[4F3-GS-10] AI application

[4F3-GS-10-01] Analysis of Stability Degree for Combustion of Waste-to-Energy Plants with Fuzzy Relational Map of Sensors

—Application of Data Selection Method

〇Akifumi Ise¹, Motohide Umano¹, Kiyotaka Kohigashi¹ (1. Hitachi Zosen Corporation)

Keywords:fuzzy logic, fuzzy relational map, Automatic Selection of Data, energy-from-waste plant, prediction of combustion state

In Waste-to-Energy plants, combustion is dependent on many complex factors and a sign of unstable states appears in other locations. We have proposed a fuzzy relational maps of many sensors (FuRMS, for short) for expressing relationships between many sensors as relations between fuzzy sets of sensor values. Waste combustion is so complex, for example, data of a combustion state contain data of a slightly different state or consist of multiple patterns. We employed the data selection method, where we construct a fuzzy relational map from a dataset and exclude the data whose evaluation values are very low for the fuzzy relational map. We iterate the process several times for a dataset of stable state. The evaluation values increased in the early iterations but decreased in the further iterations for the original data. The increase was caused by excluding data with slightly different states from the original data set, while the decrease was caused by excluding data of a pattern with a fewer data.