论文部分内容阅读
任何一项工程或一个产品的设计,都需要根据设计要求,合理选择方案,确定各种参数,以期望达到最佳的设计目标。本文建立了广义参数优化方法用以解决黑箱系统的参数优化问题,该方法以试验设计为基础,进行实验方案选择;采用人工神经网络建立因素与目标的非线性映射关系模型;利用遗传算法,获得给定参数区间的Pareto最优解集。提出的方法具有通用性,可广泛应用于各种基于试验或虚拟试验的黑箱系统多目标参数优化问题的求解。
Any project or a product design, according to the design requirements, the choice of programs to determine the various parameters in order to achieve the best design goals. In this paper, a generalized parameter optimization method is proposed to solve the problem of parameter optimization in a black box system. The method is based on the experimental design and the experimental scheme is selected. The artificial neural network is used to establish the nonlinear mapping relationship model between the factors and the target. Pareto optimal solution set for a given parameter interval. The proposed method is universal and can be widely used to solve multi-objective parameter optimization problems of black box systems based on experiments or virtual experiments.