A novel real-time optimization compensation method based on POPOA for the gold hydrometallurgy process

Abstract

The gold hydrometallurgy process has been prevalent in the hydrometallurgical industry. With a further understanding of the process, the process mechanistic model has been constructed for performing control and optimization. However, due to random disturbances and changing operating conditions, the operating performance of the nominal model-based process has changed. To further perform real-time optimization of the gold hydrometallurgy process, a novel real-time optimization (RTO) compensation method based on the process operating performance optimality assessment (POPOA) is proposed in this paper. Firstly, the operating performance of the online process is judged by the process operating performance optimality assessment algorithm. For the case that the process operating performance is the optimal grade, in which the process operating state is close to the optimal state, but it is not in the optimal state due to some small disturbances in the process, the self-optimizing control (SOC) method is realized by taking the combinations of measurements as the controlled variables (CVs), which are tracked at the optimally insensitive set-points. Furthermore, under the condition of the operating performance is non-optimal grade due to the influence of serious disturbances or obvious change of operating conditions, a similarity detection procedure is complemented based on a SOC model base for searching the CVs set-points relative to similar models. When there is no usable set-point, it is necessary to perform the SOC offline modeling to update the model, or to find similar cases in the historical optimal operational case base to realize the compensation operation strategy based on just-in-time (JIT). Specifically, when similar samples cannot be found in the historical SOC base and the optimal operational case base, the plant-wide optimization model must be rebuilt to meet the requirements of the actual industrial production process. Finally, an industrial simulation experiment of the hydrometallurgical production process is carried out. The simulation results verify the effectiveness and practicability of the proposed new RTO compensation scheme.