针铁矿法沉铁过程双层结构优化控制

针铁矿沉铁过程是锌冶炼过程中一个非常重要的环节,其中最重要的是控制氧气添加量,因此本文提出一种针铁矿沉铁过程双层结构优化控制方法.上层定义氧气利用率衡量理论消耗量与实际添加量的差别,以过程氧气利用率最高为目标优化设定级联反应器出口二价铁离子浓度下降梯度,下层以过程氧气消耗最少和出口离子浓度与上层设定值误差最小为优化目标,过程动态模型和工艺条件为约束,求解构造的非线性优化控制问题得到各反应器最优氧气添加速率.为减少不确定性干扰对系统的影响,采用一种模型参数自适应校正的方法对模型参数进行校正保证优化控制器的性能.最后根据过程离子浓度采样值计算过程实际氧气利用率作为上层优化参数重更新反应器出口二价铁离子浓度最优设定值.由于下层优化问题约束多且约束多呈非线性,采用Legendre伪谱法求解下层优化问题.仿真结果表明,所提出的双层结构优化控制方法能实现过程准确控制,减少过程氧气消耗.

Two-layer optimal control for goethite iron precipitation process

The goethite iron precipitation process is a very important part of the zinc hydrometallurgy process. Excess oxygen is usually added in the actual process to achieve iron removal, which will directly lead to massive waste of oxygen. Therefore, this paper presents a two-layer structure optimization control strategy based on Legendre pseudospectral method for goethite iron precipitation process. For the upper layer, oxygen utilization efficiency(OUE) in the process is firstly defined to measure the difference between the theoretical oxygen amount and its actual amount. On this basis, maximum OUE is the goal for optimizing the descent gradient of outlet concentration of ferrous ions. In the lower layer, least oxygen consumption and minimum error between outlet concentration of ions and the set value is the goal, and process dynamic model and process conditions are the constraints, optimal oxygen addition rate of each reactor is calculated. To overcome the impact of frequent fluctuations in production conditions on the control performance, a parameter adaptive correction approach is adopted. Finally, the actual OUE is used as the upper optimization parameter to update the outlet concentration of ferrous ions. Legendre pseudospectral method is adopted to calculate the optimization problem because of the multiple nonlinear constraints in the lower layer. Simulation results show that the two-layer structure optimal control method can dynamically adjust the oxygen addition rate to reduce the oxygen consumption, and Legendre pseudospectral method has higher precision and efficiency for multiple nonlinear constraints.