Abstract: To improve the quality of textile weaving process, and to guarantee the continuity of the entire manufacturing process, first, the application of the existing monitoring system was analyzed. Then, for the system deficiencies in data fitting processing and abnormal cases warning, the uncertainty factors in the manufacturing process were studied deeply, and the used software filtering algorithms in the system were compared and analyzed. Furthermore, through using empirical mode decomposition (EMD), de-noising experiments on weaving process data were compared, and the fitting results were simulated. Practice has proved that EMD-based method can effectively ensure the accuracy of production data, improve the quality of the textiles and guarantee the continuity of the entire manufacturing process. Meanwhile, the results can be benefit for accelerating the transition from 'labor-intensive' to 'technology-intensive' and 'brand-intensive' of the textile enterprises.

Key words: data fitting, weaving procss, application option, Empirical Mode Decomposition