Formulating standard product lead time at a textile factory using artificial neural networks
Product lead time (PLT) is difficult to be estimated in the textile industry due to problems, such as incomplete data, large product variation, and non-linearity in the time-affecting factors. This thesis proposed a methodology to formulate product lead time of textile fabric production at a textile factory using artificial neural networks. Analysis of the order fulfillment process flow of the textile company was conducted to identify the individual sequential processes that constitute product lead time. Feed forward multilayer perceptron (MLP) neural networks are developed to estimate the lead time of critical PLT processes with incomplete data and various non-linear time- affecting factors. The networks are trained in a supervised manner using back propagation algorithm. The finalized neural network models are able to estimate the lead time for each process with a good degree of accuracy and can be used as a decision making tool for quoting product lead time to customer.
B01158 | (wh) | Available |
No other version available