Emergency Load-Shedding Optimization Using Proliferation-Reduction Evolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing evolutionary algorithms for emergency load-shedding optimization face challenges in balancing global convergence and optimization speed, with large population sizes increasing computational burden and small populations leading to poor convergence and quality of optimized schemes, while data-driven surrogate models struggle to maintain accuracy in dynamic power systems.
Innovation Solution
The method employs a proliferation strategy to generate a large number of candidate schemes using different evolution search operators based on scheme distribution characteristics and a reduction strategy that pre-screens schemes using a surrogate model and validates them with time-domain simulation, updating the model with evaluation results to improve convergence and accuracy without increasing time-domain simulation frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a large candidate scheme population is used in evolutionary algorithms, then global convergence is improved, but optimization speed deteriorates due to increased computational burden from time-domain simulation
Solution Approach 1:
The patent divides the evaluation process into two segments: time-domain simulation evaluation and surrogate model evaluation. The population is evaluated through time-domain simulation only once initially, then a surrogate model is trained to rapidly evaluate subsequent candidate schemes, replacing the need for repeated time-domain simulations and resolving the contradiction between comprehensive evaluation and computational burden
Solution Approach 2:
The patent introduces a surrogate model as an intermediary between the evolutionary algorithm and time-domain simulation. This surrogate model, trained on initial simulation data, acts as a mediator that provides accurate evaluation results without requiring repeated computationally intensive simulations, thus maintaining global convergence while dramatically improving optimization speed
2Productivity
If a small candidate scheme population is used, then optimization speed is improved, but global convergence deteriorates due to reduced diversity of candidate schemes
Solution Approach 1:
The patent performs preliminary time-domain simulation evaluation on an initial parent population before training the surrogate model. This preliminary action ensures that the surrogate model is trained on diverse, high-quality data from comprehensive simulations, enabling it to accurately guide subsequent optimization with smaller populations while maintaining global convergence capabilities
3Productivity
If data-driven surrogate models are used to replace time-domain simulation, then optimization efficiency is improved, but measurement precision deteriorates due to reduced accuracy in dynamic power systems
Solution Approach 1:
The patent implements a dynamic evaluation strategy where the surrogate model is trained on a parent population evaluated through time-domain simulation, then used to evaluate offspring schemes. This dynamic approach allows the system to leverage the speed of surrogate models while periodically grounding evaluations in accurate time-domain simulations, maintaining measurement precision while achieving high optimization efficiency
Data Source
AI summary
A method and system for optimizing power grid emergency load-shedding based on proliferation and reduction evolution. The method includes the steps of obtaining upper and lower limits data of allowed load-shedding amount of each load and boundary threshold data of transient security and stability constraint indexes of a power grid; and obtaining an optimal power grid emergency load-shedding scheme based on these data and an evolutionary optimization method, wherein the key of the evolutionary optimization method to work is proliferation and reduction evolution strategies. The proliferation strategy with multiple evolution search operators is proposed to generate many temporary candidate schemes. The reduction strategy of temporary candidate schemes includes two key steps, that is, pre-screening of based on a surrogate model and validation based on time-domain simulation.