This project belongs to the field of automatic control science and technology. Resource constraints have caused the types of imported crude oil to become more complex and diverse. The intensification of environmental protection pressure has led to the continuous acceleration of oil upgrades, which has brought tremendous pressure to my country's refining and chemical enterprises. Achieving low-cost quality upgrading of oil products is an urgent problem for refining and chemical enterprises. How to extract oil chemical structure information, form accurate real-time detection and characterization of oil characteristics, and how to break through the constraints of uncertainty, non-linearity and other factors in the production process, and achieve accurate characterization and online intelligent optimization of the oil blending process. Regulation is an inevitable choice and huge challenge for the automation, intelligence and green construction of oil production and manufacturing in refining and chemical enterprises. The real-time characterization of oil characteristics and online blending optimization control technology developed in this project breaks the technical monopoly of foreign companies in the field of real-time characterization of oil characteristics and online blending, and innovates many advanced technologies for the construction and online application of non-linear models of oil blending., realizing the development and application of a series of intelligent optimization and control technologies for crude oil and refined oil blending. The core innovation work of the project is as follows: (1) A high-precision, intelligent and practical oil blending effect model and its construction method were invented. Completely different from the traditional modeling method of multi-parameter nonlinear equations, a modeling method for compensating the blending effect for the nonlinear properties of blended component oils is proposed. Based on the positive and negative effects of each component oil on different blending properties, the model accuracy is effectively guaranteed based on the linear model structure; Combined with the improved variable blending effect and online rolling correction technology, the nonlinear characteristics of key oil properties such as RON and DON are accurately characterized, laying a foundation for the reliable application of the online blending optimization control system. (2) Invented and implemented an online formulation optimization technology based on the blending effect model. Aiming at the bottleneck problem that it is difficult to obtain oil attribute information in real time during the blending process and it is difficult to apply formula optimization technology online, an online detection and real-time characterization technology for oil characteristics has been developed. Based on near infrared and Raman spectroscopy analysis technology, agile sensing of oil molecular spectral information and real-time characterization of key characteristics are realized; with the optimization goals of minimizing mass excess, minimizing blending cost and minimizing blending process time, multi-objective and multi-constraint collaborative evolutionary optimization and other algorithms achieve multi-time domain global online optimization of the oil blending process. (3) Invented and implemented a complete set of technologies for precise control of multiple indicators and industrial system integration in the oil blending process. Invented automatic switching of gasoline pipeline blending scheduling batches, gasoline pipeline blending additive control and interlocking methods, and a low-pressure drop efficient static mixing device that is conducive to oil blending. It integrates multi-agent collaborative control technology to realize the oil blending process. Intelligent scheduling execution, accurate measurement and control of additive additives, efficient oil mixing and online intelligent control of blending formulas; Develop a series of industrial application software for oil blending synthesis that integrates real-time characterization of oil characteristics, blending effect models, optimal control, and mobile monitoring. This achievement has been successfully applied in the gasoline, diesel and aviation fuel blending units of Jinling Petrochemical, the country's largest clean gasoline and aviation kerosene producer, since 2009, and has been successfully promoted and applied to the crude oil and gasoline blending units of Zhenhai Refining and Chemical Co., Ltd., the company with the largest annual refining capacity in China. The Sinopec Appraisal Committee believes that ldquo; the project has independent intellectual property rights, and the overall technology for online optimization and control has reached the international advanced level. Among them, core technologies such as blending effect model and online optimization of blending formulas have reached the international leading level. amp;rdquo; From 2016 to 2018, the cumulative new output value of application enterprises was 1.651 billion yuan, new profits were 1.261 billion yuan, new taxes were 315 million yuan, and expenditures were saved by 75.05 million yuan. It has authorized 5 national invention patents and 1 utility model, disclosed and applied for 12 national invention patents, and registered 5 software copyrights.