Decomposition-based multi-objective state transition optimization method and system

Abstract

The invention relates to the field of intelligent optimization algorithms, and discloses a decomposition-based multi-objective state transition optimization method and system to realize rapid and effective optimization solving of a multi-objective optimization problem, and the method comprises the steps: selecting an objective function corresponding to the number of optimization objectives of a to-be-optimized problem, and setting iteration times and related parameters; initializing N weight vectors, and determining the neighborhood of each weight vector; initializing a candidate solution of each objective function as an initial parent solution set; distributing a weight vector for each candidate solution, calculating an objective function of each candidate solution, determining an ideal reference point of the objective function, and dividing the objective function into N sub-problems; generating a filial population of the N sub-problems by using a state transition operator, calculating an objective function value corresponding to each sub-population, and judging whether to replace the parent solution set with a candidate solution or not according to the objective function of the sub-problem; and repeating the above steps for iteration until the set number of iterations is reached.

Description

technical field [0001] The present invention relates to the field of intelligent optimization algorithms, in particular to a multi-objective state transition optimization method and system based on decomposition. Background technique [0002] For general multi-objective optimization problems, the objectives conflict with each other, often there is no solution to make all the objectives optimal, and only optimizing one of the objectives will lead to the degradation of the other objectives. Therefore, the optimal solution of the multi-objective optimization problem A solution is a set of Pareto optimal solutions that weigh all objectives. Multi-objective optimization is to obtain the Pareto optimal solution set, and the solution set should be evenly distributed in the entire frontier as much as possible. In recent years, the use of intelligent optimization algorithms to solve multi-objective optimization problems has become a research hotspot. The multi-objective optimizatio...

AI Technical Summary

Problems solved by technology

Among them, the method based on Pareto dominance mainly uses the concept of Pareto dominance to select non-dominant solutions, and uses the diversity maintenance mechanism to enhance the selection pressure, and finally selects a Pareto non-dominant solution set with a relatively uniform distribution. When the number of objective functions increases, non-dominant solutions that grow exponentially will be generated, and the evolution of candidate solutions to Pareto optimal solutions becomes extremely slow; decomposition-based methods use weight vectors and aggregation functions to transform multi-objective optimization problems into For a series of single-objective optimization sub-problems, by optimizing these single-objective optimization sub-problems at the same time, the Pareto optimal solution set of the multi-objective optimization problem is finally obtained. This method based on decomposition is greatly affected by the weight vector, and the weight vector and the candidate solution Whether matching has a great influence on the selection of candidate solutions; the index-based method mainly solves the multi-objective optimization problem by calculating a special index and then using this index to guide the evolution process of the multi-objective optimization algorithm, but this method As the number of targets increases, the complexity of the algorithm also increases exponentially.

Images