ID 原文 译文
49006 为了解决随机选择初始群体可能会延长搜索时间这一问题,将互信息引入到算法中。 In order to solve the random initial population could extend the search time of this problem, the mutual information is introduced into the algorithm.
49007 通过计算特征与类别的相关性来确定每个特征的入选概率,根据概率值生成一个近似最优粒子,使粒子群一开始就沿着比较合理的方向搜索,从而缩短进化时间。 Through calculating the correlation of features and categories to determine the characteristics of each selected probability, according to the probability value generated an approximate optimal particle, the particle swarm in a reasonable direction from the start the search, so as to shorten the time evolution.
49008 最后,以支持向量机(support vector machine,SVM)为分类器,通过仿真实验验证了算法的可行性和有效性。 Finally, support vector machine (support vector machine, SVM) as classifier, through the simulation experiment proves the feasibility and effectiveness of the algorithm.
49009 对于适应度函数计算耗时较大的工程优化问题,采用仿生智能优化算法求解时常遇到由于适应度函数评价次数过大而导致计算量过高的瓶颈问题。 For fitness function computation is time-consuming engineering optimization problem, using bionic intelligent optimization algorithm often encountered due to the large number of fitness function evaluation and results in excessive amount of calculation of the bottleneck problem.
49010 针对上述问题,提出一种基于粒子群优化(particle swarm opti-mization,PSO)算法与高斯过程(Gaussian process,GP)机器学习方法的协同优化算法(PSO-GP)。 Aiming at these problems, puts forward a kind of based on particle swarm optimization (particle swarm opti - mization, PSO) algorithm and the Gaussian process (Gaussian process, GP) machine learning method of collaborative optimization algorithm (PSO - GP).
49011 该算法在寻优过程中采用GP近似模型来构建决策变量与适应度函数值之间的映射关系,在PSO全局寻优过程中不断地总结寻优历史经验的基础上,预测可能包含全局最优解的搜索区域,以优化粒子群飞行的方向。 The algorithm in the optimization process using GP approximation model to build the mapping relationship between decision variables and the fitness function value, constantly in the process of PSO global optimization, on the basis of summarizing historical experience for optimum forecast may contain the global optimal solution of the search area, the direction of the particle swarm optimization (PSO) to optimize the flight.
49012 多个测试函数的优化结果表明,该算法是可行的,与基本PSO算法相比,在获得全局最优解的前提下,可显著减小寻优过程中的适应度函数评价次数,寻优效率较高,在高计算代价复杂工程优化问题的求解上具有良好的应用前景。 Multiple test function optimization results show that the algorithm is feasible, compared with the basic PSO algorithm, to obtain the global optimal solution of the premise, can significantly reduce the number of fitness function evaluation in the process of optimization, optimization efficiency is higher, in the high computational cost in solving complex engineering optimization problems with good application prospect.
49013 针对准则权重信息完全的区间二型模糊多准则群决策问题,提出了相似度测量方法以及基于区间二型模糊数相似度的决策方法。 For standards of the information on attribute weights is completely interval type 2 fuzzy multiple criteria group decision making problems, puts forward the method to measure and decision making based on interval type 2 fuzzy number similarity method.
49014 该方法首先通过区间二型模糊数的集结算子计算出方案的综合准则值,根据期望值对综合准则值进行比较后确定出正、负理想方案,然后计算各方案分别和正、负理想方案之间的相似度,进一步求出贴近度系数后从而得到方案的排序。 This method first by interval type 2 fuzzy number of rally operator to calculate the scheme of comprehensive criteria values, according to the expected value of comprehensive criterion value after comparison to determine the positive and negative ideal solution, then calculate each scheme respectively, and the degree of similarity between positive and negative ideal solution, further to find the closeness coefficient solution is obtained after sorting.
49015 最后,算例分析表明了该方法的有效性和可行性。 Finally, the example analysis shows that the method is effective and feasible.