Imposing adaptive performance specifications for uncertain nonlinear systems with actuation and feedback limitations
Loading...
Date
2025-06-17
Authors
Τράκας, Παναγιώτης
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Control systems are fundamental in engineering, enabling automation and precision in fields such as robotics, power generation, autonomous vehicles, and aerospace. However, real-world constraints such as actuator saturation, limited feedback, and external disturbances pose significant challenges in achieving both stability and high performance. From a theoretical standpoint, such constraints complicate the design of control laws that guarantee robustness and precision, while in practice, they can lead to degraded system performance or even instability. This dissertation introduces a novel framework for Adaptive Performance Control (APC) for uncertain nonlinear systems subject to actuation and feedback limitations. Building on the principles of Prescribed Performance Control (PPC), we establish a systematic approach for enforcing adaptive performance specifications while ensuring compliance with real-world constraints, including control input saturation, rate limits, and partial state feedback.
The first part of the thesis is devoted to the development of the theoretical foundations of APC. It presents approximation-free control schemes for uncertain nonlinear systems, ensuring the boundedness of closed-loop signals while dynamically adjusting performance specifications based on actuation constraints, both in terms of amplitude and rate of change of the control signal. Sufficient conditions on the level of input constraints required to guarantee boundedness
of closed-loop signals, as well as conditions for achieving accurate tracking, are derived using Lyapunov analysis. Additionally, a novel output-feedback APC scheme is proposed to eliminate the reliance on full-state measurements, enhancing the robustness and applicability in practical scenarios. A nonlinear separation principle is exploited to demonstrate that the output-feedback controller can recover the performance characteristics of its state-feedback counterpart.
The second part of the thesis extends the APC methodology to three benchmark applications. The first application focuses on adaptive cruise control (ACC), where a robust control framework integrates APC into vehicular systems, ensuring compliance with actuation and performance constraints while guaranteeing collision avoidance. The second application addresses formation control of input-constrained multi-agent systems, introducing a decentralized APC approach for
leader-follower formations, experimentally validated using a swarm of quadcopters. The third application involves target tracking with collision avoidance, proposing an adaptive trajectory-tracking scheme for input-constrained mobile robots that dynamically modifies performance specifications to avoid obstacles.
The proposed APC methodology is rigorously validated through theoretical analysis, simulations, and real-world experiments, highlighting its ability to achieve high-precision control under operational constraints.
Description
Keywords
Automatic control, Input constraints, Partial feedback, Robotics