Process State Identification and Modeling in a Fluidized Bed Energy Plant by Using Artificial Neural Networks

The aims to reduce harmful emissions are influencing increasingly the modern-day production of energy, while higher objectives are set also for the efficiency of combustion processes. Therefore it is necessary to develop such data analysis and modeling methods that can respond to these demands. This paper presents an overview of how the formation of nitrogen oxides (NOx) in a circulating fluidized bed (CFB) boiler was modeled by using a sub-model -based artificial neural network (ANN) approach. In this approach, the process data is processed first by using a self-organizing map (SOM) and k-means clustering to generate subsets representing the separate process states in the boiler. These primary process states represent the higher level process conditions in the combustion, and can include for example start-ups, shutdowns, and idle times in addition to the normal process flow. However, the primary states of process may contain secondary states that represent more subtle phenomena in the process, which are more difficult to observe. The data from secondary combustion conditions can involve information on e.g. instabilities in the process. In this study, the aims were to identify these secondary process states and to show that in some cases the simulation accuracy can be improved by creating secondary sub-models. The results show that the approach presented can be a fruitful way to get new information from combustion processes.