skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Graphical models for optimal power flow

Abstract

Optimal power flow (OPF) is the central optimization problem in electric power grids. Although solved routinely in the course of power grid operations, it is known to be strongly NP-hard in general, and weakly NP-hard over tree networks. In this paper, we formulate the optimal power flow problem over tree networks as an inference problem over a tree-structured graphical model where the nodal variables are low-dimensional vectors. We adapt the standard dynamic programming algorithm for inference over a tree-structured graphical model to the OPF problem. Combining this with an interval discretization of the nodal variables, we develop an approximation algorithm for the OPF problem. Further, we use techniques from constraint programming (CP) to perform interval computations and adaptive bound propagation to obtain practically efficient algorithms. Compared to previous algorithms that solve OPF with optimality guarantees using convex relaxations, our approach is able to work for arbitrary tree-structured distribution networks and handle mixed-integer optimization problems. Further, it can be implemented in a distributed message-passing fashion that is scalable and is suitable for “smart grid” applications like control of distributed energy resources. In conclusion, numerical evaluations on several benchmark networks show that practical OPF problems can be solved effectively using this approach.

Authors:
ORCiD logo [1];  [2];  [3];  [2];  [2]
  1. California Inst. of Technology (CalTech), Pasadena, CA (United States)
  2. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
  3. Univ. of Michigan, Ann Arbor, MI (United States)
Publication Date:
Research Org.:
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE Office of Electricity Delivery and Energy Reliability (OE); USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1345156
Report Number(s):
LA-UR-16-24896
Journal ID: ISSN 1383-7133
Grant/Contract Number:
AC52-06NA25396
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
Constraints
Additional Journal Information:
Journal Volume: 22; Journal Issue: 1; Journal ID: ISSN 1383-7133
Publisher:
Springer
Country of Publication:
United States
Language:
English
Subject:
24 POWER TRANSMISSION AND DISTRIBUTION; 97 MATHEMATICS AND COMPUTING; Computer Science; Energy Sciences; Information Science

Citation Formats

Dvijotham, Krishnamurthy, Chertkov, Michael, Van Hentenryck, Pascal, Vuffray, Marc, and Misra, Sidhant. Graphical models for optimal power flow. United States: N. p., 2016. Web. doi:10.1007/s10601-016-9253-y.
Dvijotham, Krishnamurthy, Chertkov, Michael, Van Hentenryck, Pascal, Vuffray, Marc, & Misra, Sidhant. Graphical models for optimal power flow. United States. doi:10.1007/s10601-016-9253-y.
Dvijotham, Krishnamurthy, Chertkov, Michael, Van Hentenryck, Pascal, Vuffray, Marc, and Misra, Sidhant. 2016. "Graphical models for optimal power flow". United States. doi:10.1007/s10601-016-9253-y. https://www.osti.gov/servlets/purl/1345156.
@article{osti_1345156,
title = {Graphical models for optimal power flow},
author = {Dvijotham, Krishnamurthy and Chertkov, Michael and Van Hentenryck, Pascal and Vuffray, Marc and Misra, Sidhant},
abstractNote = {Optimal power flow (OPF) is the central optimization problem in electric power grids. Although solved routinely in the course of power grid operations, it is known to be strongly NP-hard in general, and weakly NP-hard over tree networks. In this paper, we formulate the optimal power flow problem over tree networks as an inference problem over a tree-structured graphical model where the nodal variables are low-dimensional vectors. We adapt the standard dynamic programming algorithm for inference over a tree-structured graphical model to the OPF problem. Combining this with an interval discretization of the nodal variables, we develop an approximation algorithm for the OPF problem. Further, we use techniques from constraint programming (CP) to perform interval computations and adaptive bound propagation to obtain practically efficient algorithms. Compared to previous algorithms that solve OPF with optimality guarantees using convex relaxations, our approach is able to work for arbitrary tree-structured distribution networks and handle mixed-integer optimization problems. Further, it can be implemented in a distributed message-passing fashion that is scalable and is suitable for “smart grid” applications like control of distributed energy resources. In conclusion, numerical evaluations on several benchmark networks show that practical OPF problems can be solved effectively using this approach.},
doi = {10.1007/s10601-016-9253-y},
journal = {Constraints},
number = 1,
volume = 22,
place = {United States},
year = 2016,
month = 9
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Save / Share:
  • Studies of effects of load modeling in optimal power flow studies using minimum cost and minimum loss objective reveal that a main disadvantage of cost minimization is the reduction of the objective via a reduction in the power demand. This inevitably results in lowering the total revenue and in most cases, reducing net revenue as well. An alternative approach for incorporating load models in security-constrained OPF (SCOPF) studies apparently avoids reducing the total power demand for the intact system, but reduces the voltages. A study of the behavior of conventional OPF solutions in the presence of loads not controlled bymore » ULTC's shows that this result in a reducing the total power demand for the intact system. In this paper, the authors propose an objective that avoids the tendency to lower the total power demand, total revenue and net revenue, for OPF neglecting contingencies (normal OPF), as well as for security-constrained OPF. The minimum cost objective is modified by subtracting the total power demand from the total fuel cost. This is equivalent to maximizing the net revenue.« less
  • Load-flow analysis utilizing the mathematical programming method is often useless for large-scale systems due to its long computing time. A new dc load-flow optimization method is proposed in which power system components are represented by electrical network elements and they are connected to each other according to the given power system configuration. Matrix sparcity is utilized for high-speed network calculation.
  • In general, the Han-Powell algorithm is evaluated as the fastest and the most reliable method for small nonlinear programming problems. However, it has one serious disadvantage when applied to a large scale problem such as an optimal power flow. This disadvantage stems from its use of non-sparse approximations to certain Hessian matrices. We propose a modified Newton method to eliminate this disadvantage. Sparsity of the Hessian is maintained by this method, and it can be modified to be a non-negative definite matrix in accordance with simple procedures. This can be implemented using quadratic approximations of power flow equations. Numerical testsmore » on a real system show the validity of the proposed method.« less
  • A simple and accurate dispatch method to economically schedule large power system generations is presented and tested. The proposed method is based on a piecewise active reactive network loss model and is more accurate than the classical approach for different system load levels. The method can be superior to optimal load flow methods with respect to convergence, computational time and storage.
  • Four modifications of a flow model for selecting the optimal parameters with various degrees of detail taking into account the dynamic and stochastic characteristics of functioning of a water resource system were examined. The interrelation between their solutions was established and experimental calculations were made, which showed the advisability of using all four modifications for successive refinement of the parameters being selected.