基于两层分块GMM-PRS的流程工业过程运行状态评价

过程运行状态评价旨在实时判断运行性能优劣程度，并追溯导致非优运行状态的原因，指导操作人员进行生产调整，保证企业经济效益.因此，对过程运行性能优劣评价的研究具有重要的理论和应用价值.本文针对定量、定性变量共存的流程工业过程运行状态评价问题，提出基于两层分块混合模型的评价方法.将流程工业过程根据其物理特性和管理方向划分子块，产生子块层和全流程层.在定量信息占主导地位的子块内，建立定量的高斯混合模型（Gaussian mixture model，GMM）.在定性信息占主导地位的子块内，建立定性概率粗糙集（Probabilistic rough set，PRS）模型.综合各子块运行状态信息，进一步判定全流程运行状态等级.针对非优运行状态等级，本文提出基于贡献率的非优原因追溯方法，在非优子块内进行原因追溯.最后，将所提方法应用于某黄金湿法冶炼生产过程，说明所提方法的可行性和有效性.

Plant-wide Process Operating Performance Assessment Based on Two-level Multi-block GMM-PRS

Process operating performance assessment judges operating performance optimal degree online, and identifies the causes for non-optimal performance to guide the production adjustment for operators. Therefore, the research on process operating performance assessment is of great significance in both theory and practical applications. To solve the plant-wide process operating performance assessment problem with coexistence of quantitative and qualitative variables, a two-level multi-block hybrid model based assessment approach is proposed in this article. According to the physical property and management direction, a plant-wide process is classified into multiple sub-blocks. Hence, there are sub-block level and global level. In a sub-block that is dominated by quantitative information, a Gaussian mixture model (GMM) is established. Accordingly, in a qualitative information dominated sub-block, a probabilistic rough set (PRS) is built. Based on the sub-block-level performance grade information, the global-level performance grade can be further judged. For the non-optimal performance grade, cause identification is implemented in the non-optimal sub-block. A contribution rate based non-optimal cause identification method is developed in this research. In the end, to illustrate the feasibility and validity, the proposed technique is applied to a gold hydrometallurgy process.