针对光伏发电功率预测精度不高的问题,提出一种结合纵横交叉算法与改进的高斯过程回归算法(crisscross optimization algorithm and weighted Gaussian process regression,CSO-WGPR)的预测模型。首先,通过加权模糊聚类对天气类型进行划...针对光伏发电功率预测精度不高的问题,提出一种结合纵横交叉算法与改进的高斯过程回归算法(crisscross optimization algorithm and weighted Gaussian process regression,CSO-WGPR)的预测模型。首先,通过加权模糊聚类对天气类型进行划分,选出与预测日相同类型的相似日样本;其次,采用单类支持向量机(One-Class supportvectormachine,One-ClassSVM)算法结合传统高斯过程回归算法,建立改进后的高斯过程回归模型(weighted Gaussianprocess regression,WGPR),减小异常值数据对预测结果的不良影响;然后,采用纵横交叉算法(crisscross optimization algorithm,CSO)优化WGPR的超参数,进一步提高模型的预测精度。以澳洲爱丽丝泉光伏系统为例进行建模预测,真实数据仿真和实验结果表明,所提预测模型在晴天、阴天、雨天类型下具有更高的预测精度,验证了该方法的有效性。展开更多
The persistently high incidence of breast cancer emphasizes the need for precise detection in its diagnosis.Computer-aided medical systems are designed to provide accurate information and reduce human errors,in which ...The persistently high incidence of breast cancer emphasizes the need for precise detection in its diagnosis.Computer-aided medical systems are designed to provide accurate information and reduce human errors,in which accurate and effective segmentation of medical images plays a pivotal role in improving clinical outcomes.Multilevel Threshold Image Segmentation(MTIS)is widely favored due to its stability and straightforward implementation.Especially when dealing with sophisticated anatomical structures,high-level thresholding is a crucial technique in identifying fine details.To enhance the accuracy of complex breast cancer image segmentation,this paper proposes an improved version of RIME optimizer EECRIME,denoted as the double Enhanced solution quality Crisscross RIME algorithm.The original RIME initially conducts an efficient optimization to target promising solutions.The double-enhanced solution quality(EESQ)mechanism is proposed for thorough exploitation without falling into local optimum.In contrast,the crisscross operations perform a further local exploration of the generated feasible solutions.The performance of EECRIME is verified with basic and advanced algorithms on IEEE CEC2017 benchmark functions.Furthermore,an EECRIME-based MTIS method in combination with Kapur’s entropy is applied to segment breast Infiltrating Ductal Carcinoma(IDC)histology images.The results demonstrate that the developed model significantly surpasses its competitors,establishing it as a practical approach for complex medical image processing.展开更多
基金supported in part by the Natural Science Foundation of Zhejiang Province(LZ22F020005)National Natural Science Foundation of China(62076185,62301367).
文摘The persistently high incidence of breast cancer emphasizes the need for precise detection in its diagnosis.Computer-aided medical systems are designed to provide accurate information and reduce human errors,in which accurate and effective segmentation of medical images plays a pivotal role in improving clinical outcomes.Multilevel Threshold Image Segmentation(MTIS)is widely favored due to its stability and straightforward implementation.Especially when dealing with sophisticated anatomical structures,high-level thresholding is a crucial technique in identifying fine details.To enhance the accuracy of complex breast cancer image segmentation,this paper proposes an improved version of RIME optimizer EECRIME,denoted as the double Enhanced solution quality Crisscross RIME algorithm.The original RIME initially conducts an efficient optimization to target promising solutions.The double-enhanced solution quality(EESQ)mechanism is proposed for thorough exploitation without falling into local optimum.In contrast,the crisscross operations perform a further local exploration of the generated feasible solutions.The performance of EECRIME is verified with basic and advanced algorithms on IEEE CEC2017 benchmark functions.Furthermore,an EECRIME-based MTIS method in combination with Kapur’s entropy is applied to segment breast Infiltrating Ductal Carcinoma(IDC)histology images.The results demonstrate that the developed model significantly surpasses its competitors,establishing it as a practical approach for complex medical image processing.
