arXiv subjects

Yicheng Zhang

Publications and source records attributed to Yicheng Zhang.

At least 55 records · Page 3Linked to original sources

arXiv2023

The suppression of Finite Size Effect within a Few Lattices

Boundary modes localized on the boundaries of a finite-size lattice experience a finite size effect (FSE) that could result in unwanted couplings, crosstalks and formation of gaps even in topological boundary modes. It is commonly believed that the FSE decays exponentially with the size of the system and thus requires many lattices before eventually becoming negligibly small. Here we identify a special type of FSE of some boundary modes that apparently vanishes at some particular wave vectors along the boundary. Meanwhile, the number of wave vectors where the FSE vanishes equals the number of lattices across the strip. We analytically prove this type of FSE in a simple model and prove this peculiar feature. We also provide a physical system consisting of a plasmonic sphere array where this FSE is present. Our work points to the possibility of almost arbitrarily tunning of the FSE, which facilitates unprecedented manipulation of the coupling strength between modes or channels such as the integration of multiple waveguides and photonic non-abelian braiding.

physics.class-ph↗

arXiv2022

Enhanced compound-protein binding affinity prediction by representing protein multimodal information via a coevolutionary strategy

Due to the lack of a method to efficiently represent the multimodal information of a protein, including its structure and sequence information, predicting compound-protein binding affinity (CPA) still suffers from low accuracy when applying machine learning methods. To overcome this limitation, in a novel end-to-end architecture (named FeatNN), we develop a coevolutionary strategy to jointly represent the structure and sequence features of proteins and ultimately optimize the mathematical models for predicting CPA. Furthermore, from the perspective of data-driven approach, we proposed a rational method that can utilize both high- and low-quality databases to optimize the accuracy and generalization ability of FeatNN in CPA prediction tasks. Notably, we visually interpret the feature interaction process between sequence and structure in the rationally designed architecture. As a result, FeatNN considerably outperforms the state-of-the-art (SOTA) baseline in virtual drug screening tasks, indicating the feasibility of this approach for practical use. FeatNN provides an outstanding method for higher CPA prediction accuracy and better generalization ability by efficiently representing multimodal information of proteins via a coevolutionary strategy.

q-bio.BM↗

arXiv2022

Direct observation of hydrodynamization and local prethermalization

Hydrodynamics accurately describes relativistic heavy-ion collision experiments well before local thermal equilibrium is established. This unexpectedly rapid onset of hydrodynamics -- which takes place on the fastest available timescale -- is called hydrodynamization. It occurs when an interacting quantum system is quenched with an energy density that is much greater than its initial energy density. During hydrodynamization, energy gets redistributed across very different energy scales. Hydrodynamization precedes local equilibration among momentum modes, which is local prethermalization to a generalized Gibbs ensemble in nearly integrable systems or local thermalization in non-integrable systems. Many theories of quantum dynamics postulate local (pre)thermalization, but the associated timescale has not been quantitatively studied. Here we use an array of 1D Bose gases to directly observe both hydrodynamization and local prethermalization. After we apply a Bragg scattering pulse, hydrodynamization is evident in the fast redistribution of energy among distant momentum modes, which occurs on timescales associated with the Bragg peak energies. Local prethermalization can be seen in the slower redistribution of occupation among nearby momentum modes. We find that the time scale for local prethermalization in our system is inversely proportional to the momenta involved. During hydrodynamization and local prethermalization, existing theories cannot quantitatively model our experiment. Exact theoretical calculations in the Tonks-Girardeau limit show qualitatively similar features.

cond-mat.quant-gas↗

arXiv2022

STGIN: A Spatial Temporal Graph-Informer Network for Long Sequence Traffic Speed Forecasting

Accurate long series forecasting of traffic information is critical for the development of intelligent traffic systems. We may benefit from the rapid growth of neural network analysis technology to better understand the underlying functioning patterns of traffic networks as a result of this progress. Due to the fact that traffic data and facility utilization circumstances are sequentially dependent on past and present situations, several related neural network techniques based on temporal dependency extraction models have been developed to solve the problem. The complicated topological road structure, on the other hand, amplifies the effect of spatial interdependence, which cannot be captured by pure temporal extraction approaches. Additionally, the typical Deep Recurrent Neural Network (RNN) topology has a constraint on global information extraction, which is required for comprehensive long-term prediction. This study proposes a new spatial-temporal neural network architecture, called Spatial-Temporal Graph-Informer (STGIN), to handle the long-term traffic parameters forecasting issue by merging the Informer and Graph Attention Network (GAT) layers for spatial and temporal relationships extraction. The attention mechanism potentially guarantees long-term prediction performance without significant information loss from distant inputs. On two real-world traffic datasets with varying horizons, experimental findings validate the long sequence prediction abilities, and further interpretation is provided.

cs.LG↗

arXiv2022

Dense-TNT: Efficient Vehicle Type Classification Neural Network Using Satellite Imagery

Accurate vehicle type classification serves a significant role in the intelligent transportation system. It is critical for ruler to understand the road conditions and usually contributive for the traffic light control system to response correspondingly to alleviate traffic congestion. New technologies and comprehensive data sources, such as aerial photos and remote sensing data, provide richer and high-dimensional information. Also, due to the rapid development of deep neural network technology, image based vehicle classification methods can better extract underlying objective features when processing data. Recently, several deep learning models have been proposed to solve the problem. However, traditional pure convolutional based approaches have constraints on global information extraction, and the complex environment, such as bad weather, seriously limits the recognition capability. To improve the vehicle type classification capability under complex environment, this study proposes a novel Densely Connected Convolutional Transformer in Transformer Neural Network (Dense-TNT) framework for the vehicle type classification by stacking Densely Connected Convolutional Network (DenseNet) and Transformer in Transformer (TNT) layers. Three-region vehicle data and four different weather conditions are deployed for recognition capability evaluation. Experimental findings validate the recognition ability of our proposed vehicle classification model with little decay, even under the heavy foggy weather condition.

cs.CV↗

arXiv2022

A SUMO Framework for Deep Reinforcement Learning Experiments Solving Electric Vehicle Charging Dispatching Problem

In modern cities, the number of Electric vehicles (EV) is increasing rapidly for their low emission and better dynamic performance, leading to increasing demand for EV charging. However, due to the limited number of EV charging facilities, catering to the huge demand for time-consuming EV charging becomes a critical problem. It is quite a challenge to dispatch EVs in the dynamic traffic environment and coordinate interaction among agents. To better serve further research on various related Deep Reinforcment Learning (DRL) EV dispatching algorithms, an efficient simulation environment is necessary to ensure success. As simulator Simulation Urban Mobility (SUMO) is one of the most widely used open-source simulators, it has great significance in creating an environment that satisfies research requirements on SUMO. We aim to improve the efficiency of EV charging station usage and save time for EV users in further work. As a result, we design an EV navigation system on the basis of the traffic simulator SUMO using Jurong Area, Singapore in this paper. Various state-of-the-art DRL algorithms are deployed on the designed testbed to validate the feasibility of the framework in terms of EV charging dispatching problems. Besides EV dispatching problems, the environment can also serve for other reinforcement learning (RL) traffic control problems

eess.SY↗

arXiv2022

AST-GIN: Attribute-Augmented Spatial-Temporal Graph Informer Network for Electric Vehicle Charging Station Availability Forecasting

Electric Vehicle (EV) charging demand and charging station availability forecasting is one of the challenges in the intelligent transportation system. With the accurate EV station situation prediction, suitable charging behaviors could be scheduled in advance to relieve range anxiety. Many existing deep learning methods are proposed to address this issue, however, due to the complex road network structure and comprehensive external factors, such as point of interests (POIs) and weather effects, many commonly used algorithms could just extract the historical usage information without considering comprehensive influence of external factors. To enhance the prediction accuracy and interpretability, the Attribute-Augmented Spatial-Temporal Graph Informer (AST-GIN) structure is proposed in this study by combining the Graph Convolutional Network (GCN) layer and the Informer layer to extract both external and internal spatial-temporal dependence of relevant transportation data. And the external factors are modeled as dynamic attributes by the attribute-augmented encoder for training. AST-GIN model is tested on the data collected in Dundee City and experimental results show the effectiveness of our model considering external factors influence over various horizon settings compared with other baselines.

cs.LG↗

arXiv2022

Statistical properties of the off-diagonal matrix elements of observables in eigenstates of integrable systems

We study the statistical properties of the off-diagonal matrix elements of observables in the energy eigenstates of integrable quantum systems. They have been found to be dense in the spin-1/2 XXZ chain, while they are sparse in noninteracting systems. We focus on the quasimomentum occupation of hard-core bosons in one dimension, and show that the distributions of the off-diagonal matrix elements are well described by generalized Gamma distributions, in both the presence and absence of translational invariance but not in the presence of localization. We also show that the results obtained for the off-diagonal matrix elements of observables in the spin-1/2 XXZ model are well described by a generalized Gamma distribution.

cond-mat.stat-mech↗

arXiv2021

Single-particle eigenstate thermalization in quantum-chaotic quadratic Hamiltonians

We study the matrix elements of local and nonlocal operators in the single-particle eigenstates of two paradigmatic quantum-chaotic quadratic Hamiltonians; the quadratic Sachdev-Ye-Kitaev (SYK2) model and the three-dimensional Anderson model below the localization transition. We show that they display eigenstate thermalization for normalized observables. Specifically, we show that the diagonal matrix elements exhibit vanishing eigenstate-to-eigenstate fluctuations, and a variance proportional to the inverse Hilbert space dimension. We also demonstrate that the ratio between the variance of the diagonal and the off-diagonal matrix elements is $2$, as predicted by the random matrix theory. We study distributions of matrix elements of observables and establish that they need not be Gaussian. We identify the class of observables for which the distributions are Gaussian.

cond-mat.stat-mech↗

arXiv2021

Emergent eigenstate solution for generalized thermalization

Generalized thermalization is a process that occurs in integrable systems in which unitary dynamics, e.g., following a quantum quench, results in states in which observables after equilibration are described by generalized Gibbs ensembles (GGEs). Here we discuss an emergent eigenstate construction that allows one to built emergent local Hamiltonians of which one eigenstate captures the entire generalized thermalization process following a global quantum quench. Specifically, we study the emergent eigenstate that describes the quantum dynamics of hard-core bosons in one dimension (1D) for which the initial state is a density wave and they evolve under a homogeneous Hamiltonian.

cond-mat.quant-gas↗

arXiv2020

Generalized hydrodynamics in strongly interacting 1D Bose gases

The dynamics of strongly interacting many-body quantum systems are notoriously complex and difficult to simulate. A new theory, generalized hydrodynamics (GHD), promises to efficiently accomplish such simulations for nearly-integrable systems. It predicts the evolution of the distribution of rapidities, which are the momenta of the quasiparticles in integrable systems. GHD was recently tested experimentally for weakly interacting atoms, but its applicability to strongly interacting systems has not been experimentally established. Here we test GHD with bundles of one-dimensional (1D) Bose gases by performing large trap quenches in both the strong and intermediate coupling regimes. We measure the evolving distribution of rapidities, and find that theory and experiment agree well over dozens of trap oscillations, for average dimensionless coupling strengths that range from 0.3 to 9.3. By also measuring momentum distributions, we gain experimental access to the interaction energy and thus to how the quasiparticles themselves evolve. The accuracy of GHD demonstrated here confirms its wide applicability to the simulation of nearly-integrable quantum dynamical systems. Future experimental studies are needed to explore GHD in spin chains, as well as the crossover between GHD and regular hydrodynamics in the presence of stronger integrability breaking perturbations.

cond-mat.quant-gas↗

arXiv2020

A Distributed Architecture for Real-time Hybrid Traffic Light Control in Urban Transportation Networks

A macroscopic model is proposed to depict the traffic dynamics involved in urban traffic systems. The link dynamics are described based on the cell-transmission model and bounded by the link capacities, while the flow dynamics are proposed based on the discharge headways and saturation flow at intersections. To fulfill the requirement of a closed-loop traffic light control strategy, an approach to estimate the branching ratios at intersections is proposed and simulations show that the convergence would be achieved under constant cyclic flow profiles. Furthermore, a system partitioning approach is proposed based congestion level identification, which is achieved via a machine learning method and a hybrid traffic network control strategy is proposed to integrate different traffic light control schemes together.

math.OC↗

arXiv2019

Information measures for a local quantum phase transition: Lattice bosons in a one-dimensional harmonic trap

We study ground-state quantum entanglement in the one-dimensional Bose-Hubbard model in the presence of a harmonic trap. We focus on two transitions that occur upon increasing the characteristic particle density: the formation of a Mott-insulating domain with site occupation one at the center of the trap (lower transition) and the emergence of a superfluid domain at the center of the Mott-insulating one (upper transition). These transitions generate discontinuities in derivatives of the total energy and have been characterized by local (nonextensive) order parameters, so we refer to them as local quantum phase transitions. We show that a second derivative of the total energy is continuous with a kink at the lower transition, and that it is discontinuous at the upper transition. We also show that bipartite entanglement entropies are order parameters for those local quantum phase transitions. We use the density-matrix renormalization group and show that the transition points extracted from entanglement measures agree with the predictions of the local density approximation in the thermodynamic limit. We discuss how to determine the transition points from results in small systems, such as the ones realized in recent optical lattice experiments that measured the second-order Renyi entanglement entropy.

cond-mat.stat-mech↗

arXiv2019

Quantum dynamics of impenetrable SU(N) fermions in one-dimensional lattices

We study quantum quench dynamics in the Fermi-Hubbard model, and its SU($N$) generalizations, in one-dimensional lattices in the limit of infinite onsite repulsion between all flavors. We consider families of initial states with generalized Neel order, namely, initial state in which there is a periodic $N$-spin pattern with consecutive fermions carrying distinct spin flavors. We introduce an exact approach to describe the quantum evolution of those systems, and study two unique transient phenomena that occur during expansion dynamics in finite lattices. The first one is the dynamical emergence of Gaussian one-body correlations during the melting of sharp (generalized) Neel domain walls. Those correlations resemble the ones in the ground state of the SU($N$) model constrained to the same spin configurations. This is explained using an emergent eigenstate solution to the quantum dynamics. The second phenomenon is the transformation of the quasimomentum distribution of the expanding strongly interacting SU($N$) gas into the rapidity distribution after long times. Finally, we study equilibration in SU($N$) gasses and show that observables after equilibration are described by a generalized Gibbs ensemble. Our approach can be used to benchmark analytical and numerical calculations of dynamics of strongly correlated SU($N$) fermions at large $U$.

cond-mat.quant-gas↗

arXiv2019

Observation of Dynamical Fermionization

We observe dynamical fermionization, where the momentum distribution of a Tonks-Girardeau (T-G) gas of strongly interacting bosons in 1D evolves from bosonic to fermionic after its axial confinement is removed. The asymptotic momentum distribution after expansion in 1D is the distribution of rapidities, which are the conserved quantities associated with many-body integrable systems. Rapidities have not previously been measured in any interacting many-body quantum system. Our measurements agree well with T-G gas theory. We also study momentum evolution after the trap depth is suddenly changed to a new non-zero value. We observe the predicted bosonic-fermionic oscillations and see deviations from the theory outside of the T-G gas limit.

cond-mat.quant-gas↗

arXiv2018

Impenetrable SU(N) fermions in one-dimensional lattices

We study SU(N) fermions in the limit of infinite on-site repulsion between all species. We focus on states in which every pair of consecutive fermions carries a different spin flavor. Since the particle order cannot be changed (because of the infinite on-site repulsion) and contiguous fermions have a different spin flavor, we refer to the corresponding constrained model as the model of distinguishable quantum particles. We introduce an exact numerical method to calculate equilibrium one-body correlations of distinguishable quantum particles based on a mapping onto noninteracting spinless fermions. In contrast to most many-body systems in one dimension, which usually exhibit either power-law or exponential decay of off-diagonal one-body correlations with distance, distinguishable quantum particles exhibit a Gaussian decay of one-body correlations in the ground state, while finite-temperature correlations are well described by stretched exponential decay.

cond-mat.quant-gas↗

arXiv2018

Information measures for a local quantum phase transition: Lattice fermions in a one-dimensional harmonic trap

We use quantum information measures to study the local quantum phase transition that occurs for trapped spinless fermions in one-dimensional lattices. We focus on the case of a harmonic confinement. The transition occurs upon increasing the characteristic density and results in the formation of a band-insulating domain in the center of the trap. We show that the ground-state bipartite entanglement entropy can be used as an order parameter to characterize this local quantum phase transition. We also study excited eigenstates by calculating the average von Neumann and second Renyi eigenstate entanglement entropies, and compare the results with the thermodynamic entropy and the mutual information of thermal states at the same energy density. While at low temperatures we observe a linear increase of the thermodynamic entropy with temperature at all characteristic densities, the average eigenstate entanglement entropies exhibit a strikingly different behavior as functions of temperature below and above the transition. They are linear in temperature below the transition but exhibit activated behavior above it. Hence, at nonvanishing energy densities above the ground state, the average eigenstate entanglement entropies carry fingerprints of the local quantum phase transition.

cond-mat.stat-mech↗

arXiv2017

Bicycle Detection Based On Multi-feature and Multi-frame Fusion in low-resolution traffic videos

As a major type of transportation equipments, bicycles, including electrical bicycles, are distributed almost everywhere in China. The accidents caused by bicycles have become a serious threat to the public safety. So bicycle detection is one major task of traffic video surveillance systems in China. In this paper, a method based on multi-feature and multi-frame fusion is presented for bicycle detection in low-resolution traffic videos. It first extracts some geometric features of objects from each frame image, then concatenate multiple features into a feature vector and use linear support vector machine (SVM) to learn a classifier, or put these features into a cascade classifier, to yield a preliminary detection result regarding whether an object is a bicycle. It further fuses these preliminary detection results from multiple frames to provide a more reliable detection decision, together with a confidence level of that decision. Experimental results show that this method based on multi-feature and multi-frame fusion can identify bicycles with high accuracy and low computational complexity. It is, therefore, applicable for real-time traffic video surveillance systems.

cs.CV↗