Genetic Algorithm for Power Distribution Network Restoration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for multi-layer restoration in electric power distribution networks face challenges in efficiently processing a large number of variables and effectively generating solutions due to the complexity of combinatorial and constrained optimization problems, particularly in achieving computational efficiency, maximizing restored loads, and maintaining radial network topology.
Innovation Solution
A method using a genetic algorithm and Reactive Tabu Search to determine network configurations for back-feed power restoration, which involves initializing chromosomes, evaluating their fitness based on power loss, voltage and current violations, switching operations, and unserved loads, and iteratively generating new candidate configurations to minimize the fitness value, ensuring radiality and avoiding energized loops.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional optimization methods are used for multi-layer restoration, then solution accuracy can be maintained, but computational efficiency deteriorates due to the large number of variables and combinatorial complexity
Solution Approach 1:
The patent segments the complex multi-layer restoration problem into multiple layers (first layer, second layer, third layer, etc.), where each layer represents a different level of restoration complexity. The algorithm processes restoration solutions layer by layer, starting from the simplest first-layer solutions and progressing to more complex multi-layer solutions. This segmentation reduces the computational burden by breaking down the large combinatorial optimization problem into smaller, more manageable sub-problems while maintaining solution accuracy through systematic exploration of all layers.
2Reliability
If the number of restoration switches is increased to maximize restored loads, then load restoration improves, but device complexity and difficulty of control increase
Solution Approach 1:
The patent performs preliminary actions by first identifying and executing first-layer restoration solutions before considering second-layer or third-layer solutions. The algorithm systematically evaluates whether first-layer sources can provide power to unserved areas before proceeding to more complex multi-layer configurations. This preliminary action approach minimizes the number of switching operations required by capturing the majority of restoration opportunities at simpler layers, thereby reducing device complexity and control difficulty while still maximizing load restoration.
3Reliability
If multi-layer restoration analysis is performed to maximize restored loads, then load restoration improves, but the complexity of the optimization problem increases
Solution Approach 1:
The patent implements a dynamic, adaptive algorithm that adjusts its search strategy based on the characteristics of each restoration layer. The algorithm dynamically transitions from exhaustive evaluation of first-layer solutions to more selective exploration of second-layer and third-layer solutions. This dynamic approach optimizes computational resources by focusing intensive analysis where it is most needed while using simpler evaluation methods for less critical layers, thereby managing optimization complexity while maximizing load restoration.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The method of determining back-feed paths is capable of efficiently solving multi-layer restoration problems by minimizing a fitness function using an iterative genetic algorithm. The method optimizes back-feeding of out-of-service areas by minimizing power loss, switching, unserved loads and voltage/current violations. The efficiency of the algorithm is further increased through the use of a reactive Tabu search to prevent duplicate candidate systems.