基于改进冠豪猪优化算法的DELM磨机负荷预测  

Load prediction of mill based on improved crested porcupine optimizer and DELM

作　　者：石建全 秦敏钦 刘坤 毕文 SHI Jianquan;QIN Minqin;LIU Kun;BI Wen(School of Mechanical Engineering,Nanjing Institute of Technology,Nanjing 211167,Jiangsu,China)

机构地区：[1]南京工程学院机械工程学院,江苏南京211167

出　　处：《矿山机械》2025年第3期28-34,共7页Mining & Processing Equipment

基　　金：江苏省高等学校基础科学研究面上项目(23KJD120005);教育部“春晖计划”项目(HZKY20220150)

摘　　要：深度极限学习机(Deep Extreme Learning Machine,DELM)的泛化能力在权重输入和隐藏层配置方面仍然存在优化空间。为提高磨机负荷预测的准确性和实用性,提出一种基于改进冠豪猪优化算法的DELM回归预测模型(CTCPO-DELM)。根据算法的收敛性,通过改进冠豪猪优化算法的防御机制,利用柯西逆累积分布函数和正切算子调整模型最佳输入权值,提高模型泛化寻优能力和预测精度。利用UCI标准数据集和磨机负荷数据集,将CTCPO-DELM模型与WOA-DELM模型和CTCPO-DELM模型进行对比试验,证明了CTCPO-DELM模型的有效性。There is still room for optimization in the generalization ability of deep extreme learning machine(DELM)in terms of weight input and hidden layer configuration.In order to improve the accuracy and practicability of the load prediction of mills,a DELM regression prediction model(CTCPO-DELM)based on the improved crested porcupine optimizer was proposed in this paper.According to the convergence of the algorithm,the defense mechanism of the improved crested porcupine optimizer,the Cauchy inverse cumulative distribution function and the tangent operator were used to adjust the optimal input weight of the model,so as to improve the generalization optimization ability and prediction accuracy of the model.Then,using the UCI standard dataset and the mill load dataset,the CTCPO-DELM model was compared with the WOA-DELM model and the CTCPO-DELM model,which proved the effectiveness of the CTCPO-DELM model.

关 键 词：磨机 深度极限学习机 冠豪猪算法 柯西逆累积分布 正切算子 负荷预测 

分 类 号：TD453[矿业工程—矿山机电]