Dear Editor, This letter proposes a multimodal data-driven reinforcement learning-based method for operational decision-making in industrial processes. Due to the frequent fluctuations of feedstock properties and oper...Dear Editor, This letter proposes a multimodal data-driven reinforcement learning-based method for operational decision-making in industrial processes. Due to the frequent fluctuations of feedstock properties and operating conditions in the industrial processes, existing data-driven methods cannot effectively adjust the operational variables. In addition, multimodal data such as images, audio.展开更多
基金supported by the National Key Research and Development Program of China (2020YFB1713800)the National Natural Science Foundation of China (92267205)+1 种基金the Hunan Provincial Innovation Foundation for Postgraduate (CX2022 0267)the Fundamental Research Funds for the Central Universities of Central South University (2022ZZTS0181)。
文摘Dear Editor, This letter proposes a multimodal data-driven reinforcement learning-based method for operational decision-making in industrial processes. Due to the frequent fluctuations of feedstock properties and operating conditions in the industrial processes, existing data-driven methods cannot effectively adjust the operational variables. In addition, multimodal data such as images, audio.
