基于GAF-ResNeXt的偏心辊式破碎机状态识别方法  

State identification method of eccentric roll crusher based on GAF-ResNeXt

作　　者：毛逸维 叶龙[2] 吴叶伟 刘万华[2] 王帅 热孜艳木·吐尔逊 MAO Yiwei;YE Long;WU Yewei;LIU Wanhua;WANG Shuai;Reziyanmu Tursun(School of Mechanical and Aerospace Engineering,Jilin University,Changchun 130025,China;Northern Heavy Industry Group Co.,Ltd.,Shenyang 110000,China;College of Biological and Agricultural Engineering,Jilin University,Changchun 130025,China)

机构地区：[1]吉林大学机械与航空航天工程学院,吉林长春130025 [2]北方重工集团有限公司,辽宁沈阳110000 [3]吉林大学生物与农业工程学院,吉林长春130025

出　　处：《重型机械》2024年第4期55-59,共5页Heavy Machinery

基　　金：国家自然科学基金项目(No.52375098);沈阳市揭榜挂帅项目“偏心辊式破碎机关键技术攻关”(No.22316120)。

摘　　要：偏心辊式破碎机需要在恶劣的环境下长时间连续作业,因此进行偏心辊式破碎机的状态识别对保证破碎机安全、稳定、高效的运行具有重大意义。基于试验数据建立了偏心辊式破碎机工况数据集,设计了基于GAF-ResNeXt的偏心辊式破碎机状态识别方法,并通过数据集验证了识别方法的有效性。Eccentric roller crushers need to operate continuously for long periods in harsh environments,making the state recognition of eccentric roller crushers crucial for ensuring their safe,stable,and efficient operation.This paper establishes an operational condition dataset for eccentric roller crushers based on experimental data,designs a state recognition method for eccentric roller crushers using GAF-ResNeXt,and validates the effectiveness of the recognition method through the dataset.

关 键 词：偏心辊式破碎机 格拉姆角场 卷积神经网络 状态识别 

分 类 号：TD451[矿业工程—矿山机电]