Prediction of conductive thread consumption of transmission lines using artificial neural networks

This paper introduces a novel approach to predict conductive thread consumption in transmission lines through the application of Artificial Neural Networks (ANNs). The study focuses on leveraging ANNs to analyse various factors influencing conductive thread usage in the production of transmission lines within textile-based electronic systems. By training the network on data collected from samples, the proposed model aims to accurately predict the conductive thread required for different transmission line configurations, enabling more efficient and cost-effective design processes in electronic textiles. In this study, transmission lines are generated using conductive thread with three different stitch types and four different stitch densities on five different fabrics, and conductive thread consumption is predicted using ANNs. The learning algorithm of the neural network is chosen as feed-forward back propagation, and the training algorithm is the Levenberg-Marquardt algorithm. Based on the obtained regression coefficient (R2=0.98683), it is suggested that the data has a linear structure, and it is expected the structured network will have a strong estimation performance.