The Monitoring and Control of Stoker-Fired Boiler Plant by Neural Networks

摘要

The development of an integrated on-line condition monitoring and cotrol system for caol fired boiler platn has the potential to maintain linger periods of optimum boiler operation whilst keepting pollutant emissions such as Carbon Monoxide and Nitrogen Oxides (NOx) to acceptable levels. Such effort can also reduce plant down time and improve plant response by regulating a better air/fuel ratio particularly during transient load following consitions which help to pervent foukling and slag formation. This paper summarises a project undertaken to developed such a system via the use of crtificial neural networks for the monitorig and control of chain grate stoker fired boiler plant.