|导师||梁自泽 ; 景奉水|
|关键词||铝电解 阴极钢棒 窄间隙旋弧焊接 工业机器人 模型预测控制|
|英文摘要|| China is the largest countries producing primary aluminum all over the world. However, the research on basic theory and key technologies of aluminum electrolysis technology in China is weak, resulting in high resource consumption, low energy utilization and copious emission of three wastes. The service life of aluminum electrolytic cells in China is only 4 to 6 years, which is 2 years less than the average life expectancy in foreign countries. Therefore, 20% of electrolyzers need to be overhauled each year. The overhaul quality is of great significance for extending electrolytic cell life and saving the cost of aluminum production. In particular, the quality of the connection between the cathode steel busbars and the steel-aluminum explosive welding plates has a significant impact on economic and technical indicators such as current distribution, power consumption, and current efficiency.|
However, ready-made automatic welding equipment is impossible to apply because of the narrow operating space of the connection joint between the cathode steel busbars and the steel-aluminum explosive welding plates. Currently this work is mainly finished through manual welding. In terriblely bad environments such as high temperatures, strong magnetic fields, high dust, high currents, toxic gases, and high noise, welding workers are of high labor intensity, high work risks, and low work efficiency. Therefore, it has significant theoretical meaning and engineering value to develop automatic welding robot replacing human workers. It can improve production stability, increase level of energy conservation and emission reduction, and promote technological progress in the aluminum electrolysis industry.
In this paper, several key problems of the cathode steel busbar welding robot for electrolytic aluminium cells are studied. The main work are summarized as follows：
1. Aiming at many problems existing in the connection of existing cathode steel bars and steel aluminum explosion welding plates, a new technical solution is proposed: the steel-aluminum explosive welding plate is positioned by the clamp of the cathode steel busbars, and a narrow weld is formed for automatic welding. A narrow gap rotating-arc welding robot system is developed to weld automatically. This technical scheme can save welding materials, improve welding quality, ameliorate current distribution and decrease the safety security threat in production operation.
2. Aiming at the working characteristics of the narrow weld between the cathode steel bars and steel aluminum explosion welding plates, the design scheme of the narrow gap rotating-arc welding robot system was proposed, and the hardware framework, software framework and robot kinematics model of the control system were established. Then several key technical problems have been solved: design rigid fixtures to effectively prevent welding distortion; control the rotating arc at the end to solve the problem of sidewall fusion; apply water-cooled sliders to prevent overflow of molten metal; make comprehensive consideration of mechanic, electricity, water, gas, magnetic, thermal and other processes in welding systems to ensure reliable operation; introduce rotary arc sensing to automatically track the narrow gaps in limited working spaces.
3.Aiming at the problem that the static model of arc welding process is difficult to apply advanced control algorithms to achieve high-precision and high-sensitivity seam tracking, the dynamic model of narrow gap rotating-arc welding process was established，based on the principle of the following dynamic process: welding arc physics, welding wire feeding and melting, power circuit model, welding wire extension, droplet transfer and weld forming. First, a sixth-order nonlinear state space equation of the system model is established. Then, the condition that the wire extension is much longer than that of the droplet radius and the droplet displacement is established, and the system equation is decoupled and reduced to a third-order four-input single-output system. Based on the small-signal linearization model near the static equilibrium point of the system, the welding current response characteristics in the influence of the wire extension and the shape of the molten pool are determined. Finally, a novel quantitative explanation is proposed for the phenomenon of bead deflection in narrow gap rotating-arc welding. In addition, based on this principle, the optimal torch rotation control strategy is planned.
4. Aiming at the problem of strong coupling, high order, disturbance and other factors in the narrow gap GMAW process of rotating arc, a receding horizon optimization strategy for open-loop voltage and welding torch height is proposed, and a predictive control system based on the assumption of states measured is designed. Firstly, a system state space disturbance model is obtained. Then, the model is discretely and dimensionlessly and incrementally processed. Fisrt, deduce predictive model of the narrow gap rotating-arc GMAW process in the framework of the state space model. Then the mathematical description and optimal control input sequence of the predictive control optimization problem are obtained. The designed predictive controller is with a “feedforward-feedback” structure. Finally, the closed-loop solution of the predictive control system is discussed when considering the influence of unmeasured disturbances. And it is of asymptotic stability and no tracking error characteristic. Then, the simulation verified that the closed-loop predictive control system designed under various perturbations of the wire feeding motor armature voltage impulse and positional slope deviation can operate stably, and there is no static difference.
5.Aiming at the unmeasurable state of narrow gap rotating-arc GMAW process system, a feedback correction strategy for estimating the state of the welding system is proposed, and a predictive control system based on state estimation is designed. First，under the condition that the system is observable, the optimal control input sequence of closed-loop predictive control system based on state observer is deduced. Specifically discuss the closed-loop system dynamic response characteristics of several cases: the pole is not configured, the pole is configured in the unit circle, the pole is configured close to the origin of the situation. Finally, the simulation verified that the closed-loop predictive control system based on the two disturbances of the armature voltage impulse and position slope deviation can eliminate the initial value deviation of the state estimation.
6. Aiming at the working environment of the aluminum electrolysis industry, a narrow gap rotating-arc welding robot system was developed. The proposed control strategy and robot system were verified by experiments. The robot is able to complete the welding of a single joint in 15 minutes, which is a significant increase over the previous 1-2 hours of manual labor. The robot welding joint voltage drop is reduced to 7mV, which is significantly lower than the artificial pressure drops of 10-30mV. And the welding quality consistency and stability are better than manual welding.
|毛一剑. 铝电解槽阴极钢棒窄间隙旋弧焊接机器人的关键技术研究[D]. 北京. 中国科学院大学,2018.|