Future-Focused Control Barrier Functions for Autonomous Vehicle Control
Mitchell Black, Mrdjan Jankovic, Abhishek Sharma, and Dimitra Panagou
Submitted to 2023 American Control Conference (under review)
In this paper, we introduce a class of future-focused control barrier functions (ff-CBF) aimed at improving traditionally myopic CBF based control design and study their efficacy in the context of an unsignaled four-way intersection crossing problem for collections of both communicating and non-communicating autonomous vehicles. Our novel ff-CBF encodes that vehicles take control actions that avoid collisions predicted under a zero-acceleration policy over an arbitrarily long future time interval. In this sense the ff-CBF defines a virtual barrier, a loosening of which we propose in the form of a relaxed future-focused CBF (rff-CBF) that allows a relaxation of the virtual ff-CBF barrier far from the physical barrier between vehicles. We study the performance of ff-CBF and rff-CBF based controllers on communicating vehicles via a series of simulated trials of the intersection scenario, and in particular highlight how the rff-CBF based controller empirically outperforms a benchmark controller from the literature by improving intersection throughput while preserving safety and feasibility. Finally, we demonstrate our proposed ff-CBF control law on an intersection scenario in the laboratory environment with a collection of 5 non-communicating AION ground rovers.
A collection of five AION R1 UGV ground rovers are controlled with a decentralized future focused control barrier function quadratic program based control law in a laboratory intersection scenario. The full control loop ran at a frequency of 20Hz, where position feedback was obtained using a Vicon motion capture system, and the extended Kalman filter output from the PX4 firmware running via the on-board Pixhawk was used for state estimation.
A relaxed version of the future focused control barrier function QP based controller is used to solve the intersection crossing problem with high empirical success.
A centralized, exponential CBF-QP controller preserves safety amongst the 4 vehicles but leads to a deadlock scenario due to the nominal, exponential CBF lacking predictive power.
Visualization of the effect of the ff-CBF. Whereas the nominal CBF is evaluated based on the locations of vehicles 1 (maize) and 2 (blue) at time t, i.e. (a) and (b), the ff-CBF evaluates safety based on the predicted future locations of the vehicles at time t + tau, i.e. (c) and (d), allowing the present control to take action to avoid predicted future danger.
The rv-CBF control scheme achieved 100% feasibility even in the turning scenario, despite the constant velocity prediction model not taking a change of heading into account. We leave any theoretical guarantees of feasibility, however, to future work.
