基于混合优化策略的拜耳法物料平衡计算

Bayer Material Balance Computation Based on Hybrid Optimization Strategy

首先分析了氧化铝生产工艺中物料平衡计算的特点,介绍了拜耳法生产工艺流程,并给出了拜耳法物料平衡计算的数学模型,接着针对传统的优化算法存在易陷入局部最优导致的收敛速度慢,精度低等问题,提出以模拟退火—单纯形法作为混合优化策略用于拜耳法物料平衡计算。最终的计算结果表明,混合优化策略有效地避免了原有两种算法的缺陷,具有较强的全局搜索能力和较高的收敛精度,其性能远远优于单一优化方法,是进行拜耳法物料平衡计算的有效方法。

This paper firstly analyzes the characteristic of the material balance computation （MBC） in alumina production, and Bayer process is briefly introduced. Then, Math model of MBC is presented. On the basis of analyzing synthetically the advantage and disadvantage of simplex method and simulate annealing algorithm, an effective hybrid optimization strategy combining simulated annealing with simplex method（SMSA） is proposed for solving material balance computation. The practical results of the MBC show that SMSA abstains effectually the disadvantages of original methods, and has higher globally convergence ability than those of single methods with both a better stability and a steady convergence. Most importantly, results demonstrate that SMSA is more feasible and efficient in MBC.