A GAN-Like Approach for Physics-Based Imitation Learning and Interactive Control

We present a simple and intuitive approach for interactive control of physically simulated characters. Our work builds upon generative adversarial networks (GAN) and reinforcement learning, and introduces an imitation learning framework where an ensemble of classifiers and an imitation policy are trained in tandem given pre-processed reference clips. The classifiers are trained to discriminate the reference motion from the motion generated by the imitation policy, while the policy is rewarded for fooling the discriminators. Using our GAN-like approach, multiple motor control policies can be trained separately to imitate different behaviors. In runtime, our system can respond to external control signal provided by the user and interactively switch between different policies. Compared to existing method, our proposed approach has the following attractive properties: 1) achieves state-of-the-art imitation performance without manually designing and fine tuning a reward function; 2) directly controls the character without having to track any target reference pose explicitly or implicitly through a phase state; and 3) supports interactive policy switching without requiring any motion generation or motion matching mechanism. We highlight the applicability of our approach in a range of imitation and interactive control tasks, while also demonstrating its ability to withstand external perturbations as well as to recover balance. Overall, our approach has low runtime cost and can be easily integrated into interactive applications and games.

Position Domain Synchronization Control For Robotic Manipulators

Developed in this thesis is a new control law focusing on the improvement of contour tracking of robotic manipulators. The new control scheme is a hybrid controller based on position domain control (PDC) and position synchronization control (PSC). On PDC, the system’s dynamics are transformed from time domain to position domain via a one-to-one mapping and the position of the master axis motion is used as reference instead of time. The elimination of the reference motion from the control input improves contouring performance relative to time domain controllers. Conversely, PSC seeks to reduce the error of the systems by diminishing the synchronization error between each agent of the system. The new control law utilizes the aforementioned techniques to maximize the contour performance. The Lyapunov method was used to prove the proposed controller’s stability. The new control law was compared to existing control schemes via simulations of linear and nonlinear contours, and was shown to provide good tracking and contouring performances.

Position Domain Synchronization Control For Robotic Manipulators

Developed in this thesis is a new control law focusing on the improvement of contour tracking of robotic manipulators. The new control scheme is a hybrid controller based on position domain control (PDC) and position synchronization control (PSC). On PDC, the system’s dynamics are transformed from time domain to position domain via a one-to-one mapping and the position of the master axis motion is used as reference instead of time. The elimination of the reference motion from the control input improves contouring performance relative to time domain controllers. Conversely, PSC seeks to reduce the error of the systems by diminishing the synchronization error between each agent of the system. The new control law utilizes the aforementioned techniques to maximize the contour performance. The Lyapunov method was used to prove the proposed controller’s stability. The new control law was compared to existing control schemes via simulations of linear and nonlinear contours, and was shown to provide good tracking and contouring performances.

Research for Space Interception Simulation under Nonlinear Time-Varied Similarity

A nonlinear time-varied similarity(NTVS) simulation algorithm is proposed to solve the problem that the process of space interception is intractable to simulate with high precision in limited platform under the linear constant similarity(LCS). Firstly, the similarity criterions based on the dimensional analysis are given in the simulation of the interception motion at high-speed, and the disadvantage of LCS is analyzed. Then, the reference motion is introduced to establish a time-varied spatio-temporal transform system which scale factors are self-adaptive with time, and the method is proved to satisfy the similarity consistency. Finally the way that linear separation and independent mapping solve the problem that the large magnitude uniform motion in approaching directions and small magnitude overload motion in lateral directions are simulated in a limited ground platform. The results show that the NTVS can simulate the whole process of interception and perform better in accuracy comparing with that via LCS.

Validation of Marker-Less System for the Assessment of Upper Joints Reaction Forces in Exoskeleton Users

This paper presents the validation of a marker-less motion capture system used to evaluate the upper limb stress of subjects using exoskeletons for locomotion. The system fuses the human skeletonization provided by commercial 3D cameras with forces exchanged by the user to the ground through upper limbs utilizing instrumented crutches. The aim is to provide a low cost, accurate, and reliable technology useful to provide the trainer a quantitative evaluation of the impact of assisted gait on the subject without the need to use an instrumented gait lab. The reaction forces at the upper limbs’ joints are measured to provide a validation focused on clinically relevant quantities for this application. The system was used simultaneously with a reference motion capture system inside a clinical gait analysis lab. An expert user performed 20 walking tests using instrumented crutches and force platforms inside the observed volume. The mechanical model was applied to data from the system and the reference motion capture, and numerical simulations were performed to assess the internal joint reaction of the subject’s upper limbs. A comparison between the two results shows a root mean square error of less than 2% of the subject’s body weight.

Diagnosis of attention-deficit hyperactivity disorder using EOG signals: a new approach

A system based on objective data was developed in the diagnosis and follow-up of attention-deficit hyperactivity disorder (ADHD) in this study. First of all, an electronic circuit, with a two-channel instrumentation amplifier designed to detect eye movements in the horizontal and vertical directions via surface electrodes, was developed to obtain the electrooculogram (EOG) signals. In order to provide a controlled analysis of eye movements during the reception of the signal, an attention test with visual stimulus software was developed. Eight patients with ADHD and eight healthy subjects were asked to monitor the stimulus images on the screen in the reference directions of the test system while recording EOG signals. According to the results of the t-test, no significant difference was found (p=0.11) between the healthy group and the reference movement information, whereas a significant difference was found between patients and the reference motion information (p=0.049). According to these results, it was seen that the number of eye movements of healthy individuals was statistically significant. In addition, they were inconsistent with the reference movement information. The level of significance was found to be low in patients. In this study, a new method is presented to test and diagnose individuals who were attention deficit.

A frame-independent comparison metric for discrete motion sequences

To evaluate the kinematic performance of designed mechanisms, a statistical-variance-based metric is proposed in this article to measure the “distance” between two discrete motion sequences: the reference motion and the given task motion. It seeks to establish a metric that is independent of the choice of the fixed frame or moving frame. Quaternions are adopted to represent the rotational part of a spatial pose, and the variance of the set of relative displacements is computed to reflect the difference between two sequences. With this variance-based metric formulation, we show that the comparison results of two spatial discrete motions are not affected by the choice of frames. Both theoretical demonstration and computational example are presented to support this conclusion. In addition, since the deviation error between the task motion and the synthesized motion measured with this metric is independent of the location of frames, those corresponding parameters could be excluded from the optimization algorithm formulated with our frame-independent metric in kinematic synthesis of mechanisms, and the complexity of the algorithm are hereby reduced. An application of a four-bar linkage synthesis problem is presented to illustrate the advantage of the proposed metric.

Why Monday is not in front of Tuesday: On the uses of English and Finnish front adpositions in sequence metaphors of time

I analyze English and Finnishfrontadpositions commonly used insequencemetaphors of time (The days ahead of Christmas are busy). Asequencemetaphor represents the order of temporal entities as their different positions on a path along which they are moving, a temporally earlier entity being ‘ahead of’ a later one, i.e. further advanced in the direction of motion. In English, the motion-relatedfrontprepositionahead ofis conventional insequencemetaphors, as opposed toin front of, which is used only occasionally and often results in ill-formedness (*Monday is in front of Tuesday). In Finnish, the dedicated two-mover adpositionedellä‘ahead of [two-mover]’ is conventional insequencemetaphors, while the generalfrontadpositionedessäis not. Based on an analysis of these adpositions in different scenarios of spatial motion, I argue that the adpositions common insequencemetaphors (ahead of, edellä) evoke a motion frame of reference (motion FoR). In the motion FoR,frontis adjacent to the direction where the moving Ground (the entity designated by the complement of the adposition) is headed. In contrast, adpositions that bear a strong association with the standard (intrinsic or relative) FoRs (in front ofandedessä) are less felicitous insequencemetaphors. Together, the two languages demonstrate how a metaphorical motion scenario in asequencemetaphor is grammatically coded as a stable, unchanging array, where both participants are steadily moving in the same direction.

Modern Control Laws for an Articulated Robotic Arm: Modeling and Simulation

The robotic manipulator has become an integral component of modern industrial automation. The current paper deals with the mathematical modeling and non-linear control of this manipulator. DH-parameters are used to derive kinematic model while the dynamics is based on Euler-Lagrange equation. Two modern control strategies, H∞ and model predictive control (MPC), are investigated to develop the control laws. For an optimal performance, the controllers have been fine-tuned through a simulation conducted in MATLAB/Simulink environment. The designed control laws are subjected to various inputs and tested for effectiveness in transient parameters like settling time and overshoot as well as steady state error. Simulation results confirm the effectiveness of the developed controllers by precisely tracking the reference motion trajectories.