Power-Steering Control Architecture for Automatic Driving

Naranjo, José Eugenio; González, Carlos; Ricardo Garcı́a; dePedro, T.; Haber, Rodolfo E. · 2005 · OpenAlex-citations

DOI: 10.1109/tits.2005.858622

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper presents a two-layer control architecture for the automatic lateral guidance (steering) of mass-produced vehicles, addressing the challenge of unmanned steering in intelligent transportation systems. The authors argue that while longitudinal control is mature, full automatic steering remains a significant research challenge. The proposed system aims to mimic human driving behavior, offering a robust alternative to complex dynamic modeling by combining fuzzy logic with classical control theory. The experimental setup utilizes a Citroën Berlingo van equipped with real-time kinematic differential global positioning system (RTK-DGPS) sensors, which provide centimeter-level positioning accuracy at a 10 Hz rate. The control architecture employs a cascade-control paradigm. The outer loop consists of a fuzzy logic controller that calculates the target steering-wheel position based on lateral and angular errors relative to a reference GPS trajectory. This controller features two operating modes—straight-road and curve-driving—with asymmetric membership functions designed to reflect human driving intuition, such as sharper reactivity for straight paths and delayed responses for right-hand turns to avoid obstacles. The inner loop utilizes a classical proportional-integral-derivative (PID) controller implemented on hardware to drive a DC motor attached to the steering column, tracking the target position at 100 Hz. The PID parameters were tuned to ensure an overdamped response, minimizing overshoot and oscillations. Experimental results on a private circuit emulating an urban environment demonstrate that the system achieves trajectory tracking comparable to human drivers. The vehicle successfully navigated straight segments and 90-degree crossroads, maintaining lane discipline. Quantitative analysis showed average angular and lateral errors of 0.8 degrees and 0.1 meters, respectively, with maximum deviations of 3.6 degrees and 0.4 meters. The fuzzy controller effectively managed the transition between straight and curved paths, while the PID inner loop accurately executed the steering commands with minimal tracking error. The study concludes that the combination of GPS-based positioning and artificial intelligence techniques, specifically the cascade of fuzzy and PID controllers, provides an effective and reliable method for automatic steering. This approach offers a practical solution for lateral control that does not require detailed vehicle dynamic models, thereby facilitating the development of autonomous driving systems for standard vehicles. The work highlights the potential of fuzzy logic to integrate qualitative human reasoning with quantitative control requirements.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• steering pattern
• control interfaces

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model