论文部分内容阅读
用神经网络降维分析法把制浆蒸煮工艺过程的样本数据降维映射到二维平面上 ,并生成纸浆得率和纸浆硬度的等值线 ,由此可直观地确定出控制纸浆硬度在一定范围内而纸浆得率较高的优化操作区域。优化区域中的任意点通过逆映射可转变成实际的操作条件。用实例演示了这一新方法的应用
Using the neural network dimension reduction analysis method, the sample data of the pulping process is dimensionally reduced to the two-dimensional plane and the contour of the pulp yield and the pulp hardness is generated. Thus, the control of the pulp hardness can be intuitively determined Range and pulp yield higher optimization of the operating area. Any point in the optimization area can be transformed into the actual operating conditions by inverse mapping. An example is given to demonstrate the application of this new method