Hybrid control for low-regular nonlinear systems: application to an embedded control for an electric vehicle

TL;DR

This paper develops a hybrid control strategy for low-regular nonlinear systems, specifically applied to embedded control of an electric vehicle, demonstrating efficiency, robustness, and low power consumption in real-world scenarios.

Contribution

It introduces a simple, optimal, and robust hybrid control approach tailored for low-regular nonlinear systems like electric vehicles with on/off engines.

Findings

Automatic velocity control matches trained human driver performance.

Power consumption of the control device is under 10 mW.

Control strategy enhances robustness and safety in vehicle operation.

Abstract

This note presents an embedded automatic control strategy for a low consumption vehicle equipped with an "on/off" engine. The main difficulties are the hybrid nature of the dynamics, the non smoothness of the dynamics of each mode, the uncertain environment, the fast changing dynamics, and low cost/ low consumption constraints for the control device. Human drivers of such vehicles frequently use an oscillating strategy, letting the velocity evolve between fixed lower and upper bounds. We present a general justification of this very simple and efficient strategy, that happens to be optimal for autonomous dynamics, robust and easily adaptable for real-time control strategy. Effective implementation in a competition prototype involved in low-consumption races shows that automatic velocity control achieves performances comparable with the results of trained human drivers. Major advantages of automatic control are improved robustness and safety. The total average power consumption for the control device is less than 10 mW.