Nonlinear Successive Compensation Method for Plant-Wide Production Index Based on Industrial Data

Due to the complexities of high-dimensionality, nonlinearity, many constraints, and process model mismatch, the plant-wide production index needs to be compensated when the production index is applied to the actual production process. This paper proposes a nonlinear successive compensation method for the plant-wide production index based on industrial data. Based on the set point of the production index obtained from the plant-wide optimization, the method uses the large amount of industrial data stored in the industrial process to establish the relationship between production index compensation and the increase of economic benefit, optimizing the solution value of the production index compensation that maximizes the economic benefit increment. Through iterative compensation, the method gradually improved the economic benefit of the production process. The proposed method was applied to the hydrometallurgical production process of a refinery. The results verified the effectiveness of the proposed method.