DNN-Based ADNMPC of an Industrial Pickling Cold-Rolled Titanium Process via Field Enhancement Heat Exchange

Hezhou Su,Biao Yang,Shenghui Guo,Shimin Zhang,Huilong Luo,Youling Li,Jinhui Peng,Wei Li,Yaming Shi,Zhuming Zhang

doi:10.1155/2015/292470

Abstract

The dynamic neural network based adaptive direct nonlinear model predictive control is designed to control an industrial microwave heating pickling cold-rolled titanium process. The identifier of the direct adaptive nonlinear model identification and the controller of the adaptive nonlinear model predictive control are designed based on series-parallel dynamic neural network training by RLS algorithm with variable incremental factor, gain, and forgetting factor. These identifier and controller are used to constitute intelligent controller for adjusting the temperature of microwave heating acid. The correctness of the controller structure, the convergence, and feasibility of the control algorithms is tested by system simulation. For a given point tracking, model mismatch simulation results show that the controller can be implemented on the system to track and overcome the mismatch system model. The control model can be achieved to track on pickling solution concentration and temperature of a given reference and overcome the disturbance.

Highlights

Pickling plates and strips are normally done in a continuous way, just by drawing the plates and strips from rolls through a cascade of picking tanks [1, 2]
The variable gain, variable forgetting factor, and resetting RLS algorithm based DNN are used to predict the temperature of acid solution in prerinsing bath (T1), in three pickling baths (T2, T3, and T4), and in rinsing bath (T5), and adaptive direct nonlinear MPC (ADNMPC) strategy is utilized to adjust these temperatures to the desired set point by adjusting input electrical power of magnetrons (Ua) as shown Figure 1 and (2)
Using these values of industrial microwave heating devices (IMHD), an open-loop simulation of the acid pickling process based on its validated model of differential equations (2)–(6) obtained from first principles was performed with a sampling interval of 1 min to obtain input-output data pairs for the neural networks (NN) training, while these validation pairs were obtained from the actual acid pickling process

Summary

Introduction

Pickling plates and strips are normally done in a continuous way, just by drawing the plates and strips from rolls through a cascade of picking tanks [1, 2]. Daosud et al [11] used neural network inverse model-based controller for the control acid concentration to be maintained at the optimum value They reported the robustness of the proposed controller showing superiority when controlling such chemical manufacturing processes that have the distributed, highly nonlinear dynamic behavior, unmodeled dynamics, and dead. To solve the microwave heating acid solution for pickling the metal products, which, as described above in literature [10,11,12, 17], is a grossly nonlinear process of unmodeled dynamic, multivariable in nature interactions between baths, distributed processes, and uncertain and time-varying parameters which cause this process to be difficult to control by conventional controllers, it is necessary to model accurately nonlinear dynamical systems and to control efficiently strategy. The main difference between the DNN-ADNMPC and the GPC is that the former uses a nonlinear NN model to identify and control microwave heating pickling process directly whereas the latter utilizes a linearized form of a nonlinear NN model to identify and control microwave heating pickling process

Description of a Cold-Rolled Titanium Pickling Process

DNN Model Identification and ADNMPC Control Strategy

Simulation Results and Discussion

Conclusions