Robust adaptive regulation of dynamically positioned ships with unknown dynamics and unknown disturbances

Abstract

We develop a robust adaptive regulating control law for dynamically positioned ships subject to unknown dynamics and bounded unknown disturbances incorporating the radial basis function (RBF) neural network (NN), the dead zone adaptive technique, and a robust control term into the vectorial backstepping approach. The RBF NNs with the dead zone adaptive laws approximate the ship unknown dynamics. The adaptive law-based robust control term compensates for unknown disturbances, NN approximation errors, and undesirable errors arising from the design procedures. The developed dynamic positioning (DP) control law regulates the ship position and heading to the desired values with arbitrarily small errors, while guaranteeing the uniform ultimate boundedness of all signals in the DP closed-loop control system of ships. High-fidelity simulations on two supply ships and comparisons demonstrate the effectiveness and the superiority of the developed DP control law. KEYWORDS robust adaptive control, ship dynamic positioning, unknown dynamics, unknown time-varying disturbances, vectorial backstepping Int J Adapt Control Signal Process. 2019;33:545-556.wileyonlinelibrary.com/journal/acs

Disturbance observer‐based finite‐time control scheme for dynamic positioning of ships subject to thruster faults

“…Thus, DP technology is crucial in shuttle tankers, oil drilling platforms, and pipe‐laying ships, especially in the deep seas 4 . PID, 1,5 LQG, 6 and intelligent control 7‐9 have been applied to implement DP in ships in the past few decades. The nonlinear control methodology has recently become popular owing to its remarkably high control precision and robustness to model uncertainties or environmental disturbances 9‐11 .…”

“…PID, 1,5 LQG, 6 and intelligent control 7‐9 have been applied to implement DP in ships in the past few decades. The nonlinear control methodology has recently become popular owing to its remarkably high control precision and robustness to model uncertainties or environmental disturbances 9‐11 . Sliding mode control (SMC) was first adopted in the DP system of a turret‐moored floating production storage and offloading vessel in 2001.…”

“…Researchers have used vectorial backstepping to design DP control schemes and achieved global exponential stability 17 . However, this technique leads to the “explosion of complexity,” which is addressed using dynamic surface control 8 and neural networks 9 . In addition to the abovementioned methodologies, model predictive control, which is an optimization‐based time‐domain control technique, can be used to handle system constraints with relatively low model accuracy 18‐20 .…”

“…However, compensating for the disturbances in the design of the control scheme is a challenge. Intelligent methods such as neural networks 9,30 and least squares support vector machines (LS‐SVM) 28 are used to estimate environmental disturbances. However, the introduction of neural networks or LS‐SVM will increase the number of online learning weights.…”

Robust trajectory tracking control of a marine surface vessel using asymmetric error constraints and output feedback

“…In References 25,26, an adaptive feed‐forward NN is used to estimate uncertainties due to disturbances and model errors in the cooperative control of an MSV. The adaptive law based on the radial basis function (RBF) NN with a deal zone is developed to approximate MSV unknown dynamics in Reference 27. With good approximation and learning performances, NNs have become powerful tools to predict the uncertainties of nonlinear systems models.…”