基于BP神经网络的高频感应淬火研究

Research on High Frequency Induction Hardening Based on BP Neural Network

本文利用正交极值法,分析了阳压、阳流、栅流和加热时间对高频感应淬火法兰盘硬化层深的影响。通过BP神经网络建立硬化层深模型,研究了真空管式电源的感应加热淬火规律,开发了产品新工艺。结果表明:BP神经网络可对硬化层深进行可靠的预测,能够有效地控制高频感应淬火的生产质量。

In this paper, the effects of positive pressure, positive current, grid current and heating time on the hardened layer depth of high-frequency induction quenching flange werw analysised by orthogonal extreme method. The hardened layer depth model was established based on BP neural network to study the induction heating quenching law of vacuum tube power supply, and the product new process was developed. The results show that BP neural network can reliably predict the hardened layer depth and effectively control the production quality of high frequency induction quenching.