Application of machine learning methods for optimizing the technical and economic performance of generating systems

Abstract

The paper is on the technical and economic performance optimization technique for thermal and power generating system using machine learning methods. The possibility of using regression analysis for parameter influence evaluation when calculating technical and economic performance in order to reach better generating unit efficiency is described. The approach to evaluate the parameter influence of a large distributed control system is presented.

Application of machine learning to predict the thermal power plant process condition

“…The first step in analyzing process variables is to join all parameters by timestamps using outer merge. Since the data is stored in the software-and-hardware historical database at different time intervals, i.e., some parameters can be written more often, the next step is to fill the empty fields with the last known states, assuming that the parameter value has not been updated yet [4]. However, after the join, the time interval between the entries may be different.…”