当前位置: X-MOL 学术Chem. Eng. Technol. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Application of Artificial Neural Networks in the Tertiary Treatment of Liquid Effluent with the Microalgae Chlorella vulgaris
Chemical Engineering & Technology ( IF 1.8 ) Pub Date : 2021-08-10 , DOI: 10.1002/ceat.202100277
Yasmin O. Carvalho 1 , Weverton V. Oliveira 2 , Rogério L. Pagano 1 , Cristina F. Silva 1, 2
Affiliation  

Modeling of biotechnological processes is difficult due to their nonlinear characteristics. Therefore, an artificial neural network was developed as a viable option for the prediction of the main variables of the tertiary treatment with Chlorella vulgaris. The network was designed with a hyperbolic tangent activation function and trained using a set of experimental and simulated data. The Levenberg-Marquardt algorithm was used to train the network; four input neurons (luminosity, ammonium ion, phosphate, and initial biomass concentrations) and two output neurons (ammonium ion and phosphate concentrations) were fixed. The network architecture with one hidden layer [4,7,2] was chosen because it presented the lowest mean square error of the test combined with high R2, indicating that the network provides a good model for use in real applications.

中文翻译:

人工神经网络在微藻小球藻三级处理中的应用

由于其非线性特性,生物技术过程的建模很困难。因此,开发了人工神经网络作为预测普通小球藻三级处理的主要变量的可行选择。该网络采用双曲正切激活函数设计,并使用一组实验和模拟数据进行训练。使用Levenberg-Marquardt算法训练网络;四个输入神经元(光度、铵离子、磷酸盐和初始生物量浓度)和两个输出神经元(铵离子和磷酸盐浓度)是固定的。选择具有一个隐藏层 [4,7,2] 的网络架构是因为它表现出测试的最低均方误差和高R 2,表明网络提供了一个很好的模型,可以在实际应用中使用。
更新日期:2021-09-20
down
wechat
bug