On learning control in industrial furnaces and boilers

This paper presents some novel contributions to the integrated learning control and supervision of industrial thermal systems such as furnaces and boilers. Three novel application case-studies have enabled to demonstrate practical feasibility and efficiency of the learning control and supervision approach in thermal systems automation. This approach provides for a methodology for solving the problem of process identification, control and supervision for complex, multi-variable, thermal plants within a time-domain setting and by using recorded, operating time series and/or rough, non-parametric process models.