摘要
以全脂大豆粉为原料,以单螺杆挤压机加工时的物料含水率、螺杆转速和机筒温度为工作参数,对挤压机的挤压膨化效果(以膨化压力来衡量)进行试验研究。利用正交旋转组合试验设计及试验数据,建立了单螺杆挤压机的人工神经网络模型。对未参与人工神经网络建模的数据进行评价和比较,结果表明,该模型对一定工作参数条件下的预测结果最大相对误差为8.76%,其结果对于生产实践具有指导作用。
The effects of varying materiel moisture, screw speed and barrel temperature on extrusion quality (expressed as pressure of extrusion) of soybean material were studied by using a screw extruder. In the paper, a artificial neural network modeling of screw extruder was established by using a orthogonal rotatory experiment design and experimental data. The maximal predicting error of the ANN model for the technological parameters condition was 8.76%, by evaluating several data sets unused for the ANN models’ learning and training. The results had supervise function to real produce.
出处
《黑龙江八一农垦大学学报》
2005年第2期89-91,共3页
journal of heilongjiang bayi agricultural university
关键词
单螺杆挤压机
人工神经网络
建模
工作参数
screw extruder
artificial neural network
modeling
manipulate parameters