This paper presents an artificial neural network (ANN)-based nonlinear model predictive visual servoing method for mobile robots. The ANN model is developed for state predictions to mitigate the unknown dynamics and parameter uncertainty issues of the physics-based (PB) model. To enhance both the model generalization and accuracy for control, a two-stage ANN training process is proposed. In a pretraining stage, highly diversified data accommodating broad operating ranges is generated by a PB kinematics model and used to train an ANN model first. In the second stage, the test data collected from the actual system, which is limited in both the diversity and the volume, are employed to further finetune the ANN weights. Path-following experiments are conducted to compare the effects of various ANN models on nonlinear model predictive control and visual servoing performance. The results confirm that the pretraining stage is necessary for improving model generalization. Without pretraining (i.e., model trained only with the test data), the robot fails to follow the entire track. Weight finetuning with the captured data further improves the tracking accuracy by 0.07–0.15 cm on average.

Inland waterway transportation is one of the most important means to transport cargo in rivers and canals. To facilitate autonomous navigation for ships in inland waterways, this paper proposes a data-driven approach for predictions and control of underactuated ships with unknown dynamics, which integrates model predictive control (MPC) with an iterative learning control (ILC) scheme. In each iteration, kernel-based linear regressors are used to identify the relations between the evolution of ship states and control inputs based on the stored data from previous iterations and the collected data during operation, so as to build the system prediction model. The data are dynamically used to fix the prediction model over iterations, as well as to improve the controller performance until it converges. The proposed approach does not require prior knowledge regarding the hydrodynamic coefficients and ship parameters, but learns from the data instead. In addition, it exploits the advantages of MPC in handling constraints with minimised overall cost. Simulation results show that the controller could start from a nominal, linear data-driven ship model and then learn to reduce the path-following errors based on the data obtained over iterations.

This paper introduces a scheme for waypoint-based path-following control for an Unmanned Robot Sailboat (URS) in the presence of actuator gain uncertainty and unknown environment disturbances. The proposed scheme has two components: intelligent guidance and an adaptive neural controller. Considering upwind and downwind navigation, an improved version of the integral Line-Of-Sight (LOS) guidance principle is developed to generate the appropriate heading reference for a URS. Associated with the integral LOS guidance law, a robust adaptive algorithm is proposed for a URS using Radial Basic Function Neural Networks (RBF-NNs) and a robust neural damping technique. In order to achieve a robust neural damping technique, one single adaptive parameter must be updated online to stabilise the effect of the gain uncertainty and the external disturbance. To ensure Semi-Global Uniform Ultimate Bounded (SGUUB) stability, the Lyapunov theory has been employed. Two simulated experiments have been conducted to illustrate that the control effects can achieve a satisfactory performance.

This paper presents a new method to design global path-following controllers for underactuated ships under both deterministic and stochastic sea loads. The path-following errors are first interpreted in a moving frame attached to the path. These errors are then to be stabilized at the origin by a design of controllers based on backstepping and Lyapunov's direct methods. Weak and strong nonlinear Lyapunov functions are introduced to overcome difficulties caused by underactuation and Hessian terms induced by stochastic differentiation rule, and to guarantee boundedness of the sway velocity. Potential projection functions are introduced to design update laws that provide bounded estimates of the mean values and covariances of the disturbances. Simulations are included to illustrate the effectiveness of the proposed approach.

This paper proposes a new kinematic controller for the waypoints guidance of robotic mobile platforms. A notable feature of the controller is its ability to process the raw sequence of waypoints to produce smooth reference velocities from control laws that are derived by taking into account a driving profile including the velocity limits, the acceleration limits, the motion modes through each waypoint (forward or backward) and the precision constraints that are required to ensure accurate waypoints traversal. A mathematical analysis demonstrates the convergence of the movements through the waypoints sequence. In addition, we present a simple way to adapt the driving profile in order that the platform reaches the last waypoint at a prescribed time. A feed-forward unit is finally described, that compensates for delays and first-order poles in the velocity response of the platform. Various simulations and experiments on real robotic platforms demonstrate the behavior and the effectiveness of the solution.

In the article an optimal control problem subject to a stationary variational inequalityis investigated. The optimal control problem is complemented with pointwise controlconstraints. The convergence of a smoothing scheme is analyzed. There, the variationalinequality is replaced by a semilinear elliptic equation. It is shown that solutions ofthe regularized optimal control problem converge to solutions of the original one. Passingto the limit in the optimality system of the regularized problem allows to proveC-stationarity of local solutions of the original problem. Moreover, convergence rateswith respect to the regularization parameter for the error in the control are obtained,which turn out to be sharp. These rates coincide with rates obtained by numericalexperiments, which are included in the paper.