Inverse model approach to “online” performance enhancement for coal-fired power plant

Abstract

This paper presents the control strategy of a fossilfuel power plant. In previous work, an advanced coordinated controller was synthesized to achieve stability with satisfying performances. This controller was implemented on-site leading to satisfactory results. However, mainly because of modeling uncertainties, the obtained performances could be non-optimal, with the risk of non-satisfaction of contract specification. In order to improve the plant behavior on-site, a new controller is added to the previous one. It consists in applying an additional control input computed from the error between the system output and a desired output trajectory. This Additional Inverse Model Controller (AIMC) is based on an inverse model of the plant and low-pass filters with small DC gain. The efficiency of the proposed control strategy is illustrated through realistic simulations and compared to the initial controller performance.