碱煮过程浸出率预测与优化研究

Prediction and Optimization of Leaching Rate during Alkali Boiling Process

针对钨碱煮过程WO_3浸出率预测困难的问题,建立了动态机理模型与最小二乘支持向量机(LSSVM)相结合的并联混合模型,在该混合模型的基础上,构建了碱煮过程优化模型,将动态浸出问题转化为带约束的优化问题,并以粒子群优化(PSO)算法对优化模型进行求解。仿真结果表明,混合模型预测精度高,优化模型效果好,提高了WO_3浸出率,降低了浸出成本。

A parallel hybrid model combined with dynamic mechanism model and LS-SVM was built to predict WO3 leaching rate during tungsten alkali boiling process. Based on this hybrid model, alkali boiling process optimization model was established to transform dynamic leaching problem into constrained optimization problem. Then optimization model was solved with particle swarm optimization （PSO） algorithm. The simulation results show that hybrid model has advantages of high prediction accuracy, good performance, high leaching rate of WO3, and low leaching cost.