Research On Control Strategies Of Electrode Regulator System For Electric Arc Furnace

Electric arc furnace is very important in steel making and the electrode regulator system is the key link of electric arc furnace. However, electrode regulator system is a complex system with high nonlinearity, parameter time-varying, variable coupling and strong random disturbance. The rapid and accurate electrode position control is one of the key factors of saving electric energy, shortening smelting period, reducing energy consumption and prolonging lining life. Therefore, the research on modeling and control of electrode regulator system for electric arc furnace has the important theoretical significance and application value. The main work of this dissertation is given as follows:Based on the analysis of the parts of electrode regulator system, the mathematical models of electric arc, hydraulic system and power supply system are established. Then, the electrical system model that is composed of the power supply system model and three electric arc models is put forth. By using the relation between the position displacement and arc length, the electrode system model can be established, which lay the foundation for the controller design of electrode regulator system.In order to solve the parameter time-varying problem in smelting process, an approximate model is derived via Taylor expansion, and the adaptive approximate model control law is realized using normalized radial basis function neural networks (NRBFNN). The robustness of the stability is established by the Lyapunov method and the proposed nonlinear controller is verified by computer simulations.The decoupling between the three phase is achieved according to the inverse model method. Firstly, an approximate model is build by means of Taylor expansion technical. And the approximate control law can be derived for elimination the coupling in the electrical controller for the three phase arc furnace. In addition, a nonlinear compensation is introduced to internal-mode configuration to ensure the system robustness in case of parameter variation and external disturbance. The robustness of the proposed controller is given. According to the question that the parameter variation and the coupling between three phases are exists simultaneously, a direct adaptive neural network controller is proposed. An equivalent model in affine-like form is derived for the original electric arc furnace generalized plant. Then, feedback linearization adaptive control is implemented based on the affine-like equivalent model. Three phase control signals are computed directly and the complete decoupling can be achieved. The weights of the neural networks are directly updated online based on the input-output measurement and the robustness of the system can be ensured when parameter variation and external disturbance. The effectiveness of the controller is illustrated by the computer simulations.All above control algorithms are based on the electric arc furnace generalized plant. However, hydraulic system can be approximated as a linear system. In order to use this feature, arc length is considered as state variable to establish system model in state space form. After constructing the mathematical model of the system in state space form, the approximate model is derived via the Taylor expansion. From the approximate model, we directly derive the inverse control law for eliminating the coupling in the electrical controller for the three-phase arc furnace. This avoids a large amount work in computation required in the online identification of the inverse model. A neural network model based extended Kalman observer is used to estimate the states as not all states are accessible. In addition, uncertainty compensation in the internal model structure is introduced for the robustness of the system. The stability is established by using the Lyapunov method. The proposed decoupling control strategy based on approximated model has some advantages such as simple computation, high robustness and easy implementation.