Parallel Primal-Dual Solver for Power Grid Unit Commitment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Solving large-scale security constrained unit commitment in power grids is computationally intensive and time-consuming due to its complexity and huge dimensions.
Innovation Solution
A power management system employing a parallel asynchronous collaborative primal dual solver that uses multiple primal and dual solvers with different initial conditions to generate resource allocation schedules by comparing upper and lower bounds, reducing computational time and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional unit commitment methods are used to solve security constrained unit commitment, then solution accuracy is maintained, but computational time and complexity increase significantly
Solution Approach 1:
The patent segments the security constrained unit commitment problem into multiple independent scenarios (base case and N-1 contingency scenarios). Each scenario is solved separately using parallel processing, allowing the complex overall problem to be divided into manageable sub-problems that can be computed simultaneously, thereby reducing total computational time while maintaining solution accuracy for each scenario.
Solution Approach 2:
The patent introduces a parallel processing dimension by implementing multiple primal and dual solvers that operate simultaneously on different scenarios. This transforms the traditionally sequential solution approach into a parallel architecture, adding a temporal dimension of concurrency that reduces overall computational time without sacrificing the precision of individual scenario solutions.
2Speed
If multiple solvers with different initial conditions are employed, then convergence speed is improved, but system complexity increases
Solution Approach 1:
The patent applies preliminary action by initializing multiple primal and dual solvers with different initial conditions before solving the unit commitment problem. This pre-positioning of diverse starting points allows the solvers to converge from different directions, increasing the likelihood of finding optimal solutions faster. The complexity management is achieved by orchestrating these pre-initialized solvers in a structured parallel framework.
Solution Approach 2:
The patent changes key parameters of the solving system by varying initial conditions across multiple solvers. By modifying initialization parameters (such as starting power outputs, commitment states) rather than the problem structure itself, the system achieves faster convergence through parameter diversity while maintaining manageable system complexity through standardized solver architectures.
3Reliability
If security constraints are included in unit commitment, then grid reliability is improved, but computational difficulty increases
Solution Approach 1:
The patent segments security constraints into separate N-1 contingency scenarios (loss of individual transmission lines or generators) that are solved in parallel. Each scenario incorporates specific security constraints relevant to that contingency, allowing comprehensive security analysis without overwhelming computational complexity in a single monolithic problem formulation.
Solution Approach 2:
The patent addresses security constraint complexity by adding a scenario dimension - solving the base case and multiple contingency scenarios as separate but parallel problems. This dimensional transformation allows each sub-problem to have manageable complexity while collectively providing comprehensive security assessment, as each scenario's constraints are processed independently and simultaneously.
Data Source
AI summary
Embodiments of the disclosure includes operating a power grid, including: generating, by a power management system of the power grid, a power grid resource allocation profile indicative of an operation of the power grid constrained by operational information of the power grid; generating a difference between a value of upper bounds from a plurality of obtained convergence paths and a value of lower bounds from the obtained convergence paths, the obtained convergence paths being based on a plurality of different initial conditions for the generated power grid resource allocation profile; and generating a resource allocation schedule for power grid resources operating within the power grid if the generated difference is smaller than a pre-determined threshold, the resource allocation schedule corresponding to a convergence path associated with the value of the upper bounds, the resource allocation schedule being configured to be received at the power grid resources.