不确定凸模型近似算法的一种改进
AN IMPROVEMENT OF THE APPROXIMATE SOLUTION TO CONVEX MODELS OF UNCERTAINTIES
-
摘要:将非概率凸模型理论与摄动理论相结合,通过有界不确定参数结构的特征值问题,对凸模型理论的一次近似算法作出一种改进.改进后的算法由于在计算中不用特征值导数,与Elishakof的算法相比,不仅拓广了凸模型理论的应用范围,而且还可提高算法的计算效率.Abstract:In terms of the eigenvalue problem with uncertain parameters,by means of non probabilistic,convex modeling combined with perturbation theory, an improvement is made on the first order approximate solution inconvex models of uncertainties. Without calculating the derivative of eigenvalues, the improved method not only can widen application of convex modeling, but also can greatly increase the efficiency ofcomputations.