论文部分内容阅读
将连续域贝叶斯优化算法(rBOA)与约束处理技术相结合,用于求解约束优化问题,其主要思想是利用约束条件影响优秀个体的选取,并间接影响概率模型,使之引导群体产生满足约束条件的子代个体,从而求得满足约束条件的最优解.将rBOA与4种不同的约束处理技术结合起来,并测试了其优化性能,实验结果表明rBOA与多目标优化法结合相比另外3种方法具有最好的优化效果,但其优化效果仍有待改进.
Combining the continuous-domain Bayesian Optimization (rBOA) algorithm and constraint processing techniques to solve constrained optimization problems, the main idea is to use constraint conditions to influence the selection of excellent individuals and to indirectly influence the probability model so that the guidance population can satisfy And get the optimal solution to meet the constraints.We combine rBOA with four different constraint processing techniques and test their optimization performance.The experimental results show that rBOA and multi-objective optimization methods combined The other three methods have the best optimization results, but the optimization results still need to be improved.