Dynamic environment optimization method based on random drift particle swarm optimization algorithm

Abstract

The invention discloses a dynamic environment optimization method based on a random drift particle swarm optimization algorithm, which belongs to the technical field of dynamic environment multi-objective optimization. The dynamic environment optimization method comprises the following algorithm processes of: (1) dividing an initialization particle swarm into a plurality of sub swarms by using hierarchical clustering; (2) updating the speeds and positions of all of the sub swarms; (3) calculating adaptive values of particles, evaluating and selecting an optimal particle; (4) performing overlapping detection, crowding detection and convergence detection on the sub swarms; (5) performing environmental change detection on the whole particle swarm; (6) judging whether stopping criteria for iteration are satisfied. The problems that dynamic environment optimization control variables are likely to be in a local optimal solution and the optimum value solving speed is low are solved, and the dynamic environment optimization method has the advantage of high robustness and adaptability.

Description

technical field [0001] The invention relates to a dynamic environment optimization method based on a random drift particle swarm optimization algorithm, and belongs to the technical field of dynamic environment multi-objective optimization. Background technique [0002] Dynamic Optimization Problems (DOPs) is to study the change footprint of tracking and locating the optimal solution when the objective function, environmental parameters and constraints change. PSO (Particle Swarm Optimization) is an evolutionary algorithm that simulates group behavior. It has the characteristics of simple operation and few parameter settings. It has been applied to dynamic environment optimization. The current particle swarm optimization algorithm has undergone certain After the iteration of , the population will converge to a satisfactory solution due to the slow reduction of the flying speed of the particles. Once the environment changes, it is difficult for the population that loses diver...

AI Technical Summary

Problems solved by technology

At present, there are not many literatures on the random drift particle swarm optimization algorithm, and most of them are used to solve static environment optimization problems. For example, in 2014, Fang Wei et al. applied the random drift particle swarm optimization algorithm to the function optimization problem. Local search capability and global search capability, whose control parameter settings have a significant impact on the optimization results

However, this is only an optimal value problem in solving static environment optimization problems, but there are very few solutions to multi-peaks function Optimization Problems (MOPs) in dynamic environments.

Images