基于小样本预测模型在机械工程上的应用

Application to the mechanical engineering based on small-sampling prediction model

为预测短期内机械生产实验的数据分布规律,以更好地指导生产和提高工作效率,充分利用计算机模拟研究的优势,基于金属切削工艺加工的部分实验数据,利用线形回归模型、灰色模型和神经网络模型对实验数据进行了分析处理并建立了小样本数据的综合预测模型。研究结果表明小样本预测模型是可行的,它对机械工程各种相关实验数据的短期预测都具有普遍适用性,同时提供了可以快速了解数据在未来分布规律的方法。

To predict data distributing rule of mechanical manufacture and experimentation in a period of time for guiding the production and improving the work efficiency, the linear regression, grey model and neural networks model were used to analyze and process partial experimental data with the advantage of cyber-simulation based on the manufacture of metalline cutting technics and established the integrated forecasting model of small data set.The results show that small-sampling prediction model is feasible, it has universal practicability that predicts all kinds of relative experimental data in the mechanical engineering and provides a quick method for finding out the data distributing rule in the future