基于量子优化的铝电解多目标控制系统理论研究

The theoretical research on multi-objective control system for aluminum reduction based on quantum optimized

针对电解铝生产节能降耗,提高能源利用率的需求,根据铝电解控制过程复杂,变量多等特点,提出了一种铝电解多目标优化控制策略。建立了基于Elman神经网络的铝电解直流电耗与电流效率的预测模型,铝电解多目标优化控制系统,采用BP神经网络控制器用以提高控制精度和动态跟踪精度,以及用量子优化的方法对BP神经网络控制器进行了优化。仿真结果表明基于量子优化算法的控制系统相比于改进型非劣分类遗传算法(NSGAII)优化的控制系统,理论上在相同的电流效率的情况下,直流电耗会明显降低。

Aiming at the demands of energy saving and consumption reduction and improving energy efficiency for aluminum reduction production, a multi - objective optimization control strategy is proposed in this paper in the light of complex control process and many variables of aluminum reduction produc- tion. The prediction model of DC power consumption and current efficiency of aluminum production and multi - objective control system for aluminum re- duction are constructed based on Elman neural network. BP neural network controller is adopted to improve the control accuracy and dynamic tracking ac- curacy and BP neural network controller is optimized by the approach of quantum optimization. The results show that the control effect based on quantum optimization control system is better than the control system based on NSGA II and the DC power consumption of aluminum reduction production using the multi -objective control system is significantly decreased at the same current efficiency in theory.