arXiv subjects

Weiming Wang

Publications and source records attributed to Weiming Wang.

62 records · Page 4Linked to original sources

arXiv2019

Design and Optimization of Conforming Lattice Structures

Inspired by natural cellular materials such as trabecular bone, lattice structures have been developed as a new type of lightweight material. In this paper we present a novel method to design lattice structures that conform with both the principal stress directions and the boundary of the optimized shape. Our method consists of two major steps: the first optimizes concurrently the shape (including its topology) and the distribution of orthotropic lattice materials inside the shape to maximize stiffness under application-specific external loads; the second takes the optimized configuration (i.e. locally-defined orientation, porosity, and anisotropy) of lattice materials from the previous step, and extracts a globally consistent lattice structure by field-aligned parameterization. Our approach is robust and works for both 2D planar and 3D volumetric domains. Numerical results and physical verifications demonstrate remarkable structural properties of conforming lattice structures generated by our method.

cs.CE↗

arXiv2019

Suction Grasp Region Prediction using Self-supervised Learning for Object Picking in Dense Clutter

This paper focuses on robotic picking tasks in cluttered scenario. Because of the diversity of poses, types of stack and complicated background in bin picking situation, it is much difficult to recognize and estimate their pose before grasping them. Here, this paper combines Resnet with U-net structure, a special framework of Convolution Neural Networks (CNN), to predict picking region without recognition and pose estimation. And it makes robotic picking system learn picking skills from scratch. At the same time, we train the network end to end with online samples. In the end of this paper, several experiments are conducted to demonstrate the performance of our methods.

cs.RO↗

arXiv2019

Combinational Q-Learning for Dou Di Zhu

Deep reinforcement learning (DRL) has gained a lot of attention in recent years, and has been proven to be able to play Atari games and Go at or above human levels. However, those games are assumed to have a small fixed number of actions and could be trained with a simple CNN network. In this paper, we study a special class of Asian popular card games called Dou Di Zhu, in which two adversarial groups of agents must consider numerous card combinations at each time step, leading to huge number of actions. We propose a novel method to handle combinatorial actions, which we call combinational Q-learning (CQL). We employ a two-stage network to reduce action space and also leverage order-invariant max-pooling operations to extract relationships between primitive actions. Results show that our method prevails over state-of-the art methods like naive Q-learning and A3C. We develop an easy-to-use card game environments and train all agents adversarially from sractch, with only knowledge of game rules and verify that our agents are comparative to humans. Our code to reproduce all reported results will be available online.

cs.LG↗

arXiv2011

Spatiotemporal dynamics in a spatial plankton system

In this paper, we investigate the complex dynamics of a spatial plankton-fish system with Holling type III functional responses. We have carried out the analytical study for both one and two dimensional system in details and found out a condition for diffusive instability of a locally stable equilibrium. Furthermore, we present a theoretical analysis of processes of pattern formation that involves organism distribution and their interaction of spatially distributed population with local diffusion. The results of numerical simulations reveal that, on increasing the value of the fish predation rates, the sequences spots $\rightarrow$ spot-stripe mixtures$\rightarrow$ stripes$\rightarrow$ hole-stripe mixtures holes$\rightarrow$ wave pattern is observed. Our study shows that the spatially extended model system has not only more complex dynamic patterns in the space, but also has spiral waves.

nlin.PS↗

arXiv2008

Complex dynamics of a Holling-type IV predator-prey model

In this paper, we focus on a spatial Holling-type IV predator-prey model which contains some important factors, such as diffusion, noise (random fluctuations) and external periodic forcing. By a brief stability and bifurcation analysis, we arrive at the Hopf and Turing bifurcation surface and derive the symbolic conditions for Hopf and Turing bifurcation in the spatial domain. Based on the stability and bifurcation analysis, we obtain spiral pattern formation via numerical simulation. Additionally, we study the model with colored noise and external periodic forcing. From the numerical results, we know that noise or external periodic forcing can induce instability and enhance the oscillation of the species, and resonant response. Our results show that modeling by reaction-diffusion equations is an appropriate tool for investigating fundamental mechanisms of complex spatiotemporal dynamics.

q-bio.PE↗

arXiv2008

Pattern formation of a predator-prey system with Ivlev-type functional response

In this paper, we investigate the emergence of a predator-prey system with Ivlev-type functional response and reaction-diffusion. We study how diffusion affects the stability of predator-prey coexistence equilibrium and derive the conditions for Hopf and Turing bifurcation in the spatial domain. Based on the bifurcation analysis, we give the spatial pattern formation, the evolution process of the system near the coexistence equilibrium point, via numerical simulation. We find that pure Hopf instability leads to the formation of spiral patterns and pure Turing instability destroys the spiral pattern and leads to the formation of chaotic spatial pattern. Furthermore, we perform three categories of initial perturbations which predators are introduced in a small domain to the coexistence equilibrium point to illustrate the emergence of spatiotemporal patterns, we also find that in the beginning of evolution of the spatial pattern, the special initial conditions have an effect on the formation of spatial patterns, though the effect is less and less with the more and more iterations. This indicates that for prey-dependent type predator-prey model, pattern formations do depend on the initial conditions, while for predator-dependent type they do not. Our results show that modeling by reaction-diffusion equations is an appropriate tool for investigating fundamental mechanisms of complex spatiotemporal dynamics.

q-bio.PE↗

arXiv2008

Spatiotemporal pattern formation of Beddington-DeAngelis-type predator-prey model

In this paper, we investigate the emergence of a predator-prey model with Beddington-DeAngelis-type functional response and reaction-diffusion. We derive the conditions for Hopf and Turing bifurcation on the spatial domain. Based on the stability and bifurcation analysis, we give the spatial pattern formation via numerical simulation, i.e., the evolution process of the model near the coexistence equilibrium point. We find that for the model we consider, pure Turing instability gives birth to the spotted pattern, pure Hopf instability gives birth to the spiral wave pattern, and both Hopf and Turing instability give birth to stripe-like pattern. Our results show that reaction-diffusion model is an appropriate tool for investigating fundamental mechanism of complex spatiotemporal dynamics. It will be useful for studying the dynamic complexity of ecosystems.

q-bio.PE↗

arXiv2007

Spatiotemporal complexity of a ratio-dependent predator-prey system

In this paper, we investigate the emergence of a ratio-dependent predator-prey system with Michaelis-Menten-type functional response and reaction-diffusion. We derive the conditions for Hopf, Turing and Wave bifurcation on a spatial domain. Furthermore, we present a theoretical analysis of evolutionary processes that involves organisms distribution and their interaction of spatially distributed population with local diffusion. The results of numerical simulations reveal that the typical dynamics of population density variation is the formation of isolated groups, i.e., stripelike or spotted or coexistence of both. Our study shows that the spatially extended model has not only more complex dynamic patterns in the space, but also chaos and spiral waves. It may help us better understand the dynamics of an aquatic community in a real marine environment.

q-bio.PE↗