On September 6, 2026, invited by the Cross-Media Big Data Joint Laboratory and IEEE CIS Shandong Chapter, Professor Zhan Zhihui — an IEEE Fellow, national high-level talent and professor at the College of Artificial Intelligence, Nankai University — visited the College of Science, China University of Petroleum (East China), and delivered a special academic report titled Data and Knowledge-Driven Evolutionary Computation. Relevant faculty members, master’s and doctoral students from the laboratory attended the report.
During the report, Professor Zhan followed the development trajectory of the evolutionary computation field, and systematically analyzed the practical bottlenecks of traditional evolutionary algorithms in solving complex optimization problems, such as low search efficiency and insufficient interpretability. Drawing on his team’s years of research achievements, he focused on expounding the core ideas of two technical routes: data-driven and knowledge-driven. By mining the inherent distribution characteristics of optimization problems based on large-scale real data, and integrating domain prior knowledge to guide the direction of population evolution, the approach opens up a collaborative path between data statistical laws and expert knowledge and experience, so as to enhance the ability of evolutionary algorithms to handle high-dimensional, multi-objective engineering optimization problems with complex constraints.
Professor Zhan introduced a series of innovative models of data-knowledge dual-driven evolutionary computation, demonstrated practical application cases of the algorithms in scenarios including industrial optimization, resource scheduling and intelligent decision-making, compared and analyzed the advantages and applicable boundaries of algorithms with different paradigms, discussed future research directions under the cross-integration of evolutionary computation with large models and agent technologies, and pointed out that embedding domain knowledge into the evolutionary iteration process is an important breakthrough to solve complex real-world optimization problems.
During the interactive exchange session, the attending faculty and students actively raised questions on topics such as algorithm parameter design, adaptation of optimization scenarios in the oil and gas field, and the combination of evolutionary computation and physics-informed models. Professor Zhan answered the questions in detail, provided guiding suggestions on research topic selection and scientific research innovation ideas, and encouraged students to conduct algorithm research based on interdisciplinary integration and oriented to the real needs of the industry.
This report broadened the academic horizons of the faculty and students in the field of intelligent optimization computing, and provided beneficial inspiration for the laboratory’s scientific research and personal development in related directions such as oil and gas artificial intelligence and intelligent computing.
[Editor:Ma Xiangyi, Zhang Yaoguang]