耐磨铸铁回火工艺的人工神经网络算法应用

Application of Artificial Neural Network Algorithm for Tempering Process of Wear Resistant Cast Iron

根据试验结果,以回火温度和保温时间为输入、硬度为输出建立了耐磨合金铸铁的人工神经网络模型。结果表明,经训练后的模型同实测结果吻合度较好,对实际生产有较强的指导意义。保温时间相同时,合金硬度随回火温度升高下降明显;回火温度不变时,合金硬度随保温时间增加略有下降。

According to the test results, the artificial neural network model of wear-resistant alloy cast iron was established with tempering temperature and holding time as input and hardness as output. The results show that the trained model is in good agreement with the measured results, which has a strong guiding significance for the actual production. When the holding time is the same, the hardness of the alloy decreases obviously with the increase of tempering temperature. When the tempering temperature is constant, the hardness of the alloy decreases slightly with the increase of holding time.