Estimation of feedwater heater parameters based on a grey-box approach

Abstract

The first-principle modeling of a feedwater heater operating in a coal-fired power unit is presented, along with a theoretical discussion concerning its structural simplifications, parameter estimation, and dynamical validation. The model is a part of the component library of modeling environments, called the Virtual Power Plant (VPP). The main purpose of the VPP is simulation of power generation installations intended for early warning diagnostic applications. The model was developed in the Matlab/Simulink package. There are two common problems associated with the modeling of dynamic systems. If an analytical model is chosen, it is very costly to determine all model parameters and that often prevents this approach from being used. If a data model is chosen, one does not have a clear interpretation of the model parameters. The paper uses the so-called grey-box approach, which combines first-principle and data-driven models. The model is represented by nonlinear state-space equations with geometrical and physical parameters deduced from the available documentation of a feedwater heater, as well as adjustable phenomenological parameters (i.e., heat transfer coefficients) that are estimated from measurement data. The paper presents the background of the method, its implementation in the Matlab/Simulink environment, the results of parameter estimation, and a discussion concerning the accuracy of the method.

Gray-box Soft Sensors in Process Industry: Current Practice, and Future Prospects in Era of Big Data

“…Simultaneous Perturbation Stochastic Approximation (SPSA) technique was used for parameter identification. Barszcz et al [57] used a serial GB approach for anomaly detection and control of feed water conditions, i.e., temperature, pressure, flow rate, etc., of a heat exchanger of a coal power plant. ANN was used as a BB model.…”

“…Application Category Process GB Type Target BB Type [55][56][57] Power plant estimation and control "thermal storage tank", "feed water heater/ heat exchanger" serial "temperature profile and the usable energy stored", "anomaly identification", "irradiation angle" "simultaneous perturbation stochastic approximation", "ANN", "fast recursive algorithm" [40][41][42][43][44] Oil and gas processing "estimation and control", "estimation", "estimation and optimization"…”

“…The GB models integrate the WB and BB models for addressing the concerns, i.e., accuracy and intuitiveness, of industrial operators.Applications of GB models are found in a variety of domains in process industries such as design, estimation, control, and monitoring. They are applied across the process industries such as iron and steel making [32][33][34][35][36], food processing [37][38][39], oil and gas processing [40-44], chemical, biochemical, and pharmaceutical [45-54], power plants [55][56][57], water treatment [58][59][60], material processing and energy materials [61-64] and industrial robots [65][66][67].The aim of the current study is to do a comprehensive review of the fundamentals of GB soft sensors, their applications in the process industry and prospects in the era of big data analytics; to the best of the authors' knowledge, no extensive review that addresses all such dimensions is reported in the literature. GB soft sensors application in selected process industries were investigated.…”

A Stacked Denoising Sparse Autoencoder Based Fault Early Warning Method for Feedwater Heater Performance Degradation

“…Generally, these methods can be classified into two major categories: mechanism-model-based methods and data-driven based methods. Barszcz et al [9,10] propose a first-principle four-volume model of FWH in which model parameters are adjusted through operational data and performance indicators are proposed. Zhao et al [11] use GSE software to build a coal-fired power plant simulation model including FWH.…”

Rational Transfer Function Model for a Double-Pipe Parallel-Flow Heat Exchanger

“…The other approach consists of using data-based techniques, also known as black-box modeling or process identification [16][17][18], or even in using artificial neural networks [19][20][21]. The so-called gray-box modeling approach, which combines qualitative prior knowledge with quantitative data, can also be used here [22][23][24].…”