Estimation of time-dependent thermal boundary conditions and online reconstruction of transient temperature field for boiler membrane water wall. (February 2020)

• Record Type: Journal Article
• Authors: Lv, Cai; Wang, Guangjun; Chen, Hong
• Is Part Of: International journal of heat and mass transfer. Volume 147(2020)
• Publication Date: 2020-02
• Subjects: Boiler -- Membrane water wall -- Time-dependent thermal boundary conditions -- Nonlinear heat transfer -- Inverse problem -- Temperature field reconstruction; Heat -- Transmission -- Periodicals; Mass transfer -- Periodicals; Chaleur -- Transmission -- Périodiques; Transfert de masse -- Périodiques; Electronic journals; 621.4022
• DOI: 10.1016/j.ijheatmasstransfer.2019.118955
• Languages: English
• ISSNs: 0017-9310

Abstract

Highlights: An adaptive inverse scheme based on boundary condition transfer is proposed. The scheme can solve the inverse problem with unknown characteristic parameter. The scheme is used to estimate thermal boundary conditions of water wall. Effectiveness of the developed scheme is verified. Abstract: The thermal boundary conditions of membrane water wall include local radiant heat flux on the fire side, fluid temperature inside the tube, and the convective heat transfer coefficient. Obtaining the above thermal boundary conditions has an important impact on the safe operation of modern power station boilers. When the above thermal boundary conditions have time-varying characteristics, the heat transfer system shows obvious nonlinearity. In view of the fact that they are difficult to measure directly and accurately, a multiple model adaptive inverse scheme based on boundary condition transfer (BCT-MMAI) is proposed. In this scheme, the estimation of convective heat transfer coefficient in the tube is transferred to the estimation of adiabatic boundary condition on the back side of water wall, that is, the local radiative heat flux on the fire side, the fluid temperature in the tube and the adiabatic boundary condition on the back side are simultaneously estimated. Firstly, according to the given discrete points of convective heat transfer coefficient (characteristic variable), the non-linear heat transfer system is divided into several linear subspaces, and the temperature prediction sub-model corresponding to each linear subspace is established. Secondly, according to the predicted and measured temperatures at the measurement points, the estimated results of the three variables are obtained by rolling optimization. Further, the weighting coefficient of each linear sub-model is constructed based on the accuracy of inversion results of adiabatic boundary condition on the back side by each linear sub-model. Finally, by weighting the inversion results of each sub-model, the estimation results of local radiant heat flux on the fire side and the fluid temperature in the tube are directly obtained. Meanwhile, the discrete values of characteristic quantity are comprehensively weighted, and the identification results of time-varying convective heat transfer coefficient are indirectly generated.