The goals of the research are to design vehicle control for achieving autonomous functionalities and to examine their implementation possibilities on small-scaled indoor vehicles and on real test vehicles.
The scientific background of the research is based on the robust and nonlinear control theory, which is used to design architectures with provable performance guarantees on the controlled system. In the design process learning-based methods are incorporated, which requires analysis on the stability and on the performance of the controlled system, especially on the safety-critic vehicle control applications. The effectiveness of the resulted methods are validated through real-life vehicle control scenarios.
The development and analysis of machine learning algorithms and optimization methods is a new direction in the planning of autonomous vehicle control. In addition, providing stability and performance guarantees for complex management architectures is a significant challenge. Research on validation and safety assurance for autonomous vehicles is one of the most challenging areas of research. This includes the generation of situations in an automated way, as the resolution of critical situations and emergencies is the basis for safe traffic. We carry out further research in the selection and analysis of suitable formal performance specifications that meet robustness requirements.
Design of controls and preparation of demonstrations related to certain functions of autonomous vehicles:
Analysis of machine learning algorithms suitable for vehicle control tasks in terms of stability and quality characteristics. Research on linearized as well as variable parameter representations of learning-based controls.
Control design of autonomous vehicle subsystems based on robust, Linear Variable Parameters (LPV) and predictive optimal methods taking into account the analysis results of machine learning algorithms.
Research on modeling and control design tasks related to autonomous vehicle maneuvers: data-based modeling, maneuvering based on wheel odometry, design of maneuvers based on motion estimation of man-made vehicles.
Some of the results (connected to a variable geometry chassis test bench, robotic vehicle demonstrations) will be demonstrated in the Indoor Demonstration Plaftorm block. Another part of the results will be implemented in the Experimental Vehicle Platform subproject (experiments related to odometry).
Németh B.: Coordination of Lateral Vehicle Control Systems using Learning-based Strategies. Energies. 2021. 14(5), 1291. (Q2, IF: 2,702)
Fényes D., Fazekas M., Németh B., Gáspár P.: Implementation of a variable-geometry suspension-based steering control system. Vehicle System Dynamics. 2021. (D1, IF: 2,581)
Fényes D., Németh B., Gáspár P.: A novel data-driven modeling and control design method for autonomous vehicles. Energies. 14(2). 517. 2021. (Q2, IF: 2,702)
Fényes D., Németh B., Gáspár P.: Design of LPV control for autonomous vehicles using the contributions of big data analysis. International Journal of Control. 2021. (D1, IF 2,78)
Fazekas M., Gáspár P., Németh B.: Calibration and Improvement of an Odometry Model with Dynamic Wheel and Lateral Dynamics Integration. Sensors. 21(2), 337. 2021. (Q1, IF 3,275)
Hegedűs T., Németh B., Gáspár P.: Design of low-complexity graph-based motion planning algorithm for autonomous vehicles. Applied Sciences. Vol 10., No. 21. 2020. (Q1, IF 2,474)