Waste incinerator emission prediction using probabilistically optimal ensemble of multi-agents

Abstract

The emission of dioxins from waste incinerators is one of the most important environmental problems today. It is known that optimization of waste incinerator controllers is a very difficult problem due to the complex nature of the dynamic environment within the incinerator. In this paper, we propose applying a probabilistically optimal ensemble technique, based on fault masking among individual classifier for N-version programming. We create an optimal ensemble of neural network trained multi-agents and use the majority voting result to predict waste incinerator emission. We show that an optimal ensemble of multi-agents greatly improves the prediction error rate of emission of dioxins. © Springer-Verlag Berlin Heidelberg 2005.