Scenario-based stochastic MPC for vehicle speed control considering the interaction with pedestrians
Anh Tran, Arun Muraleedharan, Hiroyuki Okuda, Tatsuya Suzuki
Abstract
A typical driver spends a lot of the driving time on roads shared with pedestrians and bicyclists. Unlike highway driving, when there are pedestrians and cyclists using the same space as cars, controlling the car is more complicated. This is due to the fact that the behaviors of such agents does not follow strict rules like the cars in a closed highway. Their trajectories can be expressed better with multiple probabilistic functions than deterministic ones. We suggest a scenario-based stochastic model predictive control (MPC) framework to handle this. We consider multiple pedestrian trajectories with their respective probabilities according to an Interacting Multiple-Model Kalman Filter (IMM-KF). The car dynamics and non linear constraints are considered to avoid collision. A sample-based method is used to solve this optimization problem. The control situation was simulated using MATLAB. The proposed controller is observed to give a very natural control behavior for shared road driving compared to a deterministic single scenario MPC.