Distributed Power Flow Optimization via Dual Decomposition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional methods for optimizing power flows in electric power networks are inefficient and do not effectively distribute computations, leading to sub-linear convergence and reliance on centralized solutions that do not utilize network topology.

Innovation Solution

The method employs a decomposition and coordination procedure that distributes the optimization problem into smaller, independent sub-problems using dualization of coupled constraints and semi-smooth equation theory for superlinear convergence, along with smoothing methods to ensure global convergence, even from initial parameters far from optimal.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If centralized optimization methods are used, then the optimization problem can be solved in a unified manner, but the computational efficiency and convergence speed are poor

Engineering Contradiction:
Improvecomputational efficiencyVSAvoidoptimization distribution complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent divides the centralized optimization problem into multiple distributed sub-problems, one for each bus in the power network. Each sub-problem is solved independently using local information and Lagrange multipliers, eliminating the need for a single centralized solver and significantly improving computational efficiency and convergence speed.

Inventive Principle:
Principle #1Segmentation

2Speed

If conventional subgradient method is used for dual decomposition, then the optimization problem can be distributed, but the convergence is sub-linear

Engineering Contradiction:
Improveconvergence speedVSAvoidcomputational efficiency
Core Design Contradiction:
SpeedVSProductivity

Solution Approach 1:

The patent changes the parameter update rule from the conventional subgradient method to a Newton-based method that uses second-order derivative information (Hessian matrix). This parameter change in the optimization algorithm achieves superlinear convergence while maintaining the distributed structure of the dual decomposition approach.

Inventive Principle:
Principle #35Parameter changes

3Reliability

If distributed optimization is implemented, then network topology can be exploited, but achieving global convergence from arbitrary initial parameters is difficult

Engineering Contradiction:
Improveglobal convergenceVSAvoidinitial parameter sensitivity
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The patent implements a feedback mechanism where Lagrange multipliers are updated based on the residual of the coupling constraints. This feedback loop continuously adjusts the distributed sub-problem solutions to ensure that the overall system constraints are satisfied, guaranteeing global convergence to the optimal solution regardless of initial parameter values.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2770599B1Method for optimizing power flow in electric power network
Publication Date: 2018.11.14 MITSUBISHI ELECTRIC CORP
  • EP2770599B1 patent drawingFigure 1
  • EP2770599B1 patent drawingFigure 2
  • EP2770599B1 patent drawingFigure 3

AI summary

Power flow in an electric power network is optimized by first decomposing an optimization problem into a set of disjoint parameterized optimization problems. The disjoint optimization problems are independent of each other, and the decomposition is based on dualized coupled constraints having corresponding multipliers. Each optimization problem is solved independently to obtain a corresponding solution, and a sensitivity of each solution to changes in the parameter. The parameters are updated using the corresponding solutions and the sensitivities, and iterated until convergence.