Adaptive control of small outboard-powered boats for survey applications

Abstract

Four autopilot controllers have been developed in this work that can both hold a desired heading and follow a straight line. These PID, adaptive PID, neuro-adaptive, and adaptive augmenting control algorithms have all been implemented into a numerical simulation of a 33-foot center console vessel with wind, waves, and current disturbances acting in the perpendicular (across-track) direction of the boat's desired trajectory. Each controller is tested for its ability to follow a desired heading in the presence of these disturbances and then to follow a straight line at two different throttle settings for the same disturbances. These controllers were tuned for an input thrust of 2000 N and all four controllers showed good performance with none of the controllers significantly outperforming the others when holding a constant heading and following a straight line at this engine thrust. Each controller was then tested for a reduced engine thrust of 1200 N per engine where each of the three adaptive controllers reduced heading error and across-track error by approximately 50% after a 300 second tuning period when compared to the fixed gain PID, showing that significant robustness to changes in throttle setting was gained by using an adaptive algorithm.

A robust variable-structure LQI controller for under-actuated systems via flexible online adaptation of performance-index weights

“…The closed-loop stability of the proposed adaptive control law is proved using the Lyapunov function shown in (27) [24].…”

“…The LQI controller is augmented with a superior regulator that adaptively modulates the stateweighting factors as a function of state-error variations. The baseline weight-adaptation scheme used in this research is inspired by the Fisher's gain-adjustment law due to its reasonable flexibility and good tracking capability [27]. The proposed adaptation scheme uses preconfigured dissipative and anti-dissipative functions.…”

Adaptive optimal control of under-actuated robotic systems using a self-regulating nonlinear weight-adjustment scheme: Formulation and experimental verification

“…These functions ensure the smooth commutation of state weights across all operating conditions. The weight-adjusting functions, designed to address each state-weighting factor, is expressed in (29) to (32).…”

“…In this research, the variation rates are adapted online by using the online self-regulation law, prescribed by Fisher et al in [29], due to its robust and accurate tracking ability. This online adaptation mechanism employs pre-calibrated dissipative and anti-dissipative functions…”

“…The resulting selected values of the parameters are given as; a α = 1.85, a θ = 2.41, a _ a = 11.05, a_ y = 10.05, b α = 408.15, b θ = 291.62, b _ a = 10.58, b_ y = 9.92, g * a = 2.10, g * y = 5.76, g * _ a = 1.05, g * _ y = 2.94, φ α = 1.52, φ θ = 1.45, φ _ a = 1.67, and φ _ y = 1.62.Although this scheme introduces a multitude of hyper-parameters, which makes the tuning procedure, the performance-related benefits offered by the scheme surpass this drawback. The adaptive LQR augmented with the adaptation scheme presented in(29) to(32) is denoted as the Error Magnitude-driven Self Tuning Controller (or EM-STC) in this article.Fig 4 depicts the block diagram of the basic EM-STC.…”