论文部分内容阅读
针对激光切割速度预测精度低的难题,结合层次分析法和极限学习机的优点,提出了一种新的激光切割速度预测模型。首先建立激光切割速度的工艺参数,并采用层次分析法估计每一个参数的权值,然后采用训练样本对自适应与非线性映射能力优异的极限学习机进行训练,建立激光切割速度预测模型,最后应用于钢板切割中。实验结果表明,本文模型有效避免人为确定工艺参数权值的缺陷,描述了工艺参数与激光切割速度的变化关系,提高了激光切割速度的预测精度,具有很好的推广前景。
Aiming at the problem of low prediction accuracy of laser cutting speed, a new laser cutting speed prediction model is proposed by combining the advantages of AHP and extreme learning machine. Firstly, the process parameters of laser cutting speed were established, and the weight of each parameter was estimated by AHP. Then the training samples were used to train extreme learning machine with adaptive and non-linear mapping ability, and the laser cutting speed prediction model was established. Finally Used in steel cutting. The experimental results show that the proposed model can effectively avoid the disadvantage of artificially determining the weight of the process parameters, describe the relationship between the process parameters and the laser cutting speed, and improve the prediction accuracy of the laser cutting speed, which has a good prospect of promotion.