Multi-objective cooperative control based on evolutionary process optimization of immune algorithm
FengJun Hu,HeXin Lv,Alexander V. Tuzikov
DOI: https://doi.org/10.1007/s11761-020-00295-w
2020-06-26
Service Oriented Computing and Applications
Abstract:Based on evolutionary process optimization immune evolution algorithm, a multi-balloon cooperative control strategy is proposed, which targets at the problem that an artificial immune algorithm is apt to turn premature hastily when solving multi-objective problems. First, the control model of the multi-balloon robot is transformed into a multi-objective optimization problem, and an artificial immune algorithm is used to map and control, then Gaussian perturbation is added to the cloning behavior in the evolution process to ensure that there is a difference in each progeny after cloning. Weighting the most concentrated solution of the variation results, so that the scope of the solution is more diversified. Finally, the extinction mutations are performed on the extinct antibodies, which reduce the probability of antibody death and further maintains the diversity of the population. The simulation results show that compared with the standard support vector machine,the least squares support vector machine method and the standard artificial immune algorithm, the optimization result of the improved algorithm proposed in this paper is better in the solution of multi-objective optimization problem and the effect of the multi-balloon synergy control.