Search arXiv⌕ Search

arXiv subjects

Lin Cheng

Publications and source records attributed to Lin Cheng.

At least 73 records · Page 4Linked to original sources
arXiv2023

Chance-Constrained Generic Energy Storage Operations under Decision-Dependent Uncertainty

Compared with large-scale physical batteries, aggregated and coordinated generic energy storage (GES) resources provide low-cost, but uncertain, flexibility for power grid operations. While GES can be characterized by different types of uncertainty, the literature mostly focuses on decision-independent uncertainties (DIUs), such as exogenous stochastic disturbances caused by weather conditions. Instead, this manuscript focuses on newly-introduced decision-dependent uncertainties (DDUs) and considers an optimal GES dispatch that accounts for uncertain available state-of-charge (SoC) bounds that are affected by incentive signals and discomfort levels. To incorporate DDUs, we present a novel chance-constrained optimization (CCO) approach for the day-ahead economic dispatch of GES units. Two tractable methods are presented to solve the proposed CCO problem with DDUs: (i) a robust reformulation for general but incomplete distributions of DDUs, and (ii) an iterative algorithm for specific and known distributions of DDUs. Furthermore, reliability indices are introduced to verify the applicability of the proposed approach with respect to the reliability of the response of GES units. Simulation-based analysis shows that the proposed methods yield conservative, but credible, GES dispatch strategies and reduced penalty cost by incorporating DDUs in the constraints and leveraging data-driven parameter identification. This results in improved availability and performance of coordinated GES units.

math.OC↗
arXiv2023

Intelligent diagnostic scheme for lung cancer screening with Raman spectra data by tensor network machine learning

Artificial intelligence (AI) has brought tremendous impacts on biomedical sciences from academic researches to clinical applications, such as in biomarkers' detection and diagnosis, optimization of treatment, and identification of new therapeutic targets in drug discovery. However, the contemporary AI technologies, particularly deep machine learning (ML), severely suffer from non-interpretability, which might uncontrollably lead to incorrect predictions. Interpretability is particularly crucial to ML for clinical diagnosis as the consumers must gain necessary sense of security and trust from firm grounds or convincing interpretations. In this work, we propose a tensor-network (TN)-ML method to reliably predict lung cancer patients and their stages via screening Raman spectra data of Volatile organic compounds (VOCs) in exhaled breath, which are generally suitable as biomarkers and are considered to be an ideal way for non-invasive lung cancer screening. The prediction of TN-ML is based on the mutual distances of the breath samples mapped to the quantum Hilbert space. Thanks to the quantum probabilistic interpretation, the certainty of the predictions can be quantitatively characterized. The accuracy of the samples with high certainty is almost 100$\%$. The incorrectly-classified samples exhibit obviously lower certainty, and thus can be decipherably identified as anomalies, which will be handled by human experts to guarantee high reliability. Our work sheds light on shifting the ``AI for biomedical sciences'' from the conventional non-interpretable ML schemes to the interpretable human-ML interactive approaches, for the purpose of high accuracy and reliability.

q-bio.QM↗
arXiv2022

Auction-based Operation in LEO Satellite Systems for High-Efficiency Communications

We propose an auction-based mechanism to improve the efficiency of low earth orbit satellite communication systems. The mechanism allows the ground stations to bid for downlink resources such as spectrum, satellite links, or radios, without the need to send channel status back to satellites. Simulation and experimental results show that this mechanism improves total channel capacity by dynamically leveraging the diversity among satellite-station links; reduces uplink overhead by providing lightweight and effective channel status feedback; simplifies computational complexity and improves scalability; and also provides implicit resource information stemming from the auction dynamics. This new operation mechanism provides a feasible solution for low earth orbit satellites which are sensitive to power consumption and overheating.

cs.NI↗
arXiv2022

Machine Learning Interatomic Potential for Anisotropic Thermal Transport in Bulk Hexagonal Boron Nitride

The highly anisotropic thermal conductivity in layered materials is crucial for a broad range of applications such as thermal management of electronic devices, thermal insulation, and thermoelectrics. Understanding of anisotropic thermal transport in layered materials largely depends on atomistic simulations based on density functional theory (DFT) or empirical potentials, which however suffer either low computational efficiency or accuracy. Recently, machine learning interatomic potentials (MLIPs) are emerging as a powerful tool to bridge the gap. Despite the recent progress in developing MLIPs, little attention has been paid to constructing a potential that can accurately predict the thermal properties of layered materials, which is more challenging compared with the case of isotropic materials because of the highly anisotropic bonding and weak van der Waals interactions in layered materials. Here, we introduce a MLIP within the Gaussian approximation potential (GAP) framework for bulk hexagonal boron nitride (h-BN) with a typical layered structure. The GAP can well predict the highly anisotropic phonon transport properties and thermal conductivity of bulk h-BN with DFT-level accuracy at orders of magnitude reduced cost. Our work demonstrates the ability of GAP to reproduce the subtle features of anisotropic potential energy surfaces of bulk h-BN and potentially other layered materials. Atomistic simulations based on MLIPs are expected to be able to greatly promote the understanding of phonon transport and the prediction of thermophysical properties in layered materials.

cond-mat.mtrl-sci↗
arXiv2022

Risk Assessment with Generic Energy Storage under Exogenous and Endogenous Uncertainty

Current risk assessment ignores the stochastic nature of energy storage availability itself and thus lead to potential risk during operation. This paper proposes the redefinition of generic energy storage (GES) that is allowed to offer probabilistic reserve. A data-driven unified model with exogenous and endogenous uncertainty (EXU & EDU) description is presented for four typical types of GES. Moreover, risk indices are proposed to assess the impact of overlooking (EXU & EDU) of GES. Comparative results between EXU & EDU are illustrated in distribution system with day-ahead chance-constrained optimization (CCO) and more severe risks are observed for the latter, which indicate that system operator (SO) should adopt novel strategies for EDU uncertainty.

q-fin.RM↗
arXiv2022

TSGB: Target-Selective Gradient Backprop for Probing CNN Visual Saliency

The explanation for deep neural networks has drawn extensive attention in the deep learning community over the past few years. In this work, we study the visual saliency, a.k.a. visual explanation, to interpret convolutional neural networks. Compared to iteration based saliency methods, single backward pass based saliency methods benefit from faster speed, and they are widely used in downstream visual tasks. Thus, we focus on single backward pass based methods. However, existing methods in this category struggle to uccessfully produce fine-grained saliency maps concentrating on specific target classes. That said, producing faithful saliency maps satisfying both target-selectiveness and fine-grainedness using a single backward pass is a challenging problem in the field. To mitigate this problem, we revisit the gradient flow inside the network, and find that the entangled semantics and original weights may disturb the propagation of target-relevant saliency. Inspired by those observations, we propose a novel visual saliency method, termed Target-Selective Gradient Backprop (TSGB), which leverages rectification operations to effectively emphasize target classes and further efficiently propagate the saliency to the image space, thereby generating target-selective and fine-grained saliency maps. The proposed TSGB consists of two components, namely, TSGB-Conv and TSGB-FC, which rectify the gradients for convolutional layers and fully-connected layers, respectively. Extensive qualitative and quantitative experiments on the ImageNet and Pascal VOC datasets show that the proposed method achieves more accurate and reliable results than the other competitive methods. Code is available at https://github.com/123fxdx/CNNvisualizationTSGB.

cs.CV↗
arXiv2021

Fourier-Engineered Plasmonic Lattice Resonances

Resonances in optical systems are useful for many applications, such as frequency comb generation, optical filtering, and biosensing. However, many of these applications are difficult to implement in optical metasurfaces because traditional approaches for designing multi-resonant nanostructures require significant computational and fabrication efforts. To address this challenge, we introduce the concept of Fourier lattice resonances (FLRs) in which multiple desired resonances can be chosen a priori and used to dictate the metasurface design. Because each resonance is supported by a distinct surface lattice mode, each can have a high quality factor. Here, we experimentally demonstrate several metasurfaces with arbitrarily placed resonances (e.g., at 1310 and 1550 nm) and Q-factors as high as 800 in a plasmonic platform. This flexible procedure requires only the computation of a single Fourier transform for its design, and is based on standard lithographic fabrication methods, allowing one to design and fabricate a metasurface to fit any specific, optical-cavity-based application. This work represents an important milestone towards the complete control over the transmission spectrum of a metasurface.

physics.optics↗
arXiv2021

Data driven analysis of thermal simulations, microstructure and mechanical properties of Inconel 718 thin walls deposited by metal additive manufacturing

The extreme and repeated temperature variation during additive manufacturing of metal parts has a large effect on the resulting material microstructure and properties. The ability to accurately predict this temperature field in detail, and relate it quantitatively to structure and properties, is a key step in predicting part performance and optimizing process design. In this work, a finite element simulation of the Directed Energy Deposition (DED) process is used to predict the space- and time-dependent temperature field during the multi-layer build process for Inconel 718 walls. The thermal model is validated using the dynamic infrared (IR) images captured in situ during the DED builds, showing good agreement with experimental measurements. The relationship between predicted cooling rate, microstructural features, and mechanical properties is examined, and cooling rate alone is found to be insufficient in giving quantitative property predictions. Because machine learning offers an efficient way to identify important features from series data, we apply a 1D convolutional neural network (CNN) data-driven framework to automatically extract the dominant predictive features from simulated temperature history. The relationship between the CNN-extracted features and the mechanical properties is studied. To interpret how CNN performs in intermediate layers, we visualize the extracted features produced on each convolutional layer by a trained CNN. Our results show that the results predicted by the CNN agree well with experimental measurements and give insights into physical mechanisms of microstructure evolution and mechanical properties.

physics.app-ph↗
arXiv2021

Reinforcement Learning for Standards Design

Communications standards are designed via committees of humans holding repeated meetings over months or even years until consensus is achieved. This includes decisions regarding the modulation and coding schemes to be supported over an air interface. We propose a way to "automate" the selection of the set of modulation and coding schemes to be supported over a given air interface and thereby streamline both the standards design process and the ease of extending the standard to support new modulation schemes applicable to new higher-level applications and services. Our scheme involves machine learning, whereby a constructor entity submits proposals to an evaluator entity, which returns a score for the proposal. The constructor employs reinforcement learning to iterate on its submitted proposals until a score is achieved that was previously agreed upon by both constructor and evaluator to be indicative of satisfying the required design criteria (including performance metrics for transmissions over the interface).

stat.ML↗
arXiv2021

Optimal attosecond pulses generation from oscillating plasma surfaces driven at relativistic intensity

The creation of attosecond pulses via laser-plasma interaction has been a subject of great scientific interest for more than three decades. This process is investigated by using particle-in-cell simulation with varying the plasma and laser parameters. The steepness of the density gradient at the plasma-vacuum interface is examined to see how this parameter affects the high-order harmonic generations and isolated-attosecond pulse creation. The optimal density gradient lengths $L$ are explored within the relativistic oscillating mirror mechanism. Although the ideal gradient lengths for the full width at half maximum and peak intensity of an isolated-attosecond pulse depend on the driving laser intensity independently, they are both found near $L$=0.2$λ$ for high laser intensities.

physics.plasm-ph↗
arXiv2020

GRCNN: Graph Recognition Convolutional Neural Network for Synthesizing Programs from Flow Charts

Program synthesis is the task to automatically generate programs based on user specification. In this paper, we present a framework that synthesizes programs from flow charts that serve as accurate and intuitive specifications. In order doing so, we propose a deep neural network called GRCNN that recognizes graph structure from its image. GRCNN is trained end-to-end, which can predict edge and node information of the flow chart simultaneously. Experiments show that the accuracy rate to synthesize a program is 66.4%, and the accuracy rates to recognize edge and nodes are 94.1% and 67.9%, respectively. On average, it takes about 60 milliseconds to synthesize a program.

cs.CV↗
arXiv2020

Superscattering, Superabsorption, and Nonreciprocity in Nonlinear Antennas

We propose tunable nonlinear antennas based on an epsilon-near-zero material with a large optical nonlinearity. We show that the absorption and scattering cross sections of the antennas can be controlled dynamically from a nearly superscatterer to a nearly superabsorber by changing the intensity of the laser beam. Moreover, we demonstrate that a hybrid nonlinear antenna, composed of epsilon-near-zero and high-index dielectric materials, exhibits nonreciprocal radiation patterns because of broken spatial inversion symmetry and large optical nonlinearity of the epsilon-near-zero material. By changing the intensity of the laser, the radiation pattern of the antenna can be tuned between a bidirectional and a unidirectional emission known as a Huygens source. Our study provides a novel approach toward ultrafast dynamical control of metamaterials, for applications such as beam steering and optical limiting.

physics.optics↗
arXiv2020

Exploring the parameter reusability of CNN

In recent times, using small data to train networks has become a hot topic in the field of deep learning. Reusing pre-trained parameters is one of the most important strategies to address the issue of semi-supervised and transfer learning. However, the fundamental reason for the success of these methods is still unclear. In this paper, we propose a solution that can not only judge whether a given network is reusable or not based on the performance of reusing convolution kernels but also judge which layers' parameters of the given network can be reused, based on the performance of reusing corresponding parameters and, ultimately, judge whether those parameters are reusable or not in a target task based on the root mean square error (RMSE) of the corresponding convolution kernels. Specifically, we define that the success of a CNN's parameter reuse depends upon two conditions: first, the network is a reusable network; and second, the RMSE between the convolution kernels from the source domain and target domain is small enough. The experimental results demonstrate that the performance of reused parameters applied to target tasks, when these conditions are met, is significantly improved.

eess.IV↗
arXiv2020

Learning Object Scale With Click Supervision for Object Detection

Weakly-supervised object detection has recently attracted increasing attention since it only requires image-levelannotations. However, the performance obtained by existingmethods is still far from being satisfactory compared with fully-supervised object detection methods. To achieve a good trade-off between annotation cost and object detection performance,we propose a simple yet effective method which incorporatesCNN visualization with click supervision to generate the pseudoground-truths (i.e., bounding boxes). These pseudo ground-truthscan be used to train a fully-supervised detector. To estimatethe object scale, we firstly adopt a proposal selection algorithmto preserve high-quality proposals, and then generate ClassActivation Maps (CAMs) for these preserved proposals by theproposed CNN visualization algorithm called Spatial AttentionCAM. Finally, we fuse these CAMs together to generate pseudoground-truths and train a fully-supervised object detector withthese ground-truths. Experimental results on the PASCAL VOC2007 and VOC 2012 datasets show that the proposed methodcan obtain much higher accuracy for estimating the object scale,compared with the state-of-the-art image-level based methodsand the center-click based method

cs.CV↗
arXiv2019

Dynamic Bandwidth Allocation in Small-Cell Networks: An Economics Approach

We propose and experimentally demonstrate a bandwidth allocation method based on the comparative advantage of spectral efficiency among users in a multi-tone small-cell radio access system with frequency-selective fading channels. The method allocates frequency resources by ranking the comparative advantage of the spectrum measured at the receivers ends. It improves the overall spectral efficiency of the access system with low implementation complexity and independently of power loading. In a two-user wireless transmission experiment, we observe up to 23.1% average capacity improvement by using the proposed method.

cs.NI↗
arXiv2019

Hallucinated Adversarial Learning for Robust Visual Tracking

Humans can easily learn new concepts from just a single exemplar, mainly due to their remarkable ability to imagine or hallucinate what the unseen exemplar may look like in different settings. Incorporating such an ability to hallucinate diverse new samples of the tracked instance can help the trackers alleviate the over-fitting problem in the low-data tracking regime. To achieve this, we propose an effective adversarial approach, denoted as adversarial "hallucinator" (AH), for robust visual tracking. The proposed AH is designed to firstly learn transferable non-linear deformations between a pair of same-identity instances, and then apply these deformations to an unseen tracked instance in order to generate diverse positive training samples. By incorporating AH into an online tracking-by-detection framework, we propose the hallucinated adversarial tracker (HAT), which jointly optimizes AH with an online classifier (e.g., MDNet) in an end-to-end manner. In addition, a novel selective deformation transfer (SDT) method is presented to better select the deformations which are more suitable for transfer. Extensive experiments on 3 popular benchmarks demonstrate that our HAT achieves the state-of-the-art performance.

cs.CV↗
arXiv2019

Fully Distributed DC Optimal Power Flow Based on Distributed Economic Dispatch and Distributed State Estimation

Optimal power flow (OPF) is an important technique for power systems to achieve optimal operation while satisfying multiple constraints. The traditional OPF are mostly centralized methods which are executed in the centralized control center. This paper introduces a totally Distributed DC Optimal Power Flow (DDCOPF) method for future power systems which have more and more distributed generators. The proposed method is based on the Distributed Economic Dispatch (DED) method and the Distributed State Estimation (DSE) method. In this proposed scheme, the DED method is used to achieve the optimal power dispatch with the lowest cost, and the DSE method provides power flow information of the power system to the proposed DDCOPF algorithm. In the proposed method, the Auto-Regressive (AR) model is used to predict the load variation so that the proposed algorithm can prevent overflow. In addition, a method called constraint algorithm is developed to correct the results of DED with the proposed correction algorithm and penalty term so that the constraints for the power system will not be violated. Different from existing research, the proposed method is completely distributed without need for any centralized facility.

eess.SY↗