The utilization of heavy-duty gas turbines in power plant, coupled with a combined cycle system (CCPP), represents a significant alternative to conventional coal-fired boiler turbine units, primarily due to its swift load response capabilities and environmentally benign emissions profile. In China, the gas turbines control system (TCS) have been predominantly supplied and stringently regulated by international vendors. The domestically developed TCS needs extensive validation prior to its commercial deployment. The adequate fidelity model of the CCPP to represents both static and dynamic characteristics plays a critical role in this process. Hybrid modeling provides a promising option to tackle such complex problems. In hybrid modeling, the process-related principles are represented by mechanistic models, while the coefficients added to the first-principal models as adjustments to present unknown process characterizes, are described by data-driven models. This article highlights the significance of hybrid modeling in multiscale CCPP processes. The common design methodology is first introduced. The piece-by-piece devices modeling are then performed following the mass and energy flow. A novel control-oriented hardware-in-the-loop (HIL) simulator for TCS validation is introduced, which helps to reduce costs, mitigate the risk of equipment damage compared to comprehensive and dangerous on-site testing and expedites the commissioning process. Concluding remarks are provided, and future perspectives are pointed out at the end.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Hybrid Data-Driven and Mechanistic Modeling Approaches for Gas Turbine CCPP Control System Validation

  • Ke Wu,
  • Zheng Ji,
  • Jianliang Li,
  • Lei Huang,
  • Zijing Chen

摘要

The utilization of heavy-duty gas turbines in power plant, coupled with a combined cycle system (CCPP), represents a significant alternative to conventional coal-fired boiler turbine units, primarily due to its swift load response capabilities and environmentally benign emissions profile. In China, the gas turbines control system (TCS) have been predominantly supplied and stringently regulated by international vendors. The domestically developed TCS needs extensive validation prior to its commercial deployment. The adequate fidelity model of the CCPP to represents both static and dynamic characteristics plays a critical role in this process. Hybrid modeling provides a promising option to tackle such complex problems. In hybrid modeling, the process-related principles are represented by mechanistic models, while the coefficients added to the first-principal models as adjustments to present unknown process characterizes, are described by data-driven models. This article highlights the significance of hybrid modeling in multiscale CCPP processes. The common design methodology is first introduced. The piece-by-piece devices modeling are then performed following the mass and energy flow. A novel control-oriented hardware-in-the-loop (HIL) simulator for TCS validation is introduced, which helps to reduce costs, mitigate the risk of equipment damage compared to comprehensive and dangerous on-site testing and expedites the commissioning process. Concluding remarks are provided, and future perspectives are pointed out at the end.