Recursive Learning-Based Bilinear Subspace Identification for Online Modeling and Predictive Control of a Complicated Industrial Process

Abstract

In this paper, a recursive learning based bilinear subspace identification (R-B-SI) algorithm is proposed for online modeling and data-driven predictive control of blast furnace (BF) ironmaking process with strong nonlinear time-varying dynamics. Different from the existing linear SI algorithms, the R-B-SI algorithm can make full use of the process data information by adding the Kronecker product term of input data and the Kronecker product term between input and output data into the data block Hankel matrices. Moreover, the parameters of the R-B-SI model can be updated online with the latest data using recursive learning. In order to adjust the sensitivity of the algorithm to the data at different sampling times, the forgetting factor is introduced into the involved recursive learning as well. Therefore, the nonlinear time-varying dynamics of process can be well captured and described by the R-B-SI model. As a result, the data-driven predictive controller designed by the R-B-SI prediction model with parameter adaption is always capable of achieving satisfactory control performances regardless of the variation in various process operating conditions. Data experiments using actual industrial data have verified the practicability and superiority of the proposed methods. INDEX TERMS Recursive bilinear subspace identification, online modeling, recursive learning, data-driven predictive control, blast furnace ironmaking process, nonlinear time-varying dynamics.

Fault diagnosis and fault‐tolerant control for bilinear stochastic distribution systems with actuator fault

“…are the control input vector, the weight output vector and the state vector of the system, respectively, 𝛾(y, u(k)) is the output PDF, h is the dimension of the input. The matrices A, B, C, N i are the system matrices to be identified, which are identified by bilinear subspace identification 28 in this paper.…”

Historical overview of refractory lining in the blast furnace

“…The melting points of some of the most commonly employed metals and refractories are depicted in Figure 2. [17,18]. As shown in Figure 3, if melting point is the only criterion, a range of refractory materials are suitable.…”

A Collaborative Optimization Strategy for Energy Reduction in Ironmaking Digital Twin

“…In Figure 12, it is shown that ACF and PACF of #1 BF are terminated within the confidence band at 5 and 2 orders respectively. Therefore, the prototype of this energy model for #1 BF can be set as ARMA (2,5). 2000 samples of seven BFs have been captured from an ironmaking plant with a production of ten million tons per year.…”

“…To enhance the transparency of soft-margin support vector machine, Chen et al proposed a novel algorithm to solve the shortcomings of black box model in BF [4]. Be-sides, Zhou et al proposed a bilinear subspace identification algorithm based on recursive learning to learn the nonlinear time varying dynamics of a BF [5]. Moreover, Li et al judged the transformation tendency of silicon content in hot metal by fuzzy classifier [6].…”

“…Width represents the predicted coke ratio and radius represents the practical one. (c) Bottom plot is the out-of-sample prediction performance of ARMA(2,5).…”