenarios. This architecture comprises two key components: 1) RAID-Net, a novel attention-based Recurrent Neural Network that predicts relevant interactions along the MPC prediction horizon between the autonomous vehicle and the surrounding vehicles using Lagrangian duality, and 2) a reduced Stochastic MPC problem that eliminates irrelevant collision avoidance constraints, enhancing computational efficiency. Our approach is demonstrated in a simulated traffic intersection with interactive surrounding vehicles, showcasing a 12x speed-up in solving the motion planning problem. A video demonstrating the proposed architecture in multiple complex traffic scenarios can be found here: https://youtu.be/-pRiOnPb9_c. GitHub: https://github.com/MPC-Berkeley/hmpc_raidnet

Problem

• Enables scalable MPC in multi-modal traffic scenarios
• Predicts autonomous vehicle interactions using duality-based methods
• Reduces computational cost by eliminating irrelevant constraints

Innovation

• Hierarchical architecture for scalable real-time MPC
• RAID-Net predicts interactions using Lagrangian duality
• Reduced Stochastic MPC enhances computational efficiency