基于时间序列汽车轮毂尺寸偏差趋势预测

Prediction of Vehicle Wheel Hub Dimensional Deviation Based on Time Series

文章选用指数平滑法和ARIMA法两种典型的时间序列算法作为预测模型,对流水线生产的汽车轮毂倒锥大尺寸值偏差趋势进行预测。以MATLAB作为工具进行实验计算,依据预测数据和实际观察值的RMSE和期望误差的对比来比较这两种模型的预测效果,得出在预测汽车轮毂倒锥大尺寸值偏差趋势时,二次指数平滑法的预测效果更好,所以最终采用此模型,从而为宏观调控刀具补偿值提供有力的依据。

In this paper,two typical time series algorithms,namely exponential smoothing method and ARIMA method,are used as prediction models to predict the deviation trend of large-scale values of inverted cones of automobile hubs produced by pipelines.The experimental calculation was carried out with MATLAB as the tool,according to RMSE and the expected error of predicted data and actual observation,the prediction effects of the two models are compared.It is concluded that the second exponential smoothing effect is better when predicting the deviation trend of the large size of the inverted cone of the automobile hub,so the model is finally adopted,providing a strong basis for macro-control tool compensation values.