The Research on Life Prediction of the Hoist Shaft Based on BP Neural Network

摘要

To evaluate and forecast lifespan of hoist shaft by ANN (Artificial Neural Networks), an evaluation model based on forecast lifespan is set up. A new theme is proposed to forecast remaining life of hoist shaft in which the mutation in cross-section of the bending stress acts as an input unit and remaining life acts as the output unit on the different condition. In order to forecast remaining life, it is essential to construct 5-9-1 BP (Back Propagation) network models. By studying the specific instance to forecast remaining life of 2JK hoist shaft as well as examine the accuracy of life prediction model.