Hierarchical Nonlinear Model Predictive Control for Combined Cycle Start-up Optimization

Adrian Tica (1) , Hervé Guéguen (1) , Didier Dumur (2) , Frans Davelaar (3) , Damien Faille (3)

1 IETR - Institut d'Électronique et des Technologies du numéRique
2 E3S - Supélec Sciences des Systèmes
3 EDF R&D STEP - Simulation et Traitement de l'information pour l'Exploitation des systèmes de Production

A hierarchical model predictive control (H-MPC) structure is proposed to improve the start-up performances of Combined Cycle Power Plants (CCPPs) start-up. The structure includes two layers. At each layer, the control problem aims at deriving the profile of the gas turbine (GT) load at different time scales. To achieve this, the profile is assumed to be described by parameterized functions, whose parameters are computed by solving an optimal time control problem, based on a model developed in the modeling language Modelica and subject to a number of constraints on the plant variables. Numerical results prove the potential advantages of the proposed approach.

Domains: Automatic

Adrian Tica, Hervé Guéguen, Didier Dumur, Frans Davelaar, Damien Faille. Hierarchical Nonlinear Model Predictive Control for Combined Cycle Start-up Optimization. 2012 IEEE 51st IEEE Conference on Decision and Control (CDC), Dec 2012, Maui, Hawaii, United States. ⟨10.1109/CDC.2012.6425843⟩. ⟨hal-00747361⟩