由表2~4可知，三层BP神经网络预测模型预测A细纱棉结指标的数据集平均

It can be seen from Tables 2~4 that the average relative error of the data set of the three-layer BP neural network prediction model to predict the A spun yarn neps index is 10.63%, and the maximum relative error of the predicted data can reach 34.07%; the data set to predict the B spun yarn neps index The average relative error is 11.45%, and the maximum relative error of the predicted data can reach 33.97%; the average relative error of the data set for predicting the C spun neps index is 18.97%, and the maximum relative error of the predicted data can reach 46.28%.

As can be seen from Table 2 to 4, the three-layer BP neural network prediction model predicts that the average relative error of the data set of A fine yarn cotton junction index is 10.63%, the maximum relative error of the prediction data can reach 34.07%;

From tables 2 to 4, it can be seen that the average relative error of the data set predicted by the three-layer BP neural network prediction model is 10.63%, and the maximum relative error of the prediction data can reach 34.07%; the average relative error of the data set predicted by the model is 11.45%, and the maximum relative error of the prediction data can reach 33.97%; the average relative error of the data set predicted by the model is 18.97%, The maximum relative error of the prediction data is 46.28%.<br>