【摘要】 针对轧辊偏心问题提出了一种基于热轧带钢轧制力的轧辊偏心自动识别方法。该方法通过实时采集和分析轧制力数据，结合数据预处理、二次采样、多项式拟合等技术，实现了轧辊偏心的在线自动识别。通过计算相邻轧辊周长区间轧制力的相似度，建立了轧辊偏心状态评分模型，为生产现场及时发现轧机轧辊缺陷提供了数据支持。研究结果表明，该方法能够有效识别轧辊偏心状态，为提高轧制产品质量和强化设备维护提供重要参考。更多还原

【Abstract】 This paper proposes an automatic identification method of roll eccentricity based on the rolling force of hot-rolled strip. This method realizes the online automatic identification of roll eccentricity by real-time acquisition and analysis of rolling force data, combined with data preprocessing, secondary sampling, polynomial fitting and other technologies. By calculating the similarity of rolling forces in adjacent roll circumference intervals, a roll eccentricity state scoring model is established, which provides data support for timely detection of mill roll defects in the production site. The research results show that this method can effectively identify the roll eccentricity state, providing an important reference for improving the quality of rolled products and equipment maintenance.更多还原