您的位置：首页 > 论文页面
一种输入变量加权的诺西肽发酵过程菌体浓度测量建模方法
发表时间：2016-02-28 浏览量：2147 下载量：728
| 全部作者： | 杨强大，赵强 |
| 作者单位： | 东北大学冶金学院 |
| 摘 要： | 诺西肽发酵过程中菌体浓度难以在线测量，给控制与优化带来困难。针对该问题，利用软测量技术来实现菌体浓度的在线估计，并提出一种输入变量加权的建模方法。首先，以诺西肽发酵过程非结构模型为基础，根据隐函数存在定理进行输入变量的合理选择；然后，利用神经网络建立表征输入变量与菌体浓度之间函数关系的软测量模型，并采用粒子群优化（particle swarm optimization，PSO）算法同步求取各输入变量的权重及神经网络的连接权和阈值。基于实际生产数据进行实验研究，结果验证了所提方法的有效性。 |
| 关 键 词： | 自动控制技术；软测量；输入变量加权；诺西肽发酵过程；神经网络；粒子群优化 |
| Title： | An input-variable weighted modeling method for soft sensor of biomass in nosiheptide fermentation process |
| Author： | YANG Qiangda, ZHAO Qiang |
| Organization： | School of Metallurgy, Northeastern University |
| Abstract： | It is hard to measure the biomass on-line in nosiheptide fermentation process, which brings difficulties to control and optimization of this process. Aiming at this problem, soft sensor technique was applied to fulfil the on-line estimation of biomass, and an input-variable weighted modeling method was proposed. Based on the unstructured model of nosiheptide fermentation process, the input variables were selected according to the implicit function existence theorem. Next, a soft-sensor model representing the functional relationship between the input variables and biomass was developed using artificial neural networks, where the weight of each input variable as well as the weights and thresholds of the corresponding neural network were determined simultaneously by the particle swarm optimization (PSO) algorithm. Experimental study was also carried out based on the practical production data from a lab-scale nosiheptide fermentation process. The testing results show the effectiveness of the proposed method. |
| Key words： | autocontrol technology; soft sensor; input-variable weighted; nosiheptide fermentation process; neural network; particle swarm optimization |
| 发表期数： | 2016年2月第4期 |
| 引用格式： | 杨强大，赵强. 一种输入变量加权的诺西肽发酵过程菌体浓度测量建模方法[J]. 中国科技论文在线精品论文，2016，9（4）：357-363. |
2
评论数 0请您登录
暂无评论
下载全文
多维论文