Searcharxiv⌕ Search

arXiv subjects

Hongbin Sun

Publications and source records attributed to Hongbin Sun.

At least 91 records · Page 5Linked to original sources
arXiv2021

threaTrace: Detecting and Tracing Host-based Threats in Node Level Through Provenance Graph Learning

Host-based threats such as Program Attack, Malware Implantation, and Advanced Persistent Threats (APT), are commonly adopted by modern attackers. Recent studies propose leveraging the rich contextual information in data provenance to detect threats in a host. Data provenance is a directed acyclic graph constructed from system audit data. Nodes in a provenance graph represent system entities (e.g., $processes$ and $files$) and edges represent system calls in the direction of information flow. However, previous studies, which extract features of the whole provenance graph, are not sensitive to the small number of threat-related entities and thus result in low performance when hunting stealthy threats. We present threaTrace, an anomaly-based detector that detects host-based threats at system entity level without prior knowledge of attack patterns. We tailor GraphSAGE, an inductive graph neural network, to learn every benign entity's role in a provenance graph. threaTrace is a real-time system, which is scalable of monitoring a long-term running host and capable of detecting host-based intrusion in their early phase. We evaluate threaTrace on three public datasets. The results show that threaTrace outperforms three state-of-the-art host intrusion detection systems.

cs.CR↗
arXiv2021

A Federated Learning Framework for Smart Grids: Securing Power Traces in Collaborative Learning

With the deployment of smart sensors and advancements in communication technologies, big data analytics have become vastly popular in the smart grid domain, informing stakeholders of the best power utilization strategy. However, these power-related data are stored and owned by different parties. For example, power consumption data are stored in numerous transformer stations across cities; mobility data of the population, which are important indicators of power consumption, are held by mobile companies. Direct data sharing might compromise party benefits, individual privacy and even national security. Inspired by the federated learning scheme from Google AI, we propose a federated learning framework for smart grids, which enables collaborative learning of power consumption patterns without leaking individual power traces. Horizontal federated learning is employed when data are scattered in the sample space; vertical federated learning, on the other hand, is designed for the case with data scattered in the feature space. Case studies show that, with proper encryption schemes such as Paillier encryption, the machine learning models constructed from the proposed framework are lossless, privacy-preserving and effective. Finally, the promising future of federated learning in other facets of the smart grid is discussed, including electric vehicles, distributed generation/consumption and integrated energy systems.

cs.LG↗
arXiv2021

Virtual domination of 3-manifolds II

For any closed oriented 3-manifold $M$ with positive simplicial volume and any closed oriented 3-manifold $N$, we prove that there exists a finite cover $M'$ of $M$ that admits a degree-1 map $f:M'\to M$, i.e. M virtually 1-dominates N. This result generalizes previous virtual domination results with closed hyperbolic domain to more general domains.

math.GT↗
arXiv2021

Distributionally Robust Frequency Constrained Scheduling for an Integrated Electricity-Gas System

Power systems are shifted from conventional bulk generation toward renewable generation. This trend leads to the frequency security problem due to the decline of system inertia. On the other hand, natural gas-fired units are frequently scheduled to provide operational flexibility due to their fast adjustment ability. The interdependence between power and natural gas systems is thus intensified. In this paper, we study the frequency constrained scheduling problem from the perspective of an integrated electricity-gas system under variable wind power. We propose a distributionally robust (DR) chance constrained optimization model to co-optimize the unit commitment and virtual inertia provision from wind farm systems. This model incorporates both frequency constraints and natural gas system (NGS) operational constraints and addresses the wind power uncertainty by designing DR joint chance constraints. We show that this model admits a mixed-integer second-order cone programming. Case studies demonstrate that the proposed approach can provide a highly reliable and computationally efficient solution and show the importance of incorporating NGS operational constraints in the frequency constrained scheduling problem.

math.OC↗
arXiv2021

A Scenario-oriented Approach to Multi-period Energy-Reserve Joint Procurement and Pricing

In [1], a single-period co-optimization model of energy and reserve is considered to better illustrate the properties of the co-optimization model and the associated market mechanism. To make the discussion more general, in this paper, the single-period co-optimization model (II) in [1] will be extended into a one-shot multi-period co-optimization model, where the coupling of ramping and reserve in multi-period operations will be considered. Also, the multi-period pricing approach and settlement process associated with the multi-period co-optimization model will be discussed. In addition, in case study, the average system total costs of the proposed multi-period model and the multi-period traditional model are compared, and the effectiveness of some market properties in multi-period operation is tested.

eess.SY↗
arXiv2021

MMOCR: A Comprehensive Toolbox for Text Detection, Recognition and Understanding

We present MMOCR-an open-source toolbox which provides a comprehensive pipeline for text detection and recognition, as well as their downstream tasks such as named entity recognition and key information extraction. MMOCR implements 14 state-of-the-art algorithms, which is significantly more than all the existing open-source OCR projects we are aware of to date. To facilitate future research and industrial applications of text recognition-related problems, we also provide a large number of trained models and detailed benchmarks to give insights into the performance of text detection, recognition and understanding. MMOCR is publicly released at https://github.com/open-mmlab/mmocr.

cs.CV↗
arXiv2021

Two-stage Planning for Electricity-Gas Coupled Integrated Energy System with CCUS Considering Carbon Tax and Price Uncertainty

In this article, we propose two-stage planning models for Electricity-Gas Coupled Integrated Energy System (EGC-IES), in which traditional thermal power plants (TTPPs) are considered to be retrofitted into carbon capture power plants (CCPPs), with power to gas (PtG) coupling CCPPs to gas system. The sizing and siting of carbon capture, utilisation and storage (CCUS)/PtG facilities, as well as the operation cost of TTPPs/CCPPs/gas sources/PtG, are all considered in the proposed model, including penalty on carbon emissions and revenue of CCUS. With changing policy on climate change and carbon emission regulation, the uncertainties of carbon price and carbon tax are also analysed and considered in the proposed planning model. The stochastic planning, and robust planning methods are introduced to verify mutually through economic and carbon indices. The proposed methods' effectiveness in reducing carbon emissions, increasing profit of CCUS from EGC-IES are demonstrated through various cases and discussions.

eess.SY↗
arXiv2021

A Scenario-oriented Approach to Energy-Reserve Joint Procurement and Pricing

We consider some crucial problems related to the secure and reliable operation of power systems with high renewable penetrations: how much reserve should we procure, how should reserve resources distribute among different locations, and how should we price reserve and charge uncertainty sources. These issues have so far been largely addressed empirically. In this paper, we first develop a scenario-oriented energy-reserve co-optimization model, which directly connects reserve procurement with possible outages and load/renewable power fluctuations without the need for empirical reserve requirements. Accordingly, reserve can be optimally procured system-wide to handle all possible future uncertainties with the minimum expected system total cost. Based on the proposed model, marginal pricing approaches are developed for energy and reserve, respectively. Locational uniform pricing is established for energy, and the similar property is also established for the combination of reserve and re-dispatch. In addition, properties of cost recovery for generators and revenue adequacy for the system operator are also proven.

eess.SY↗
arXiv2021

An Integration and Operation Framework of Geothermal Heat Pumps in Distribution Networks

The application of the energy-efficient thermal and energy management mechanism in Geothermal Heat Pumps (GHPs) is indispensable to reduce the overall energy consumption and carbon emission across the building sector. Besides, in the Virtual Power Plant (VPP) system, the demand response of clustered GHP systems can improve the operating flexibility of the power grid. This paper presents an integration and operation framework of GHPs in the distribution network, applying a layered communication and optimization method to coordinate multiple clustered GHPs in a community. In the proposed hierarchical operation scheme, the operator of regional GHPs collects the thermal zone information and the disturbance prediction of buildings in a short time granularity, predicts the energy demand, and transmits the information to an aggregator. Using a novel linearized optimal power flow model, the aggregator coordinates and aggregates load resources of GHP systems in the distribution network. In this way, GHP systems with thermal and energy management mechanisms can be applied to achieve the demand response in the VPP and offer more energy flexibility to the community.

eess.SY↗
arXiv2021

A Multivariate Density Forecast Approach for Online Power System Security Assessment

A multivariate density forecast model based on deep learning is designed in this paper to forecast the joint cumulative distribution functions (JCDFs) of multiple security margins in power systems. Differing from existing multivariate density forecast models, the proposed method requires no a priori hypotheses on the distribution of forecasting targets. In addition, based on the universal approximation capability of neural networks, the value domain of the proposed approach has been proven to include all continuous JCDFs. The forecasted JCDF is further employed to calculate the deterministic security assessment index evaluating the security level of future power system operations. Numerical tests verify the superiority of the proposed method over current multivariate density forecast models. The deterministic security assessment index is demonstrated to be more informative for operators than security margins as well.

eess.SY↗
arXiv2021

A Conditional Value-at-Risk Based Planning Model for Integrated Energy System with Energy Storage and Renewables

Owing to the potential higher energy supply efficiency and operation flexibility, integrated energy system (IES), which usually includes electric power, gas and heating/cooling systems, is considered as one of the primary forms of energy carrier in the future. However, with the increasing complexity of multiple energy devices and systems integration, IES planning is facing a significant challenge in terms of risk assessment. To this end, an energy hub (EH) planning model considering renewable energy sources (RES) and energy storage system (ESS) integration is proposed in this paper, in which the risk is measured by Conditional Value-at-Risk (CVaR). The proposed IES planning model includes two stages: 1) investment planning on equipment types and capacity (e.g., energy converters, distributed RES and ESS) and 2) optimizing the potential risk loss in operation scenarios along with confidence level and risk preference. The problem solving is accelerated by Benders Decomposition and Improved Backward Scenario Reduction Method. The numerical results illustrate the effectiveness of proposed method in balancing the potential operation risk and investment cost. Moreover, the effectiveness of reducing potential operation risk by introducing ESS and RES are also verified.

eess.SY↗
arXiv2021

End-to-End Object Detection with Fully Convolutional Network

Mainstream object detectors based on the fully convolutional network has achieved impressive performance. While most of them still need a hand-designed non-maximum suppression (NMS) post-processing, which impedes fully end-to-end training. In this paper, we give the analysis of discarding NMS, where the results reveal that a proper label assignment plays a crucial role. To this end, for fully convolutional detectors, we introduce a Prediction-aware One-To-One (POTO) label assignment for classification to enable end-to-end detection, which obtains comparable performance with NMS. Besides, a simple 3D Max Filtering (3DMF) is proposed to utilize the multi-scale features and improve the discriminability of convolutions in the local region. With these techniques, our end-to-end framework achieves competitive performance against many state-of-the-art detectors with NMS on COCO and CrowdHuman datasets. The code is available at https://github.com/Megvii-BaseDetection/DeFCN .

cs.CV↗
arXiv2021

Spatial Dual-Modality Graph Reasoning for Key Information Extraction

Key information extraction from document images is of paramount importance in office automation. Conventional template matching based approaches fail to generalize well to document images of unseen templates, and are not robust against text recognition errors. In this paper, we propose an end-to-end Spatial Dual-Modality Graph Reasoning method (SDMG-R) to extract key information from unstructured document images. We model document images as dual-modality graphs, nodes of which encode both the visual and textual features of detected text regions, and edges of which represent the spatial relations between neighboring text regions. The key information extraction is solved by iteratively propagating messages along graph edges and reasoning the categories of graph nodes. In order to roundly evaluate our proposed method as well as boost the future research, we release a new dataset named WildReceipt, which is collected and annotated tailored for the evaluation of key information extraction from document images of unseen templates in the wild. It contains 25 key information categories, a total of about 69000 text boxes, and is about 2 times larger than the existing public datasets. Extensive experiments validate that all information including visual features, textual features and spatial relations can benefit key information extraction. It has been shown that SDMG-R can effectively extract key information from document images of unseen templates, and obtain new state-of-the-art results on the recent popular benchmark SROIE and our WildReceipt. Our code and dataset will be publicly released.

cs.CV↗
arXiv2021

All finitely generated 3-manifold groups are Grothendieck rigid

In this paper, we prove that all finitely generated 3-manifold groups are Grothendieck rigid. More precisely, for any finitely generated 3-manifold group $G$ and any finitely generated proper subgroup $H<G$, we prove that the inclusion induced homomorphism $\widehat{i}:\widehat{H}\to \widehat{G}$ on profinite completions is not an isomorphism.

math.GT↗
arXiv2020

A Data-Driven Warm Start Approach for Convex Relaxation in Optimal Gas Flow

In this letter, we propose a data-driven warm start approach, empowered by artificial neural networks, to boost the efficiency of convex relaxations in optimal gas flow. Case studies show that this approach significantly decreases the number of iterations for the convex-concave procedure algorithm, and optimality and feasibility of the solution can still be guaranteed.

math.OC↗
arXiv2020

Rethinking Learnable Tree Filter for Generic Feature Transform

The Learnable Tree Filter presents a remarkable approach to model structure-preserving relations for semantic segmentation. Nevertheless, the intrinsic geometric constraint forces it to focus on the regions with close spatial distance, hindering the effective long-range interactions. To relax the geometric constraint, we give the analysis by reformulating it as a Markov Random Field and introduce a learnable unary term. Besides, we propose a learnable spanning tree algorithm to replace the original non-differentiable one, which further improves the flexibility and robustness. With the above improvements, our method can better capture long-range dependencies and preserve structural details with linear complexity, which is extended to several vision tasks for more generic feature transform. Extensive experiments on object detection/instance segmentation demonstrate the consistent improvements over the original version. For semantic segmentation, we achieve leading performance (82.1% mIoU) on the Cityscapes benchmark without bells-and-whistles. Code is available at https://github.com/StevenGrove/LearnableTreeFilterV2.

cs.CV↗
arXiv2020

Fine-Grained Dynamic Head for Object Detection

The Feature Pyramid Network (FPN) presents a remarkable approach to alleviate the scale variance in object representation by performing instance-level assignments. Nevertheless, this strategy ignores the distinct characteristics of different sub-regions in an instance. To this end, we propose a fine-grained dynamic head to conditionally select a pixel-level combination of FPN features from different scales for each instance, which further releases the ability of multi-scale feature representation. Moreover, we design a spatial gate with the new activation function to reduce computational complexity dramatically through spatially sparse convolutions. Extensive experiments demonstrate the effectiveness and efficiency of the proposed method on several state-of-the-art detection benchmarks. Code is available at https://github.com/StevenGrove/DynamicHead.

cs.CV↗