Aluminum Electrolysis Multi-objective Control System Based On Quantum Optimized

In current, our country suggests requirements that we should establishment resource-saving society and environment-friendly society to reduce the concentration of PM2.5, saving energy and improve overall energy efficiency. As aluminum electrolytic industry is high energy consumption, heavy pollution industries, it is the focus of optimal adjustment. It is an important way of aluminum electrolytic industrial energy conservation by energy saving control and optimal control for aluminum electrolytic industry. As aluminum electrolytic control system is a complex multi-variable, non-linear control system, there are many variables to impact of energy efficiency and the current aluminum electrolysis. How to adopt multi-objective optimization control strategy to achieve energy saving effect is the focus of research.Firstly, we analysis the technological characteristics of the aluminum electrolysis comprehensively, then we proposed multi-objective optimization control strategy of aluminum electrolysis according to the aluminum electrolysis process is complex, variable and more features. Aim to material and energy balance model, combined with the main process parameters and analysis of the main process variables affect aluminum electrolysis, we construction of the aluminum electrolysis process mechanism model. Since the mechanism model is based on ideal conditions, so the error is larger in practical application, the control effect is not work well. Therefore, a prediction model based on Elman neural network is established.We constructed the overall structure control system of multi-objective optimization of aluminum electrolytic. We use a BP neural network controller to improve control accuracy and dynamic tracking accuracy. However, due to the inherent shortcomings of BP neural network, we need to be further improved. Therefore, we presented by quantum optimization method for multi-objective optimization problem. In the proposed algorithm, chromosomes are comprised of quantum bits encoded by real number, and chromosomes are renovated by quantum rotating gates. A quantum selection, a quantum crossover operation, a quantum mutation are applied to evolve the population. A strategy which can drop as the index level and adjust dynamic circularly is proposed in the aspect of qubit phase incremental space; a s election sort to construct non-dominated set is introduced. A diversity preservation strategy based on P areto individual gravitational is proposed. These improvements make the algorithm to optimize efficiency improved significantly. The test results show that the algorithm can solve multi-objective optimization problem better. It has strong global search capability compared to other classical multi-objective optimization algorithm.In order to make aluminum electrolytic multi-target control system to produce better output, we use the improved algorithm optimize BP neural network and proposed specific improvement program, elaborate specific implementation steps quantum multi-objective genetic algorithms with BP neural network integration. We optimize BP neural network controller of weights, thresholds and learning modalities by quantum multi- objective genetic algorithm, give full play to the global search ability quantum multi-objective genetic algorithm to improve the convergence rate and generalization ability of neural networks.Finally, we verify the effect on t he output of aluminum electrolytic multi-target control system by computer simulation. To compare the effect in two target output by employing a BP neural network optimization of aluminum electrolytic multi-objective optimization control system based on NSGAII. The simulation results show that the quantum-based multi-objective optimization algorithm electrolytic control system, at the same current efficiency, the DC power consumption significantly decreased. Therefore, we can conclude that the control system has progressiveness in terms of energy saving aluminum electrolysis and the control strategy feasibility of the proposed aluminum electrolytic multi-objective optimization in this paper.