复杂石化过程数据和领域知识融合的能效评价与系统优化理论及关键技术

The complex petrochemical industry is one of high energy consumption industries in China, which has characteristics of large production scales, long process and complex productions. The factors that affect the energy consumption, conversion and utilization are extremely complex. How to improve the energy efficiency, reduce the energy consumption cost and improve the production efficiency are the problems urgent to be solved. So the research of energy efficiency evaluation and system optimization has important scientific significance and application prospect. This project analyzes the complex logical relations of the material stream, the energy stream, the information stream and the value stream in the complex petrochemical process from the perspective of the global and the system. And this project studies formal knowledge expression and the extraction method of the energy efficiency process system. The Multi-level functionalization, Quantitative and logical intelligent modeling method based on the integration of data and domain knowledge is proposed to build the superstructure model of energy efficiency process complex network. Meanwhile, studying the energy efficiency analysis and evaluation methods, which establish the multi-level, multi-dimensional and multi-index evaluation system of energy efficiency to identify the system energy efficiency under different indicators bottlenecks and diagnose the operation efficiency and energy saving potential. In addition, in order to study the coordination optimization between the local, the local and the overall of the superstructure model, a multi-level, large-scale and multi-objective optimization method of energy efficiency is proposed. Furthermore, the whole process optimized resource configuration, installation and unit operation scheme of coordinated operation are obtained. The proposed theory and methods are verified by application in a complex ethylene production process, which provide theoretical basis and method guidance for the energy saving and reducing consumption and optimizing the operation of the petrochemical industry.

石化工业是我国的高能耗行业之一，其生产规模大、流程长，能源消耗、转化和利用过程的工艺、影响因素及相互关系极为复杂，如何提高能源效率、节能降耗、提高生产效率是迫切需要解决的问题，研究过程能效评价与系统优化具有重要的科学意义和应用前景。本项目从全局系统分析过程物流、能流、信息流和价值流的复杂逻辑关系，研究过程系统的知识表示与提取方法；提出一种基于数据和领域知识融合的多层次功能化、定量化、逻辑化的过程系统超结构智能建模方法；研究过程系统能效评价方法，建立多层次、多维度、多指标能效评价体系，识别系统能效瓶颈、运营效率与节能潜力；研究超结构模型局部、整体之间的协同优化，提出一种多目标大规模多层次协同的能效优化方法，以及资源优化配置、装置与单元协同运行的操作方案。以乙烯生产流程为应用对象，验证所提理论和方法的有效性与普适性，为石化工业节能降耗与优化运行提供理论依据和方法指导。

为有效实现石化过程节能降耗、提高能源效率和生产效率，本课题从分析石化过程物流、能流、信息流和价值流之间各种交互行为与复杂逻辑关系出发，研究复杂石化过程数据和领域知识融合的能效评价与系统优化理论及关键技术。基于石化过程运行大数据与机理的有效融合，提出了面向石化过程多元信息处理、知识表述与提取方法；深入理解石化过程单元-装置-系统的复杂结构，提出了面向复杂石化过程用能系统的多层次功能化、定量化、逻辑化的超结构建模方法，建立了具有过程行为表达能力强，又具有高效计算能力的能效分析模型。基于复杂石化过程产、供、用能系统及目标，考虑资源、经济和环境排放等方面构建了能效、碳排、能源强度、活动强度、结构指标等的多层次、多维度、多指标能效评价体系；基于DEA交叉模型，使用能源环境性能指数表示不同决策单元的总体性能，提出了一种改进的环境DEA交叉模型的能源效率评价方法，构建了过程单元-装置-系统不同层次的能效综合评价模型，识别不合理的能流分配与节能潜力。综合考虑超结构模型中各指标间关系，建立了面向资源优化配置的多目标优化模型，提出了多种多目标协同综合优化方法和一种基于可扩展P-Graph推理的原料调度优化方法，形成了乙烯生产过程的优化操作与资源配置最佳方案。最后，研发了石化过程能效评价与优化系统，并在三家石化企业多套装置应用，取得了显著的经济效益和社会效益，验证了所提方法的有效性，为石化工业节能降耗与优化运行提供了理论依据和方法指导。整个项目按年度计划顺利开展工作，考核指标全部完成，达到预期目标。组织了4次国际、国内学术会议并做了8次大会邀请报告，发表学术论文90篇，其中SCI收录论文52篇、CAA A类期刊论文26篇、EI收录论文29篇，授权发明专利6项、申请发明专利15项，登记软件著作权2项，3家石化企业应用成效显著，培养出站博士后3人、毕业博士生6人和硕士生18人。