arXiv subjects

Guangyi Liu

Publications and source records attributed to Guangyi Liu.

At least 145 records · Page 8Linked to original sources

arXiv2019

Communication-Constrained Routing and Traffic Control for Autonomous Vehicles

Autonomous vehicles (AV) is an advanced technology that can bring convenience, improve the road-network throughput, and reduce traffic accidents. To enable higher levels of automation (LoA), massive amounts of sensory data need to be uploaded to the network for processing, and then, maneuvering decisions must be returned to the AV. Furthermore, passengers might have a higher transmission rate demands for various data-hungry and delay-sensitive applications.

eess.SY↗

arXiv2019

A Layered Architecture for Active Perception: Image Classification using Deep Reinforcement Learning

We propose a planning and perception mechanism for a robot (agent), that can only observe the underlying environment partially, in order to solve an image classification problem. A three-layer architecture is suggested that consists of a meta-layer that decides the intermediate goals, an action-layer that selects local actions as the agent navigates towards a goal, and a classification-layer that evaluates the reward and makes a prediction. We design and implement these layers using deep reinforcement learning. A generalized policy gradient algorithm is utilized to learn the parameters of these layers to maximize the expected reward. Our proposed methodology is tested on the MNIST dataset of handwritten digits, which provides us with a level of explainability while interpreting the agent's intermediate goals and course of action.

cs.LG↗

arXiv2019

Unbalanced Multi-Phase Distribution Grid Topology Estimation and Bus Phase Identification

There is an increasing need for monitoring and controlling uncertainties brought by distributed energy resources in distribution grids. For such goal, accurate multi-phase topology is the basis for correlating measurements in unbalanced distribution networks. Unfortunately, such topology knowledge is often unavailable due to limited investment, especially for \revv{low-voltage} distribution grids. Also, the bus phase labeling information is inaccurate due to human errors or outdated records. For this challenge, this paper utilizes smart meter data for an information-theoretic approach to learn the topology of distribution grids. Specifically, multi-phase unbalanced systems are converted into symmetrical components, namely positive, negative, and zero sequences. Then, this paper proves that the Chow-Liu algorithm finds the topology by utilizing power flow equations and the conditional independence relationships implied by the radial multi-phase structure of distribution grids with the presence of incorrect bus phase labels. At last, by utilizing Carson's equation, this paper proves that the bus phase connection can be correctly identified using voltage measurements. For validation, IEEE systems are simulated using three real data sets. The simulation results demonstrate that the algorithm is highly accurate for finding multi-phase topology even with strong load unbalancing condition and DERs. This ensures close monitoring and controlling DERs in distribution grids.

eess.SY↗

arXiv2019

Enhancement of Power Equipment Management Using Knowledge Graph

Accurate retrieval of the power equipment information plays an important role in guiding the full-lifecycle management of power system assets. Because of data duplication, database decentralization, weak data relations, and sluggish data updates, the power asset management system eager to adopt a new strategy to avoid the information losses, bias, and improve the data storage efficiency and extraction process. Knowledge graph has been widely developed in large part owing to its schema-less nature. It enables the knowledge graph to grow seamlessly and allows new relations addition and entities insertion when needed. This study proposes an approach for constructing power equipment knowledge graph by merging existing multi-source heterogeneous power equipment related data. A graph-search method to illustrate exhaustive results to the desired information based on the constructed knowledge graph is proposed. A case of a 500 kV station example is then demonstrated to show relevant search results and to explain that the knowledge graph can improve the efficiency of power equipment management.

cs.AI↗

arXiv2019

A High-Performance Energy Management System based on Evolving Graph

As the fast growth and large integration of distributed generation, renewable energy resource, energy storage system and load response, the modern power system operation becomes much more complicated with increasing uncertainties and frequent changes. Increased operation risks are introduced to the existing commercial Energy Management System (EMS), due to its limited computational capability. In this paper, a high-performance EMS analysis framework based on the evolving graph is developed. A power grid is first modeled as an evolving graph and then the power system dynamic analysis applications, like network topology processing (NTP), state estimation (SE), power flow (PF), and contingency analysis (CA), are efficiently implemented on the system evolving graph to build a high-performance EMS analysis framework. Its computation performance is field tested using a 2749-bus power system in Sichuan, China. The results illustrate that the proposed EMS remarkably speeds up the computation performance and reaches the goal of real-time power system analysis.

cs.DC↗

arXiv2019

Fast Grid Splitting Detection for N-1 Contingency Analysis by Graph Computing

In this study, a graph-computing based grid splitting detection algorithm is proposed for contingency analysis in a graph-based EMS (Energy Management System). The graph model of a power system is established by storing its bus-branch information into the corresponding vertex objects and edge objects of the graph database. Numerical comparison to an up-to-date serial computing algorithm is also investigated. Online tests on a real power system of China State Grid with 2752 buses and 3290 branches show that a 6 times speedup can be achieved, which lays a good foundation for advanced contingency analysis.

cs.DC↗

arXiv2019

Optimal Battery Energy Storage Placement for Transient Voltage Stability Enhancement

A placement problem for multiple Battery Energy Storage System (BESS) units is formulated towards power system transient voltage stability enhancement in this paper. The problem is solved by the Cross-Entropy (CE) optimization method. A simulation-based approach is adopted to incorporate higher-order dynamics and nonlinearities of generators and loads. The objective is to maximize the voltage stability index, which is set up based on certain grid-codes. Formulations of the optimization problem are then discussed. Finally, the proposed approach is implemented in MATLAB/DIgSILENT and tested on the New England 39-Bus system. Results indicate that installing BESS units at the optimized location can alleviate transient voltage instability issue compared with the original system with no BESS. The CE placement algorithm is also compared with the classic PSO (Particle Swarm Optimization) method, and its superiority is demonstrated in terms of fewer iterations for convergence with better solution qualities.

math.OC↗

arXiv2019

Graph Computing based Fast Screening in Contingency Analysis

During last decades, contingency analysis has been facing challenges from significant load demand increase and high penetrations of intermittent renewable energy, fluctuant responsive loads and non-linear power electronic interfaces. It requires an advanced approach for high-performance contingency analysis as a safeguard of the power system operation. In this paper, a graph-based method is employed for N-1 contingency analysis (CA) fast screening. At first, bi-directional breadth-first search (BFS) is proposed and adopted on graph model to detect the potential shedding component in contingency analysis. It implements hierarchical parallelism of the graph traverse and speedup its process. Then, the idea of evolving graph is introduced in this paper to improve computation performance. For each contingency scenario, N-1 contingency graph quickly derives from system graph in basic status, and parallelly analyzes each contingency scenario using graph computing. The efficiency and effectiveness of the proposed approach have been tested and verified by IEEE 118-bus system and a practical case SC 2645-bus system.

cs.DC↗

arXiv2019

Location Planning of Fast Charging Station considering its Impact on the Power Grid Assets

Under the ambition of boosting Plug-in Electric Vehicle (PEV) charging speed to a level comparable to the traditional refueling, Fast Charging Station (FCS) has been integrated into power distribution system. The location planning of FCS must allow for satisfactory charging service for PEV users as well as mitigate the detrimental effects on power grid caused by uncertainty and impulsiveness of charging demand. This paper proposed a location planning model for FCS, taking into account its impacts on the critical power grid assets. The multi-objective planning model simultaneously considered the role of FCS in the electricity and transportation sectors. This planning model is solved by the cross-entropy (CE) method. The validity and effectiveness of the CE approach have been demonstrated on a synthetic coupled network.

math.OC↗

arXiv2019

Substation One-Line Diagram Automatic Generation and Visualization

In Energy Management System (EMS) applications and many other off-line planning and study tools, one-line diagram (OLND) of the whole system and stations is a straightforward view for planners and operators to design, monitor, analyze, and control the power system. Large-scale power system OLND is usually manually developed and maintained. The work is tedious, time-consuming and ease to make mistake. Meanwhile, the manually created diagrams are hard to be shared among the on-line and off-line systems. To save the time and efforts to draw and maintain OLNDs, and provide the capability to share the OLNDs, a tool to automatically develop substation based upon Common Information Model (CIM) standard is needed. Currently, there is no standard rule to draw the substation OLND. Besides, the substation layouts can be altered from the typical formats in textbooks based on factors of economy, efficiency, engineering practice, etc. This paper presents a tool on substation OLND automatic generation and visualization. This tool takes the substation CIM/E model as input, then automatically computes the coordinates of all components and generates the substation OLND based on its components attributes and connectivity relations. Evaluation of the proposed approach is presented using a real provincial power system. Over 95\% of substation OLNDs are decently presented and the rest are corner cases, needing extra effort to do specific reconfiguration.

cs.OH↗

arXiv2019

A Graph Computation based Sequential Power Flow Calculation for Large-Scale ACDC Systems

This paper proposes a graph computation based sequential power flow calculation method for Line Commutated Converter (LCC) based large-scale AC/DC systems to achieve a high computing performance. Based on the graph theory, the complex AC/DC system is first converted to a graph model and stored in a graph database. Then, the hybrid system is divided into several isolated areas with graph partition algorithm by decoupling AC and DC networks. Thus, the power flow analysis can be executed in parallel for each independent area with the new selected slack buses. Furthermore, for each area, the node-based parallel computing (NPC) and hierarchical parallel computing (HPC) used in graph computation are employed to speed up fast decoupled power flow (FDPF). Comprehensive case studies on the IEEE 300-bus, polished South Carolina 12,000-bus system and a China 11,119-bus system are performed to demonstrate the accuracy and efficiency of the proposed method

cs.DC↗

arXiv2019

Graph Computing based Distributed Fast Decoupled Power Flow Analysis

Power flow analysis plays a fundamental and critical role in the energy management system (EMS). It is required to well accommodate large and complex power system. To achieve a high performance and accurate power flow analysis, a graph computing based distributed power flow analysis approach is proposed in this paper. Firstly, a power system network is divided into multiple areas. Slack buses are selected for each area and, at each SCADA sampling period, the inter-area transmission line power flows are equivalently allocated as extra load injections to corresponding buses. Then, the system network is converted into multiple independent areas. In this way, the power flow analysis could be conducted in parallel for each area and the solved system states could be guaranteed without compromise of accuracy. Besides, for each area, graph computing based fast decoupled power flow (FDPF) is employed to quickly analyze system states. IEEE 118-bus system and MP 10790-bus system are employed to verify the results accuracy and present the promising computation performance of the proposed approach.

cs.DC↗

arXiv2019

Simplify Power Flow Calculation Using Terminal Circuit and PMU Measurements

Power flow calculation methods have been developed in decades using power injections and Newton-Raphson method. The nonlinear characteristics of the power flow to the bus voltage require Jacobian matrix reformation and refactorization in each iteration. Power network is composed by resistors, reactors, and capacitors which is a linear circuit when investigating the node voltages with the node current injections. To take the advantage of the linearity, this paper proposed to use terminal circuit model and PMU voltage phase angle measurements to simplify power flow calculation. When updating current injections at the righthand side of power flow equations and using PMU voltage phase angle measurements representing PV buses voltage phase angle, the Jacobian matrix is constant during the iteration. The simplification reduces the computational efforts and improves the computation efficiency. The proposed method is tested on the IEEE 14-bus and IEEE 118-bus standard systems. The results are validated by traditional power flow solution and the computation efficiency is demonstrated.

eess.SP↗

arXiv2018

Graph Based Power Flow Calculation for Energy Management System

Power flow calculation in EMS is required to accommodate a large and complex power system. To achieve a faster than real-time calculation, a graph based power flow calculation is proposed in this paper. Graph database and graph computing advantages in power system calculations are presented. A linear solver for power flow application is formulated and decomposed in nodal parallelism and hierarchical parallelism to fully utilize graph parallel computing capability. Comparison of the algorithm with traditional sequential programs shows significant benefits on computation efficiency. Case studies on practical large-scale systems provide supporting evidence that the new algorithm is promising for online computing for EMS.

eess.SP↗

arXiv2018

Power Market Price Forecasting via Deep Learning

A study on power market price forecasting by deep learning is presented. As one of the most successful deep learning frameworks, the LSTM (Long short-term memory) neural network is utilized. The hourly prices data from the New England and PJM day-ahead markets are used in this study. First, a LSTM network is formulated and trained. Then the raw input and output data are preprocessed by unit scaling, and the trained network is tested on the real price data under different input lengths, forecasting horizons and data sizes. Its performance is also compared with other existing methods. The forecasted results demonstrate that, the LSTM deep neural network can outperform the others under different application settings in this problem.

cs.LG↗

arXiv2018

Urban MV and LV Distribution Grid Topology Estimation via Group Lasso

The increasing penetration of distributed energy resources poses numerous reliability issues to the urban distribution grid. The topology estimation is a critical step to ensure the robustness of distribution grid operation. However, the bus connectivity and grid topology estimation are usually hard in distribution grids. For example, it is technically challenging and costly to monitor the bus connectivity in urban grids, e.g., underground lines. It is also inappropriate to use the radial topology assumption exclusively because the grids of metropolitan cities and regions with dense loads could be with many mesh structures. To resolve these drawbacks, we propose a data-driven topology estimation method for MV and LV distribution grids by only utilizing the historical smart meter measurements. Particularly, a probabilistic graphical model is utilized to capture the statistical dependencies amongst bus voltages. We prove that the bus connectivity and grid topology estimation problems, in radial and mesh structures, can be formulated as a linear regression with a least absolute shrinkage regularization on grouped variables (\textit{group lasso}). Simulations show highly accurate results in eight MV and LV distribution networks at different sizes and 22 topology configurations using PG\&E residential smart meter data.

stat.ML↗

arXiv2018

Convolutional Graph Auto-encoder: A Deep Generative Neural Architecture for Probabilistic Spatio-temporal Solar Irradiance Forecasting

Machine Learning on graph-structured data is an important and omnipresent task for a vast variety of applications including anomaly detection and dynamic network analysis. In this paper, a deep generative model is introduced to capture continuous probability densities corresponding to the nodes of an arbitrary graph. In contrast to all learning formulations in the area of discriminative pattern recognition, we propose a scalable generative optimization/algorithm theoretically proved to capture distributions at the nodes of a graph. Our model is able to generate samples from the probability densities learned at each node. This probabilistic data generation model, i.e. convolutional graph auto-encoder (CGAE), is devised based on the localized first-order approximation of spectral graph convolutions, deep learning, and the variational Bayesian inference. We apply our CGAE to a new problem, the spatio-temporal probabilistic solar irradiance prediction. Multiple solar radiation measurement sites in a wide area in northern states of the US are modeled as an undirected graph. Using our proposed model, the distribution of future irradiance given historical radiation observations is estimated for every site/node. Numerical results on the National Solar Radiation Database show state-of-the-art performance for probabilistic radiation prediction on geographically distributed irradiance data in terms of reliability, sharpness, and continuous ranked probability score.

cs.LG↗

arXiv2018

Power Flow Analysis Using Graph based Combination of Iterative Methods and Vertex Contraction Approach

Compared with relational database (RDB), graph database (GDB) is a more intuitive expression of the real world. Each node in the GDB is a both storage and logic unit. Since it is connected to its neighboring nodes through edges, and its neighboring information could be easily obtained in one-step graph traversal. It is able to conduct local computation independently and all nodes can do their local work in parallel. Then the whole system can be maximally analyzed and assessed in parallel to largely improve the computation performance without sacrificing the precision of final results. This paper firstly introduces graph database, power system graph modeling and potential graph computing applications in power systems. Two iterative methods based on graph database and PageRank are presented and their convergence are discussed. Vertex contraction is proposed to improve the performance by eliminating zero-impedance branch. A combination of the two iterative methods is proposed to make use of their advantages. Testing results based on a provincial 1425-bus system demonstrate that the proposed comprehensive approach is a good candidate for power flow analysis.

cs.DC↗