Automatic Parameter Tuning of Self-Driving Vehicles

TL;DR

This paper introduces an automatic parameter tuning method for self-driving vehicles that uses cost functions and local optimization to improve trajectory planning based on expert demonstrations.

Contribution

It presents a novel approach combining cost functions and local optimization to automatically tune parameters in autonomous driving systems, outperforming manual tuning.

Findings

Automatic tuning improves initial parameters significantly.

Method handles noisy demonstration data effectively.

Case study confirms effectiveness in lane following scenario.

Abstract

Modern automated driving solutions utilize trajectory planning and control components with numerous parameters that need to be tuned for different driving situations and vehicle types to achieve optimal performance. This paper proposes a method to automatically tune such parameters to resemble expert demonstrations. We utilize a cost function which captures deviations of the closed-loop operation of the controller from the recorded desired driving behavior. Parameter tuning is then accomplished by using local optimization techniques. Three optimization alternatives are compared in a case study, where a trajectory planner is tuned for lane following in a real-world driving scenario. The results suggest that the proposed approach improves manually tuned initial parameters significantly even with respect to noisy demonstration data.