Robust Control of Robot Manipulator based on QFT and H∞

Robust controllers have been developed by both control techniques QFT and H∞ applied in the waist, shoulder and elbow of a manipulator of 6 degrees of freedom. The design is based on the identification of a linear model of the robot dynamics which represents the non-linearity of the system using parametric uncertainty. QFT control methodology is used to tune the robust PID-controller and pre-filters of the system, and H∞ controllers are obtained by designing the weighting functions and using the MATLAB hinfopt tool. Finally the performance of robust controllers is compared designed based on the calculation and analysis of some behavioral indices.

Pedestrian motion centroid model based on robot dynamics

In recent years, with the rapid development of computer vision technology, image-based human body research has become an important task, such as pedestrian target detection, trajectory tracking, posture estimation and behaviour recognition. The centre of mass is one of the important characteristics that can reflect the phenomenon of pedestrian movement. This paper first introduces the biped robot model in robotics, starting from forward and inverse kinematics, to find the mapping relationship between the position of each joint and the pose of the end effector. Then, corresponding to the skeleton model of the human joint points, the characteristics of the bone posture and joint angle are determined. The moment of inertia factor is introduced, and the motion superposition of different joint points is considered to establish a pedestrian motion centroid model. By calculating the equivalent dynamic centroid, the pedestrian kinematics law can be explored and the pedestrian movement mechanism can be more deeply recognized.

Performing Energy-Efficient Pick-and-Place Motions for High-Speed Robots by using Variable Stiffness Springs

Typically, for pick-and-place robots operating at high speeds, an enormous amount of energy is lost during the robot braking phase. This is due to the fact that, during such operational phase, most of the energy is dissipated as heat on the braking resistances of the motor drivers. In order to increase the energy-efficiency during the high-speed pick-and-place cycles, this paper investigates the use of variable stiffness springs (VSS) in parallel configuration with the motors. These springs store the energy during the braking phase, instead of dissipating it. The energy is then released to actuate the robot in a next displacement phase. This design approach is combined with a motion generator which seeks to optimize trajectories for input torques reduction (and thus of energy consumption), through solving a boundary value problem (BVP) based on the robot dynamics. Experimental results of the suggested approach on a five-bar mechanism show the drastic reduction of input torques, and therefore of energetic losses.

An insect-scale robot reveals the effects of different body dynamics regimes during open-loop running in feature-laden terrain

The transition from the lab to natural environments is an archetypal challenge in robotics. While larger robots can manage complex limb-ground interactions using sensing and control, such strategies are difficult to implement on small platforms where space and power are limited. The Harvard Ambulatory Microrobot (HAMR) is an insect-scale quadruped capable of effective open-loop running on featureless, hard substrates. Inspired by the predominantly feedforward strategy of rapidly-running cockroaches on uneven terrain [Sponberg, 2007], we used HAMR to explore open-loop running on two 3D printed heterogeneous terrains generated using fractional Brownian motion. The ``pocked'' terrain had foot-scale features throughout while the ``jagged'' terrain features increased in height in the direction of travel. We measured the performance of trot and pronk gaits while varying limb amplitude and stride frequency. The frequencies tested encompassed different dynamics regimes: body resonance (10-25~Hz) and kinematic running (30-40~Hz), with dynamics typical of biological running and walking, respectively, and limb-transmission resonance (45-60~Hz). On the featureless and pocked terrains, low mechanical cost-of-transport (mCoT) kinematic running combinations performed best without systematic differences between trot and pronk; indicating that if terrain features are not too tall, a robot can transition from homo- to heterogeneous environments in open-loop. Pronk bypassed taller features than trot on the jagged terrain, and higher mCoT, lower frequency running was more often effective. While increasing input power to the robot improved performance in general, lower frequency pronking on jagged terrain allowed the robot to bypass taller features compared with the same input power at higher frequencies. This was correlated with the increased variation in center-of-mass orientation occurring at frequencies near body resonance. This study established that appropriate choice of robot dynamics, as mediated by gait, frequency, and limb amplitude, can expand the terrains accessible to microrobots without the addition of sensing or closed-loop control.

Comparison of Control Algorithms by Simulating Power Consumption of Differential Drive Mobile Robot Motion Control in Vineyard Row

The paper deals with comparing electricity power consumption of various control algorithms by simulating differential mobile robot motion control in a vineyard row. In field of autonomous mobile robotics, the quality of control is a crucial aspect. Besides the precision of control, the energy consumption for motion is becoming an increasingly demanding characteristic of a controller due to the increasing costs of fossil fuels and electricity. A simulation model of a differential drive mobile robot motion in a vineyard row was created, including robot dynamics for evaluating motion consumption, and there were implemented commonly used PID, Fuzzy, and LQ control algorithms, the task of which was to navigate the robot through the centre of vineyard row section by measuring distances from trellises on both robot sides. The comparison was carried out using Matlab software and the best results in terms of both power consumption and control accuracy were achieved by LQI controller. The designed model for navigating the robot through the vineyard row centre and optimized controllers were implemented in a real robot and tested under real conditions.

End-Effector Force and Joint Torque Estimation of a 7-DoF Robotic Manipulator Using Deep Learning

When a mobile robotic manipulator interacts with other robots, people, or the environment in general, the end-effector forces need to be measured to assess if a task has been completed successfully. Traditionally used force or torque estimation methods are usually based on observers, which require knowledge of the robot dynamics. Contrary to this, our approach involves two methods based on deep neural networks: robot end-effector force estimation and joint torque estimation. These methods require no knowledge of robot dynamics and are computationally effective but require a force sensor under the robot base. Several different architectures were considered for the tasks, and the best ones were identified among those tested. First, the data for training the networks were obtained in simulation. The trained networks showed reasonably good performance, especially using the LSTM architecture (with a root mean squared error (RMSE) of 0.1533 N for end-effector force estimation and 0.5115 Nm for joint torque estimation). Afterward, data were collected on a real Franka Emika Panda robot and then used to train the same networks for joint torque estimation. The obtained results are slightly worse than in simulation (0.5115 Nm vs. 0.6189 Nm, according to the RMSE metric) but still reasonably good, showing the validity of the proposed approach.

Balance Recoverability and Control of Bipedal Walkers with Foot Slip

Low-friction foot/ground contacts present a particular challenge for stable bipedal walkers. The slippage of the stance foot introduces complexity in robot dynamics and the general locomotion stability results cannot be applied directly. We relax the commonly used assumption of non-slip contact between the walker foot and the ground and examine bipedal dynamics under foot slip. Using a two-mass linear inverted pendulum model, we introduce the concept of balance recoverability and use it to quantify the balanced or fall-prone walking gaits. Balance recoverability also serves as the basis for the design of the balance recovery controller. We design the within- or multi-step recovery controller to assist the walker to avoid fall. The controller performance is validated through simulation results and robustness is demonstrated in the presence of measurement noises as well as variations of foot/ground friction conditions. In addition, the proposed methods and models are used to analyze the data from human walking experiments. The multiple subject experiments validate and illustrate the balance recoverability concept and analyses.

Predicting Flying Robot Dynamics with Deep Learning

With the rapid increase in the power of computing and technological advances in robotics, research in the field of robotics has rapidly become very expansive. Being able to accurately predict movements of a robot is vital to many applications within this field, allowing for more precise simulation and prototyping as well as more accurate control of robotic systems. In this paper, we present an adaptable neural network that accurately predicts the movement of quadcopter robotic agents which can be expanded to encompass many more robots and applications given the requisite data, producing accurate results within a small margin for error.

Learning Neuroplastic Matching of Robot Dynamics in Closed-loop CPGs

Legged robots have the potential to show locomotion performance with reduced control effort and energy efficiency by leveraging elastic structures inspired by animals' elastic tendons and muscles. However, it remains a challenge to match the natural dynamics of complex legged robots and their control task dynamics. Here we present a framework to match control task dynamics and natural dynamics based on the neuroelasticity and neuroplasticity concept. Inspired by animals we design quadruped robot Morti with strong natural dynamics as a testing platform. It is controlled through a bioinspired closed-loop central pattern generator (CPG) that is designed to neuroelastically mitigate short term perturbations using sparse contact feedback. We use the amount of neuroelastic activity as a proxy to quantify the dynamics' mismatching. By minimizing neuroelastic activity, we neuroplastically match the control task dynamics to the robot's natural dynamics. Through matching the robot learns to walk within one hour with only sparse feedback and improves its energy efficiency without explicitly minimizing it in the cost function.

An Historical Perspective on the Control of Robotic Manipulators

This article is an historical overview of control theory applied to robotic manipulators, with an emphasis on the early fundamental theoretical foundations of robot control. It discusses properties of robot dynamics that enable application of advanced control methods followed by robust and adaptive control of manipulators. It also discusses nonlinear control of underactuated robots and teleoperators. Expected final online publication date for the Annual Review of Control, Robotics, and Autonomous Systems, Volume 5 is May 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.