Two-Step Genetic Algorithm for Pre-Cooling Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining optimal pre-cooling strategy parameters for energy consumption optimization are computationally intensive and costly due to the high number of possible combinations of start and end times for cooling periods, leading to high computation time and cost.
Innovation Solution
A two-step genetic algorithm is employed to optimize energy consumption by identifying candidate solutions and iteratively refining them to determine optimal parameters for the pre-cooling strategy, reducing computation time and cost.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a numerical model is used to exhaustively simulate the cost of energy consumption for all possible combinations of pre-cooling times, then the optimal parameters for pre-cooling strategy can be determined, but the computation time and cost become excessively high
Solution Approach 1:
The patent divides the exhaustive search space into multiple segments by implementing a two-step genetic algorithm. The first step performs a coarse search with a larger population size to identify promising regions, while the second step refines the solution with a smaller population size focused on the identified regions. This segmentation approach maintains optimization accuracy while significantly reducing computation time compared to exhaustive simulation of all possible combinations.
Solution Approach 2:
The patent applies preliminary action by performing an initial coarse optimization phase before the final refinement phase. The first step of the two-step genetic algorithm pre-identifies optimal or near-optimal regions in the parameter space, which then guides the second step to focus computational resources on refining those specific areas. This preliminary action avoids wasting computational time on obviously suboptimal combinations.
2Reliability
If the number of possible combinations of pre-cooling times is increased to ensure optimal parameter selection, then the quality of optimization improves, but the computation cost increases significantly
Solution Approach 1:
The patent implements dynamics by adapting the population size and search strategy at different stages of the optimization process. The two-step genetic algorithm dynamically adjusts computational resources: the first step uses a larger population to explore the search space broadly, while the second step uses a smaller population to exploit promising regions. This dynamic approach ensures high optimization quality while minimizing computation cost.
Solution Approach 2:
The patent applies parameter changes by modifying the population size and other algorithm parameters between the two steps. The first step uses parameters optimized for exploration (larger population), while the second step uses parameters optimized for exploitation (smaller population). This parameter adaptation allows the system to maintain reliability of optimization results while reducing overall computation cost.
Data Source
AI summary
In order to reduce computation time and cost involved with determining one or more optimal parameters for a pre-cooling strategy, for a modeled system, a two-step genetic algorithms is used to optimize energy consumption of the modeled system with respect to cost of the energy consumption. A first step of the two-step genetic algorithms determines a population of potential solutions that are used to initialize a second step of the two-step genetic algorithm. The second step of the two-step genetic algorithm determines the one or more optimal parameters for the pre-cooling strategy from the population output by the first genetic algorithm.