基于改进郊狼算法与极限学习机的工业金刚石检测  

Industrial diamond detection method based on improved coyote optimization algorithm and extreme learning machine

作　　者：杨建新 兰小平 赵振 杨一铭 王波 YANG Jianxin;LAN Xiaoping;ZHAO Zhen;YANG Yiming;WANG Bo(Information Center of China North Industries Group Corporation,Beijing 100089,China)

机构地区：[1]中国兵器工业信息中心,北京100089

出　　处：《计算机集成制造系统》2023年第2期449-459,共11页Computer Integrated Manufacturing Systems

摘　　要：为了提高工业金刚石的检测效率、保障产品质量,提出一种基于改进郊狼算法与极限学习机的工业金刚石检测方法。将工业金刚石视频图像按照一定时间序列分解为一组较为平稳的、形态单一的二维图像数据;利用深度卷积网络Inception-V3对多视角二维图像数据建立预测模型;在此基础上,以预测结果为输入构建极限学习机模型,并利用反向学习和莱维飞行改进的郊狼算法优化极限学习机输入权值和阈值,提高工业金刚石模型的检测精度。最后将该模型的检测结果与基本极限学习机、差分进化算法、粒子群优化算法和基本郊狼算法优化的极限学习机模型检测结果比较表明,该模型具有良好的检测精度和泛化能力,对于工业金刚石的质量检测具有指导意义。To improve the detection efficiency of industrial diamond and ensure product quality,an industrial diamond detection method based on an improved Coyote Optimization Algorithm(COA)and Extreme Learning Machine(ELM)was proposed.The video images of industrial diamond was decomposed into a group of relatively stable and single-dimensional image data according to a certain time series;the deep convolution network Inception-V3 was used to establish a prediction model for multi-perspective 2D image data.On this basis,the prediction results were used as input to construct the ELM model,and the COA improved by reverse learning and Levy flight was used to optimize the input weights and thresholds of ELM to improve the detection accuracy of the industrial diamond model.The detection results of the model were compared with basic ELM,and those of ELM model optimized by Differential Evolution algorithm(DE),Particle Swarm Optimization algorithm(PSO)and basic COA.The comparative experimental results showed that the model had good detection accuracy and generalization ability,which had guiding significance for the qualitative detection of industrial diamond.

关 键 词：工业金刚石 极限学习机 郊狼优化算法 反向学习 莱维飞行 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP391[自动化与计算机技术—控制科学与工程]