USING ANFIS FOR MAINTENANCE PLANNING OF TEXTILE MACHINES

Abstract

Considering the complexity of the wear of textile machines, a mathematical modeling of this phenomenon is not available in the existing literature. Based on the advantages of both fuzzy logic and neural networks, a neuro-fuzzy approach seems to be well suited for the prediction of the maintenance activities of textile machines. Therefore, the Adaptive Neuro Fuzzy Inference System (ANFIS) was proposed in this study to plan the maintenance activities of textile machines. The research was performed on the PEGASUS sewing machine at a working speed of 3100 stitches/minute. The sewing material used in the experiments was cotton, while NM 80 sewing needles were employed. The vibrations along the OZ axis and level of noise of the PEGASUS sewing machine were measured with appropriate devices. The ANFIS of the Matlab® software was used to plan the maintenance of textile machines. The amplitude of vibration and level of noise of the needles were used as inputs in the ANFIS system. The output of the ANFIS system was the time to replacement of the needles. The performance of the developed ANFIS system was expressed through the RMSE and MAPE measures. Considering their values, it may be pointed out that the ANFIS prediction system demonstrates good performance.

Keywords

• wear
• vibration
• noise
• anfis
• replacement