Parametric sensitivity analysis of maintainability design

Because of fuzzy and correlative parameters of maintainability design, which are the factors to influent the maintainability design, traditional evaluating methods of maintainability design fail to distinctly express the influence of design factors on maintainability. This paper applies a method linking fuzzy mathematic theory and neural network to establish a network for analyzing the parametric sensitivity of maintainability design. By this network, parameters of maintainability design are regarded as the input of the network and the result of maintainability design is referred to the output of the network. Obviously, when the parameters are changed partially, the input of the network is different. As a result, the output of the network is changed too. Therefore, with fluctuant parameters of maintainability design, we can analyze the sensitive degree of maintainability on equipment by the different outputs.