Estimating the robot state within a known map is an essential problem for mobile robot; it is also referred to “localization”. Even LiDAR-based localization is practical in many applications, it is difficult to achieve global localization with LiDAR only for its low-dimension feedback, especially in environments with repetitive geometric features. A sensor-fusion-based localization system is introduced in this paper, which has the capability of addressing the global localization problem. Both LiDAR and vision sensors are integrated, making use of the rich information introduced by vision sensor and the robustness from LiDAR. A hybrid grid-map is built for global localization, and a visual global descriptor is applied to speed up the localization convergence, combined with a pose refining pipeline for improving the localization accuracy. Also, a trigger mechanism is introduced to solve kidnapped problem and verify the relocalization result. The experiments under different conditions are designed to evaluate the performance of the proposed approach, as well as a comparison with the existing localization systems. According to the experimental results, our system is able to solve the global localization problem, and the sensor-fusion mechanism in our system has an improved performance.

Humanoid robots (HR) equipped with a sophisticated facial character analysis (FCA) algorithm can able to initiate crucial improvements in human–robot interactions. This paper, for the first time in the literature, proposes a three-stage FCA algorithm for the HR. At the initial stage of this algorithm, the HR detects the face with the Viola–Jones algorithm, and then important facial distance measurements are obtained with the geometric-based facial distance measurement technique. Finally, the measured facial distances are evaluated with the physiognomy science to reveal the characteristic properties of a person. Even though the proposed algorithm can be implemented to all HR, in this paper, it has been specifically applied to NAO HR. The reliability of the developed FCA algorithm is verified by analyzing each terminal decision about the character and its connection with the measured facial distances in the anatomy science.

Autonomous landing on the deck of a boat or an unmanned surface vehicle (USV) is the minimum requirement for increasing the autonomy of water monitoring missions. This paper introduces an end-to-end control technique based on deep reinforcement learning for landing an unmanned aerial vehicle on a visual marker located on the deck of a USV. The solution proposed consists of a hierarchy of Deep Q-Networks (DQNs) used as high-level navigation policies that address the two phases of the flight: the marker detection and the descending manoeuvre. Few technical improvements have been proposed to stabilize the learning process, such as the combination of vanilla and double DQNs, and a partitioned buffer replay. Simulated studies proved the robustness of the proposed algorithm against different perturbations acting on the marine vessel. The performances obtained are comparable with a state-of-the-art method based on template matching.

As for the obstacle avoidance and formation control for the multi-robot systems, this paper presents an obstacle-avoidance method based on the improved artificial potential field (IAPF) and PID adaptive tracking control algorithm. In order to analyze the dynamics and kinematics of the robot, the mathematical model of the robot is built. Then we construct the motion situational awareness map (MSAM), which can map the environment information around the robot on the MSAM. Based on the MSAM, the IAPF functions are established. We employ the rotating potential field to solve the local minima and oscillations. As for collisions between robots, we build the repulsive potential function and priority model among the robots. Afterwards, the PID adaptive tracking algorithm is utilized to multi-robot formation control. To demonstrate the validity of the proposed method, a series of simulation results confirm that the approaches proposed in this paper can successfully address the obstacle- and collision-avoidance problem while reaching formation.

This paper aims at estimating the tremor torque using extended Kalman filter (EKF) applied to a two-link 3-DOF robot with nonlinear dynamics modelled using Lie-group and Lie-algebra theory. Later, it is generalised to d number of links with (d + 1) -DOF. The configuration of each link at any time is described by its rotation relative to the preceding link. Using this formulation, an elegant formula for the kinetic energy of the (d + 1) -DOF system is obtained as a quadratic form in the angular velocities with coefficients being highly nonlinear trigonometric functions of the angles. Properties of the Lie algebra generators and the Lie adjoint map are used to arrive at this expression. Further, the gravitational potential energy and the torque potential energy are expressed as nonlinear trigonometrical functions of the angles using properties of the SO(3) group. The input torque comprises a nonrandom intentional torque component and a highly nonlinear tremor torque component. The tremor torque is modelled as a stochastic differential equation (sde) satisfying Ornstein–Uhlenbeck (OU) process with diffusion and damping coefficients. Further, the tremor is treated as the disturbance. The Euler–Lagrange equations for the angles are derived. These form a system of sdes, and the EKF is used to get a more accurate disturbance estimate than that provided by the usual disturbance observer. The EKF is based on noisy angle measurements and yields as a bonus the angle and angular velocity estimates on a real-time basis. The parameters in the OU process model of the tremor torque, and parameters of the Fourier components of the intentional torque have also been estimated.

This paper introduces an approach of collaborative control for individual robots to collaboratively perform a common task, without the need for a centralized controller to coordinate the group. The approach is illustrated by an application example involving multiple robots performing a collaborative task to achieve a common goal. The objective of this example problem is to control multiple robots that are connected to an object through elastic cables in order to bring the object to a target position. There is no communication between the robots, and hence each robot is unaware of how the other robots are going to react at any instant. Only the information pertaining to the object and the target is available to all the robots at any instant. Genetic fuzzy system (GFS) is used to develop controller for each of the robots. The nonlinearity of fuzzy logic systems coupled with the search capability of genetic algorithms provides a tool to design controllers for such collaborative tasks. A set of training scenarios are developed to train the individual robot controllers for this task. The trained controllers are then tested on an extensive set of scenarios. This paper describes the development process of GFS controllers for dynamic case involving systems consisting of three robots. It is also shown that the GFS controllers are scalable for the more complex systems involving more than three robots.

In this paper, a neural network (NN)-based tracking controller is proposed for a general class of type (m,s) wheeled mobile manipulators (WMMs) subjected to model uncertainties with prescribed transient and steady-state performance specifications. First, an input–output model of WMMs is derived by introducing proper output equations. Then, the prescribed performance technique is employed to propose a proportional integral derivative trajectory tracking controller for WMMs to ensure that the tracking errors converge to a smaller, arbitrary ultimate bound with a predefined maximum overshoot/undershoot and convergence speed. The learning capabilities of multilayer NNs are incorporated into the controller to approximate the uncertain nonlinear dynamics of the robot. An adaptive saturation-type controller is utilized to compensate NN estimation errors and external disturbances. A Lyapunov-based stability analysis is used to demonstrate that the tracking errors are uniformly ultimately bounded and converge to a small neighborhood of zero with a guaranteed prescribed performance. Numerical computer simulations are presented to show the effectiveness of the proposed controller.

Robots often need to accomplish some complex tasks such as surveillance, response and obstacle avoidance. In this paper, a dynamic search method is proposed to generate optimal robot trajectories satisfying complex task requirement in uncertain environment. The LTL-A* algorithm is presented to generate a global optimal path and the A* algorithm is provided to modify the global optimal path. The task is specified by a linear temporal logic (LTL) formula, and a weighted transition system according to the known information in uncertain environment is modeled to describe the robot motion. Subsequently, a product automaton is constructed by combining the transition system with the task requirement. Based on the product automaton, the LTL-A* algorithm is proposed to generate a global optimal path. The local path planning based on the A* algorithm is employed to deal with the environment change during the process of tracking the global optimal path for the robot. The results of the simulation and experiments show that the proposed method can not only meet the complex task requirement in uncertain environment but also improve the search efficiency.

In this study, the kinematics and dynamics of a single actuator wave (SAW)-like robot are explored. Comprising a helical spine and links, SAW has the potential for miniaturization. A kinematic model for SAW is firstly established, and the dynamic equation of motion is derived based on Kane’s method. For validation, the motion of SAW is simulated using both MATLAB and ADAMS, and the comparison of results demonstrates the effectiveness of the theoretical models. Then the inverse dynamic analysis is performed to reveal the power consumption. Finally, robot prototypes are developed and tested to confirm the robot velocity predicted by simulations.

For a real-time robotic prosthetic control, gait event detection plays an important role. In this paper, one novel sensor was proposed to realize gait event detection. The sensor includes one strain gauge bridge, which can reflect the entire deformation of carbon-fiber footplate on a robotic prosthesis. Three unilateral transtibial amputees participated in the experiments. Experimental results show that using the proposed sensor method, gait event detection (stance phase and swing phase) accuracy is approximately 100%. Based on the detected gait events, three locomotion modes (sit, stand, and walk) and the corresponding transition modes could be determined. Difference between different gait event detection systems was further conducted.

Reducing the energy consumed by a car-like mobile robot makes it possible to move at a lower cost, yet it takes more working time. This paper proposes an optimization algorithm for trajectories with optimal times and analyzes the consequences of restricting the energy consumed on the trajectory obtained for a car-like robot. When modeling the dynamic behavior of the vehicle, it is necessary to consider its inertial parameters, the behavior of the motor, and the basic properties of the tire in its interaction with the ground. To obtain collision-free, minimum-time trajectories quadratic sequential optimization techniques are used, where the objective function is the time taken by the robot to move between two given configurations. This is subject to constraints relating to the vehicle and tires as well as the energy consumed, which is the basis for this paper. We work with a real random distribution of consumed energy values following a normal Gaussian distribution in order to analyze its influence on the trajectories obtained by the vehicle. The energy consumed, the time taken, the maximum velocity reached, and the distance traveled are analyzed in order to characterize the properties of the trajectories obtained. The proposed algorithm has been applied to 101 examples, showing that the computational times needed to obtain the solutions are always lower than those required to realize the trajectories. The results obtained allow us to reach conclusions about the energy efficiency of the trajectories.