金氰化浸出过程自适应优化

Adaptive optimization for gold cyanidation leaching process

以某湿法冶炼厂金氰化浸出过程为背景,研究了浸出过程的自适应实时优化问题。针对实际过程中的模型失配问题,提出了一种金氰化浸出过程的修正项自适应实时优化策略,利用实际过程测量值及梯度信息不断修正原优化问题,使其迭代收敛到实际过程的最优设定点。仿真结果表明在过程输出存在适量的测量噪声时,对于模型参数不确定性、结构不确定性以及工况改变3种情况,该方法经过数次的迭代最终都能够收敛到实际过程的最优设定点附近,节约了生产成本,而且不需要模型更新步骤,这为湿法冶金全流程优化控制的顺利实施奠定了重要基础。

The adaptive real-time optimization （RTO） of gold cyanidation leaching process in a hydrometallurgy plant was investigated. To solve plant-model mismatch, an adaptive real-time optimization strategy based on the modifier adaptation method was proposed, and the real plant data and gradient information were used to correct the original optimization problem iteratively to drive its solution to converge to the optimal set point for the plant. The simulation results showed that in the presence of moderate measurement noise and model uncertainty, the iterates based on the proposed adaptive strategy could converge to the optimal set point for the plant after several iterations and moreover the step of parameter estimation was not necessary, laying an important foundation for the successful implementation of the plant-wide optimization and control for hydrometallurgy process.