等离子切割过程参数智能决策的研究

摘要

针对等离子切割过程非线性、工艺参数相互耦合、参数变化引起的时变性等导致获得最佳切割工艺参数较困难从而影响切割质量和效率的问题,论文利用人工神经网络在非线性系统决策中良好的性能,建立了基于人工神经网络的切割过程参数智能决策模型,实验表明,该模型能够在切割过程中较准确的预测工艺指标参数,预测输出值与样本值间的平均误差小于6％,满足了等离子切割工艺参数决策的要求.%By taking into account these problems such as the nonlinear cutting process,the multi-parameter close coupling of cutting technology process and time variation caused by changing parameter, we take advantage of neural networks' good performance in nonlinear system identification, established the intelligent decision system based on neural network, The results show that the method can accurately predict the output value and the error between predict value and sample values is less than 6% ,it can satisfy requirement of parameter decision in plasma cutting process.