This paper presents a novel approach to collision and obstacle avoidance in fixed-wing unmanned aerial systems (UASs), vehicles with high speed and high inertia, operating in proximal or congested settings. A unique reformulation of classical artificial potential field (APF) navigational approaches, adaptively morphing the functions' shape considering six-degrees-of-freedom (6DOF) dynamic characteristics and constraints of fixed-wing aircraft, is fitted to an online predictive and prioritized waypoint planning algorithm for generation of evasive paths during abrupt encounters. The time-varying waypoint horizons output from the navigation unit are integrated into a combined guidance and nonlinear model predictive control scheme. Real-time avoidance capabilities are demonstrated in full nonlinear 6DOF simulation of a large unmanned aircraft showcasing evasion efficiency with respect to classical methods and collision free operation in a congested urban scenario.

The nonlinear and unsteady nature of aircraft aerodynamics in the presence of adverse conditions and external disturbances, together with a limited range of flight variables makes the use of the linear control theory inadequate in such conditions. To address these constraints and significantly enhance aircraft control capabilities, this brief presents an adaptive framework for a robust nonlinear model predictive control (NMPC). Control algorithms are tested on a 1100-pound unmanned aerial system, with nonlinear, coupled, and unstable open-loop dynamics, subjected to environmental disturbances and measurement noise. Given the usual frequency content exclusion between disturbance and noise, this solution addresses the lack of robustness in model predictive control by the inclusion of frequency-dependent weighting matrices and a nonlinear version of the mixed sensitivity approach. Furthermore, real-time aerodynamic parameter estimation and predictive model updating is carried out by online adaptive artificial neural networks. Through assessment and validity of control algorithms, it is demonstrated that two originally competing control concepts, robustness and performance, are integrally attained in real time. This is usually unreachable in the classical NMPC framework for complex systems.

This work presents a closed-loop position control of a mobile robot, which is capable of moving from its current position to a target point by manipulating its linear and angular velocities. The main objective of this paper is to modify an existing control law based on the kinematic model to improve the response when the robot is backwards oriented and to reach the destination point in less time and with a shorter trajectory. Stability of the proposed control law is validated by Lyapunov Criterion. Some procedures are implemented to test this approach both in simulation with MATLAB, and experimentally with the Khepera IV robot.

Due to the increase in complexity in autonomous vehicles, most of the existing control systems are proving to be inadequate. Reinforcement Learning is gaining traction as it is posed to overcome these difficulties in a natural way. This approach allows an agent that interacts with the environment to get rewards for appropriate actions, learning to improve its performance continuously. The article describes the design and development of an algorithm to control the position of a wheeled mobile robot using Reinforcement Learning. One main advantage of this approach concerning traditional control algorithms is that the learning process is carried out automatically with a recursive procedure forward in time. Moreover, given the fidelity of the model for the particular implementation described in this work, the whole learning process can be carried out in simulation. This fact avoids damages to the actual robot during the learning stage. It shows that the position control of the robot (or similar specific tasks) can be done without the need to know the dynamic model of the system explicitly. Its main drawback is that the learning stage can take a long time to finish and that it depends on the complexity of the task and the availability of adequate hardware resources. This work provides a comparison between the proposed approach and traditional existing control laws in simulation and real environments. The article also discusses the main effects of using different controlled variables in the performance of the developed control law.