Searcharxiv⌕ Search

arXiv subjects

Tian Yu

Publications and source records attributed to Tian Yu.

31 records · Page 2Linked to original sources

Edge Deep Learning Enabled Freezing of Gait Detection in Parkinson's Patients

This paper presents the design of a wireless sensor network for detecting and alerting the freezing of gait (FoG) symptoms in patients with Parkinson's disease. Three sensor nodes, each integrating a 3-axis accelerometer, can be placed on a patient at ankle, thigh, and truck. Each sensor node can independently detect FoG using an on-device deep learning (DL) model, featuring a squeeze and excitation convolutional neural network (CNN). In a validation using a public dataset, the prototype developed achieved a FoG detection sensitivity of 88.8% and an F1 score of 85.34%, using less than 20 k trainable parameters per sensor node. Once FoG is detected, an auditory signal will be generated to alert users, and the alarm signal will also be sent to mobile phones for further actions if needed. The sensor node can be easily recharged wirelessly by inductive coupling. The system is self-contained and processes all user data locally without streaming data to external devices or the cloud, thus eliminating the cybersecurity risks and power penalty associated with wireless data transmission. The developed methodology can be used in a wide range of applications.

eess.SP↗

Reinforcement Learning Based User-Guided Motion Planning for Human-Robot Collaboration

Robots are good at performing repetitive tasks in modern manufacturing industries. However, robot motions are mostly planned and preprogrammed with a notable lack of adaptivity to task changes. Even for slightly changed tasks, the whole system must be reprogrammed by robotics experts. Therefore, it is highly desirable to have a flexible motion planning method, with which robots can adapt to specific task changes in unstructured environments, such as production systems or warehouses, with little or no intervention from non-expert personnel. In this paper, we propose a user-guided motion planning algorithm in combination with the reinforcement learning (RL) method to enable robots automatically generate their motion plans for new tasks by learning from a few kinesthetic human demonstrations. To achieve adaptive motion plans for a specific application environment, e.g., desk assembly or warehouse loading/unloading, a library is built by abstracting features of common human demonstrated tasks. The definition of semantical similarity between features in the library and features of a new task is proposed and further used to construct the reward function in RL. The RL policy can automatically generate motion plans for a new task if it determines that new task constraints can be satisfied with the current library and request additional human demonstrations. Multiple experiments conducted on common tasks and scenarios demonstrate that the proposed user-guided RL-assisted motion planning method is effective.

eess.SY↗

Bistability and equilibria of creased annular sheets and strips

A creased thin disk is generally bistable since the crease could be pushed through to form a stable cone-like inverted state with an elastic singularity corresponding to the vertex of the conical surface. In a recent study, we found that this bistability could be destroyed by removing the singularity through cutting a hole around the vertex, depending on the size and shape of the hole. This paper extends our recent work and is based on the following observations in tabletop models of creased disks with circular holes: (i) reducing the circumference of the creased disk by removing an annular sector could increase the hole size to be as large as the disk without destroying the bistability, (ii) with a single crease, the circular hole could be as large as the disk without loss of the bistability, and (iii) a family of stable inverted states can be obtained by inverting the disk almost anywhere along the crease. An inextensible strip model is implemented to investigate these phenomena. We formulate a minimal facet of the creased disk as a two-point boundary value problem with the creases modeled as nonlinear hinges, and use numerical continuation to conduct parametric studies. Specifically, we focus on geometric parameters which include an angle deficit that determines the circumference of the disk, the rest crease angle, the number of evenly distributed creases, and an eccentricity that determines the position of the hole on the crease. Our numerical results confirm the qualitative observations in (i)-(iii) and further reveal unexpected results caused by the coupling between these geometric parameters. Our results demonstrate that by varying the geometry of a simply creased disk, surprisingly rich nonlinear behaviors can be obtained, which shed new light on the mechanics and design of origami, kirigami, and morphable structures.

cond-mat.soft↗

Wetting dynamics in an angular channel

We analyze the dynamics of liquid filling in a thin, slightly inflated rectangular channel driven by capillary forces. We show that although the amount of liquid $m$ in the channel increases in time following the classical Lucas-Washburn law, $m \propto t^{1/2}$, the prefactor is very sensitive to the deformation of the channel because the filling takes place by the growth of two parts, the bulk part (where the cross-section is completely filled by the liquid), and the finger part (where the cross-section is partially filled). We calculate the time dependence of $m$ accounting for the coupling between the two parts and show that the prefactor for the filling can be reduced significantly by a slight deformation of the rectangular channel, e.g., the prefactor is reduced 50% for a strain of 0.1%. This offers an explanation for the large deviation in the value of the prefactor reported previously.

cond-mat.soft↗

Numerical modeling of static equilibria and bifurcations in bigons and bigon rings

In this study, we explore the mechanics of a bigon and a bigon ring from a combination of experiments and numerical simulations. A bigon is a simple elastic network consisting of two initially straight strips that are deformed to intersect with each other through a fixed intersection angle at each end. A bigon ring is a novel multistable structure composed of a series of bigons arranged to form a loop. We find that a bigon ring usually contains several families of stable states and one of them is a multiply-covered loop, which is similar to the folding behavior of a bandsaw blade. To model bigons and bigon rings, we propose a numerical framework combining several existing techniques to study mechanics of elastic networks consisting of thin strips. Each strip is modeled as a Kirchhoff rod, and the entire strip network is formulated as a two-point boundary value problem (BVP) that can be solved by a general-purpose BVP solver. Together with numerical continuation, we apply the numerical framework to study static equilibria and bifurcations of the bigons and bigon rings. Both numerical and experimental results show that the intersection angle and the aspect ratio of the strip's cross section contribute to the bistability of a bigon and the multistability of a bigon ring; the latter also depends on the number of bigon cells in the ring. The numerical results further reveal interesting connections among various stable states in a bigon ring. Our numerical framework can be applied to general elastic rod networks that may contain flexible joints, naturally curved strips of different lengths, etc. The folding and multistable behaviors of a bigon ring may inspire the design of novel deployable and morphable structures

cond-mat.soft↗

Wetting equilibrium in a rectangular channel

When a capillary channel with corners is wetted by a fluid, there are regions where the fluid fills the whole cross-section and regions where only the corners are filled by the fluid. The fluid fraction of the partially-filled region, $s^*$, is an important quantity related to the capillary pressure. We calculate the value of $s^*$ for channels with a cross-section slightly deviated from a rectangle: the height is larger in the center than those on the two short sides. We find that a small change in the cross-section geometry leads to a huge change of $s^*$. This result is consistent with experimental observations.

cond-mat.soft↗

Mastering the working sequence in human-robot collaborative assembly based on reinforcement learning

A long-standing goal of the Human-Robot Collaboration (HRC) in manufacturing systems is to increase the collaborative working efficiency. In line with the trend of Industry 4.0 to build up the smart manufacturing system, the Co-robot in the HRC system deserves better designing to be more self-organized and to find the superhuman proficiency by self-learning. Inspired by the impressive machine learning algorithms developed by Google Deep Mind like Alphago Zero, in this paper, the human-robot collaborative assembly working process is formatted into a chessboard and the selection of moves in the chessboard is used to analogize the decision making by both human and robot in the HRC assembly working process. To obtain the optimal policy of working sequence to maximize the working efficiency, the robot is trained with a self-play algorithm based on reinforcement learning, without guidance or domain knowledge beyond game rules. A neural network is also trained to predict the distribution of the priority of move selections and whether a working sequence is the one resulting in the maximum of the HRC efficiency. An adjustable desk assembly is used to demonstrate the proposed HRC assembly algorithm and its efficiency.

cs.RO↗

Exact and approximate mechanisms for pure bending of sheets

Direct measurement of the moment-curvature response of sheets or wires up to high curvatures can aid in modeling creasing, pleating, and other forming operations. We consider theoretical issues related to the geometry of pure bending. We present a linkage design that, for homogeneous deformation of the sample, results in an exact pure bending state up to arbitrarily high curvature, by combining a cochleoidal trajectory with an angle-doubling mechanism. The full mechanism is single degree of freedom and position controlled, a desirable feature for measurement of soft materials. We also present an optimal approximation to the cochleoid using a more easily implemented circular trajectory, and compare this with an existing commercial system for fabric testing. We quantify the error of the approximate test by calculating the deformation of an Euler \emph{elastica}, a structure with linear moment-curvature response. While the circular mechanism can approximate the exact boundary conditions quite well up to moderately large curvatures, the resulting curvature of the test sample still deviates significantly from homogeneous response. We also briefly discuss expectations of localization behavior in pure bending tests.

cond-mat.soft↗

Dynamics of Taylor Rising

We study the dynamics of liquid climbing in a narrow and tilting corner, inspired by recent work on liquid transportation on the peristome surface of Nepenthes alata. Considering the balance of gravity, interfacial tension and viscous force, we derive a partial differential equation for the meniscus profile, and numerically study the behavior of the solution for various tilting angle $β$. We show that the liquid height $h(t)$ at time $t$ satisfy the same scaling law found for vertical corner, i.e., $h(t) \propto t^{1/3}$ for large $t$, but the coefficient depends on the tilting angle $β$. The coefficient can be calculated approximately by Onsager principle, and the result agrees well with that obtained by numerical calculation. Our model can be applied for a weakly curved corner and may provide guidance to the design of biomimetic surfaces for liquid transportation.

cond-mat.soft↗

Capillary imbibition in a square tube

When a square tube is brought in contact with bulk liquid, the liquid wets the corners of the tube, and creates finger-like wetted region. The wetting of the liquid then takes place with the growth of two parts, the bulk part where the cross section is entirely filled with the liquid and the finger part where the cross section of the tube is partially filled. In the previous works, the growth of these two parts has been discussed separately. Here we conduct the analysis by explicitly accounting for the coupling of the two parts. We propose coupled equations for the liquid imbibition in both parts and show that (a) the length of each part, $h_0$ and $h_1$, both increases in time $t$ following the Lucas-Washburn's law, $h_0 \sim t^{1/2}$ and $h_1 \sim t^{1/2}$, but that (b) the coefficients are different from those obtained in the previous analysis which ignored the coupling.

cond-mat.soft↗

Overcurvature induced multistability of linked conical frusta: How a `bendy straw' holds its shape

We study the origins of multiple mechanically stable states exhibited by an elastic shell comprising multiple conical frusta, a geometry common to reconfigurable corrugated structures such as `bendy straws'. This multistability is characterized by mechanical stability of axially extended and collapsed states, as well as a partially inverted `bent' state that exhibits stability in any azimuthal direction. To understand the origin of this behavior, we study how geometry and internal stress affect the stability of linked conical frusta. We find that tuning geometrical parameters such as the frustum heights and cone angles can provide axial bistability, whereas stability in the bent state requires a sufficient amount of internal pre-stress, resulting from a mismatch between the natural and geometric curvatures of the shell. We analyze the latter effect through curvature analysis during deformation using X-ray computed tomography (CT), and with a simple mechanical model that captures the qualitative behavior of these highly reconfigurable systems.

cond-mat.soft↗

Bifurcations of buckled, clamped anisotropic rods and thin bands under lateral end translations

Motivated by observations of snap-through phenomena in buckled elastic strips subject to clamping and lateral end translations, we experimentally explore the multi-stability and bifurcations of thin bands of various widths and compare these results with numerical continuation of a perfectly anisotropic Kirchhoff rod. Our choice of boundary conditions is not easily satisfied by the anisotropic structures, forcing a cooperation between bending and twisting deformations. We find that, despite clear physical differences between rods and strips, a naive Kirchhoff model works surprisingly well as an organizing framework for the experimental observations. In the context of this model, we observe that anisotropy creates new states and alters the connectivity between existing states. Our results are a preliminary look at relatively unstudied boundary conditions for rods and strips that may arise in a variety of engineering applications, and may guide the avoidance of jump phenomena in such settings. We also briefly comment on the limitations of current strip models.

cond-mat.soft↗

Polar State in Freestanding Strontium Titanate Nanoparticles

Monodispersed strontium titanate nanoparticles were prepared and studied in detail. It is found that ~10 nm as-prepared stoichiometric nanoparticles are in a polar structural state (with possibly ferroelectric properties) over a broad temperature range. A tetragonal structure, with possible reduction of the electronic hybridization is found as the particle size is reduced. In the 10 nm particles, no change in the local Ti-off centering is seen between 20 and 300 K. The results indicate that nanoscale motifs of SrTiO3 may be utilized in data storage as assembled nano-particle arrays in applications where chemical stability, temperature stability and low toxicity are critical issues.

cond-mat.mtrl-sci↗