Subspace identification and predictive control of boiler-turbine coordinated system

作者：Wu Xiao; Shen Jiong^*

来源：Journal of Southeast University(Natural Science Edition), 2012, 42(SUPPL2): 281-286.

DOI：10.3969/j.issn.1001-0505.2012.S2.011

• Advanced controller
• Boiler turbine coordinated system
• Boiler turbine system
• Data-driven modeling
• Identification model
• Multi-model
• Predictive controller
• Subspace identification

A novel data-driven modeling strategy for boiler-turbine coordinated system using subspace identification and multi-model method is proposed. According to the decomposition and synthesis principle of multi-modeling, the original nonlinear system is divided into several local models, and subspace method is used to identify the local state-space model by obtaining the input and output data corresponding to the local models. Then coherence relations of each local model in the swithching points are used to transform all local models to the common basis to build the integrated multi-model system. Due to its data-driven feature, the proposed modeling method can easily be adapted to different types of systems without knowing the underlying system models, and the state-space form of the resulting model makes it suitable for advanced controller design. Based on the identification model, a constrained multi-model predictive controller (MMPC) is designed to operate the boiler-turbine system in a wide range of operation. Simulation results demonstrate the effectiveness of the proposed approach.

• 出版日期2012
• 单位东南大学