针对纺织企业对小数据及含有突变数据的纱线单强指标难以预测的问题，提出了

Aiming at the problem that textile companies cannot predict the single-strength indicators of yarns with small data and mutation data, a method for forecasting single-strength yarns that combines gray correlation and nearest neighbor is proposed. In this paper, 26 indicators such as raw cotton quality indicators and key process parameters are used as input factors for the prediction model, and the single yarn strength is predicted. The experimental results show that the relative error of the new method to predict the yarn single strength index is basically within 10.0%, while the relative error of the individual data predicted by the traditional three-layer BP neural network is very large. Compared with the three-layer BP neural network, the new method also has a good prediction effect on the single-strength index of small sample yarn varieties with mutation data.

In view of the difficult prediction of small data and the single-strength index of yarn containing mutation data in textile enterprises, a method of predicting the single strength of yarn that combines gray correlation and the nearest neighbor is proposed. In this paper, 26 indicators, such as raw cotton quality index and key process parameters, are used as input factors for the prediction model, and the yarn single strength is predicted. The experimental results show that the relative error of the new method predicts the single-force index of yarn is basically within 10.0%, while the individual data of the traditional three-layer BP neural network predicts the relative error is very large. Compared with the three-layer BP neural network, the new method also has a good prediction effect on the single-strength index of small sample yarn variety with mutation data.

In order to solve the problem that it is difficult to predict the single strength index of yarn with small data and mutation data in textile enterprises, a prediction method of single strength of yarn combining grey correlation and nearest neighbor is proposed. In this paper, 26 indexes such as raw cotton quality index and key process parameters are used as input factors of the prediction model, and yarn strength is predicted. The experimental results show that the relative error of the new method is less than 10.0%, while that of the traditional three-layer BP neural network is very large. Compared with the three-layer BP neural network, the new method also has a good prediction effect on the single strength index of small sample yarn varieties with mutation data.<br>