Abstract
Artificial neural networks, the artificial intelligence systems that imitate functions of biological neurons, have been widely used in different areas due to their variety and ability to conform to specificities of different applications. When it comes to application of artificial neural networks in bioprocess modeling, their task usually represents prediction or forecasting the values of dependent variables (outputs) based on given values of independent variables (inputs). Although bioprocess model is the 'black box' and remains unknown, which could represent the obstacle in bioprocess analysis, neural networks have shown better ability in prediction of bioprocess results comparing to other modeling methods, such as RSM (Response Surface Methodology) and mathematical modeling. Obtained model could be further used for bioprocess optimization, commonly performed using genetic algorithms. This study provides the review of the main characteristics and applications of artificial neural networks in modeling and optimization of biofuels (bioethanol, biogas and biohydrogen) production.
Highlights
Artificial neural networks (ANNs) represent artificial intelligence systems created with an aim to imitate functions of human brain and biological neurons and to be applied in solving different problems which require performing of complex cognitive operations
Depending on the applied software, obtained bioprocess model usually remains unknown (“black box”) (Jokić et al, 2011), which could represent an obstacle in bioprocess analysis, artificial neural networks show great potential and many advantages comparing to traditional bioprocess modeling approaches
Bioethanol production Due to the great interest in production of bioethanol from different agroindustrial remains in the past decade (Dželetović and Mihailović, 2011), a large number of studies concerning the application of artificial neural networks and genetic algorithms in modeling and optimization of these bioprocesses, as well as in modeling and optimization of other biofuels’ production processes, could be found in the literature (Table 1)
Summary
Artificial neural networks (ANNs) represent artificial intelligence systems created with an aim to imitate functions of human brain and biological neurons and to be applied in solving different problems which require performing of complex cognitive operations. ANNs have been used in different areas, such as pattern classification, clustering/categorization, function approximation, prediction/forecasting, associative memory and control (Jain et al, 1996). Their wide application is enabled due to variety of neural network types and their adaptability to certain application. The most usually used neural networks are based on multilayer perceptron architecture. Neural network architecture significantly affects its performance and usually adapts to concrete problem’s demands. One of the neural networks elementary traits is their ability to learn, i. E. generalization ability based on the experience that network acquires in the training process One of the neural networks elementary traits is their ability to learn, i. e. generalization ability based on the experience that network acquires in the training process
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
