Multi-target shuffled frog-leaping algorithm based on multilevel message feedback

Abstract

The invention discloses a multi-target shuffled frog-leaping algorithm based on multilevel message feedback. The multi-target shuffled frog-leaping algorithm comprises the steps of optimizing a standard shuffled frog-leaping optimizing layer, a frog evolving and learning layer and an external file information exchanging layer, wherein the standard shuffled frog-leaping optimizing layer is used for obtaining new positions of the frog and comparing advantages and disadvantages of old positions with those of the new positions; entering the frog evolving and learning layer if the new positions are inferior to the old positions; obtaining Pareto domination solutions in the optimizing process and storing the Pareto domination solutions in an external file; extracting a certain number of non-domination solutions to conduct information exchange from the external file according to a preset strategy after every arriving ends so as to improve the qualities of the solutions in the file and provide a globally optimal solution for later frog optimizing and frog evolving and learning. The multi-target shuffled frog-leaping algorithm based on the multilevel message feedback has the advantages of being visual, simple, clear, universal and the like, making good use of complete information of the positions of the frog in the arriving process, and powerfully strengthening the frog ability of jumping out of the locally optimal solution.

Description

technical field [0001] The invention belongs to the technical field of data processing, in particular to a multi-objective hybrid leapfrog algorithm based on multi-level information feedback. Background technique [0002] In many engineering and scientific researches in the real world, multi-objective optimization problems (MOP) is a key problem that must be solved, and there is no very effective solution method so far. Different from single-objective optimization problems, MOP is designed based on traditional mathematical concepts due to its extensiveness and difficulty in solving, and the fact that different design conditions have different weight requirements for each objective, and the dimensions and magnitudes of each objective are different. The application of multi-objective optimization methods in practical problems often shows a certain degree of fragility. This is because MOP often has multiple conflicting goals, and the improvement of one goal may cause the perfo...

AI Technical Summary

Problems solved by technology

[0005] When traditional mathematical optimization methods are applied to multi-objective optimization problems (MOP) of complex systems, since different design conditions have different weight requirements for each objective, and the dimensions and magnitudes of each objective are different, many optimization problems need to be optimized. There are often conflicts between objectives, so it is difficult to effectively deal with MOP when the multi-objective optimization method designed based on traditional mathematical concepts is applied to practical problems.

Images