Research in Reliability Modeling of WEDM Based on Neural Network

Because of the small size of the wire-electronic discharging machine (WEDM) reliability data, the reliability model of WEDM can not be set up accurately by the traditional reliability modeling methods. Therefore, the neural network is used in this paper to solve this problem. The algebraic algorithm, a new algorithm of neural network, is used in the proposed method. Based on statistically analyzing the reliability data of WEDM, the general idea and basic method is proposed for setting up reliability model by neural network, and a new way of reliability modeling is found out for WEDM reliability modeling. After an example is calculated, it is shown that the calculation time of the proposed method is 3 times, and the mean absolute error of the proposed method is 0.0156, as well as the mean absolute error of the traditional method is 0.0703. It is proved that the model which is set up by neural network can map the actual data more accurate than the traditional model. It has a good reference to WEDM reliability researching.