Model Predictive Control with Gaussian-Process-Supported Dynamical Constraints for Autonomous Vehicles
This work addresses safety-critical control for autonomous vehicles interacting with human drivers, representing an incremental improvement over existing model predictive control methods.
The researchers tackled the problem of autonomous vehicle control in human-driven environments by developing a model predictive control approach that uses Gaussian processes to predict human driving behavior with uncertainty quantification. Their method achieved repeated feasibility and probabilistic constraint satisfaction with high probability in simulations using real-world measurement data.
We propose a model predictive control approach for autonomous vehicles that exploits learned Gaussian processes for predicting human driving behavior. The proposed approach employs the uncertainty about the GP's prediction to achieve safety. A multi-mode predictive control approach considers the possible intentions of the human drivers. While the intentions are represented by different Gaussian processes, their probabilities foreseen in the observed behaviors are determined by a suitable online classification. Intentions below a certain probability threshold are neglected to improve performance. The proposed multi-mode model predictive control approach with Gaussian process regression support enables repeated feasibility and probabilistic constraint satisfaction with high probability. The approach is underlined in simulation, considering real-world measurements for training the Gaussian processes.