文献类型：期刊文献

英文题名：An adjustable robust optimization model under dynamic informer-based framework for industrial renewable energy systems

作者：Ma, Aoqun[1,2,3];Shen, Feifei[3];Li, Zhi[2,3];Peng, Xin[2,3];Zhong, Weimin[1,2,3]

机构：[1]East China Univ Sci & Technol, State Key Lab Chem Engn & Low Carbon Technol, Shanghai 200237, Peoples R China;[2]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China;[3]East China Univ Sci & Technol, Sch Informat Sci & Engn, Shanghai 200237, Peoples R China

年份：2025

卷号：197

外文期刊名：PROCESS SAFETY AND ENVIRONMENTAL PROTECTION

收录：;EI(收录号：20251518197619);WOS:【SCI-EXPANDED(收录号:WOS:001466668300001)】；

基金：This work was supported by the the National Key Research and Development Program of China (2023YFB3307800) , the Shanghai Pilot Program for Basic Research (22TQ1400100-16) , National Natural Sci-ence Foundation of China (62173145, 62303186, 62403201) and Fundamental Research Funds for the Central Universities.

语种：英文

外文关键词：Energy management; Deep learning; Uncertainty; Robust optimization; Integrated renewable energy system

摘要：Introducing renewable energy sources into industrial energy systems is an effective way to reduce greenhouse gas emissions. However, the time-varying uncertainty of renewable energy causes significant high operation cost and extreme scenarios in energy management for large-scale industrial systems. Current optimization methods include extensive uncertain parameters but fail to utilize real-time data, limiting their scalability. Therefore, we propose an adjustable robust optimization method with a dynamic framework based on the Informer prediction model, which captures temporal dynamics for accurate real-time forecasts. Firstly, a hybrid modeling method is applied to build up block models based on the process mechanism. Then the Informer method is applied to derive predictive data information. The proposed method combines hybrid modeling with long-term predictions, dynamically adjusting uncertainty set boundaries using real-time errors and penalty costs. A case study on a practical industrial system shows the method reduces total costs by 4.46 % and constraint conflicts by 53 % compared to traditional robust optimization. Additionally, it lowers test costs by 6820$ and reduces constraint conflicts by 47.6 % compared to Adaptive Robust Optimization. This approach enhances economic performance and reliability in energy scheduling under uncertainty. The proposed framework offers innovative solutions for optimizing the reduction of carbon emissions in industrial energy systems.