Abstract
This study illustrates an approach to provide an easily manageable tool for end-users to predict the biogas production rate to ease the transition to a flexible and proactive operation of digesters at both biogas plants as well as at wastewater treatment plants. To provide a high flexibility to end-users but keep BioTOOL easily manageable, the approach uses a pre-trained artificial neural networks library consisting of different variable combinations. In BioTOOL’s workflow, end-user will be asked for current measurements disregarding the combination of those. BioTOOL uses a seek-and-implement method to find the correct stored network for the entered input. This approach results in predictions similar to the optimized pre-trained network. It has been demonstrated that the idea of a flexible prediction tool could be fully realized.
Originalsprache | Englisch |
---|---|
Seiten (von - bis) | 87-94 |
Seitenumfang | 8 |
Fachzeitschrift | Environmental Modeling and Assessment |
Jahrgang | 24 |
Ausgabenummer | 1 |
DOIs | |
Publikationsstatus | Veröffentlicht - 1 Feb. 2019 |