Details

Title:

A Comparison between the Classical Control System and the Neural Network Approach for the Control of a Small Scale (800 Kwth) Downdraft Fixed Bed Biomass Gasifier

Author(s):

Stefanski, M., Siedlecki, M.

Document(s):

Abstract:

Fixed bed gasifiers offer a number of advantages when considering small combined heat and power (CHP) energy generation systems based on biomass gasification. In particular their simple and therefore cheap construction makes these reactors attractive for the use in distributed energy systems. In order to produce a reliable and safe CHP installation, particular attention needs to be given to the process control system. This article presents a comparison between classic control algorithms versus control systems based on artificial neural networks. The work is targeted at the application in the 800 kWth downdraft fixed bed gasifier operated by the Institute of Power Engineering. Although a neural network requires a very careful identification of the appropriate input signals, its great advantage, however, is its adaptive capability, and self­learning during the process. In the paper an overview of the automated process control issues related to the operation of a fixed bed reactor is presented together with a brief description of the main process phenomena. The influence of the main input parameters on the reactor operation is presented shortly. Next to the conclusions regarding the most suitable control approach, suggestions are made for the new monitoring tools and hardware modifications necessary to execute the desired control actions.

Keywords:

control systems, fixed bed, gasification

Topic:

R&D on Biomass Conversion Technologies for Heating, Electricity and Chemicals

Subtopic:

Gasification for power, CHP and polygeneration

Event:

21st European Biomass Conference and Exhibition

Session:

2BO.4.2

Pages:

473 - 479

ISBN:

978-88-89407-53-0

Paper DOI:

10.5071/21stEUBCE2013-2BO.4.2

Price:

FREE