AI applications in the power-to-methanol process and fuel cells: a short discussion

Cui, X. (Lecturer), Liso, V. (Lecturer), Kær, S. K. (Lecturer)

Activity: Talks and presentationsConference presentations

Description

The power-to-methanol process provides an option for the energy storage from renewable sources as well as CO2 utilization. Moreover, the methanol product can be used for fuel cells and vehicles directly or indirectly, which can contribute to the concept of “methanol economy” or “hydrogen economy”. To complete this route, different processes could be involved: (1) the production of H2 by the electrolysis technology; (2) CO2 capture (e.g., from industrial emission or biomass); (3) the processes of methanol synthesis from CO2-rich syngas and methanol purification; (4) fuel cell applications with methanol fuel directly (by direct methanol fuel cell) or indirectly (by reformed methanol fuel cell system). There are many challenges in the above processes with complex units (e.g., fuel cells) and systems (e.g., wind power system with load uncertainties), where the technology of artificial intelligence (AI) may be able to help in the aspect of process operations, materials design and fault diagnosis. This talk presents a short discussion on the potentials of AI applications in these processes, and particularly introduces previous work on the fault diagnosis of high-temperature proton exchange membrane fuel cell stack based on an artificial neural network classifier.
Period11 Jan 2020
Event titleThe First International Conference on Energy and AI
Event typeConference
LocationTianjin, China
Degree of RecognitionInternational

Keywords

  • Power to methanol
  • AI
  • green methanol
  • wind power
  • methanol synthesis
  • fuel cell system
  • fault diagnosis
  • machine learning