基于BP人工神经网络的磨料水射流切割深度模型

Cutting depth model of abrasive water jet based on artificial BP neural network

在已知的磨料水射流切割混凝土实验数据的基础上,应用BP人工神经网络理论,并且借助MATLAB神经网络工具箱,建立磨料水射流切割深度模型.模型中包括的射流参数有射流压力、靶距、磨料粒径、磨料流量、磨料喷嘴直径、磨料喷嘴长度及横移速度.通过模型预测结果与实验结果的比较,表明模型具有一定的精度.

Cutting depth model of abrasive water jet was established based on the experimental data of abrasive water jet cutting concrete by applying artificial BP neural network and MATLAB neural network tool box. The model includes several jet parameters： jet pressure, standoff distance, diameter of abrasive, mass flow rate of abrasive, diameter of abrasive nozzle, length of abrasive nozzle and traversing velocity of jet. The result of comparison between the model-predicted data and experimental data shows that the model is of reliable precision.