The hardness prediction model was established by support vector regression(SVR).In order to avoid exaggerating the contribution of very tiny alloying elements,a weighted fuzzy C-means(WFCM)algorithm was proposed for d...The hardness prediction model was established by support vector regression(SVR).In order to avoid exaggerating the contribution of very tiny alloying elements,a weighted fuzzy C-means(WFCM)algorithm was proposed for data clustering using improved Mahalanobis distance based on random forest importance values,which could play a full role of important features and avoid clustering center overlap.The samples were divided into two classes.The top 10 features of each class were selected to form two feature subsets for better performance of the model.The dimension and dispersion of features decreased in such feature subsets.Comparing four machine learning algorithms,SVR had the best performance and was chosen to modeling.The hyper-parameters of the SVR model were optimized by particle swarm optimization.The samples in validation set were classified according to minimum distance of sample to clustering centers,and then the SVR model trained by feature subset of corresponding class was used for prediction.Compared with the feature subset of original data set,the predicted values of model trained by feature subsets of classified samples by WFCM had higher correlation coefficient and lower root mean square error.It indicated that WFCM was an effective method to reduce the dispersion of features and improve the accuracy of model.展开更多
Probabilistic damage tolerance is a critical method to understand and communicate risk and safety.This paper reviews recent research on the probabilistic damage tolerance design for life-limited parts.The vision of th...Probabilistic damage tolerance is a critical method to understand and communicate risk and safety.This paper reviews recent research on the probabilistic damage tolerance design for life-limited parts.The vision of the probabilistic damage tolerance assessment is provided.Five core parts of the probabilistic damage tolerance method are introduced separately,including the anomaly distribution,stress processing and zone definition,fatigue and fracture calculation method,probability of failure(POF)calculation method,and the combination with residual stress induced by the manufacturing process.The above currently-available risk assessment methods provide practical tools for failure risk predictions and are applied by the airworthiness regulations.However,new problems are exposed with the development of the aeroengines.The time-consuming anomaly distribution derivation process restricts the development of the anomaly distribution,especially for the developing aviation industries with little empirical data.Additionally,the strong transient characteristic is prominent because of the significant temperature differences during the take-off and climbing periods.The complex loads then challenge the fatigue and fracture calculation model.Besides,high computational efficiency is required because various variables are considered to calculate the POF.Therefore,new technologies for the probabilistic damage tolerance assessment are provided,including the efficient anomaly distribution acquisition method based on small samples,the zone definition method considering transient process,and stress intensity factor(SIF)solutions under arbitrary stress distributions combined with the machine learning method.Then,an efficient numerical integration method for calculating failure risk based on the probability density evolution theory is proposed.Meanwhile,the influence of the manufacturing process on residual stress and the failure risk of the rotors is explored.The development of the probabilistic damage tolerance method can meet the requ展开更多
基金supported by the National Research and Development Project of China (2020YFB2008400).
文摘The hardness prediction model was established by support vector regression(SVR).In order to avoid exaggerating the contribution of very tiny alloying elements,a weighted fuzzy C-means(WFCM)algorithm was proposed for data clustering using improved Mahalanobis distance based on random forest importance values,which could play a full role of important features and avoid clustering center overlap.The samples were divided into two classes.The top 10 features of each class were selected to form two feature subsets for better performance of the model.The dimension and dispersion of features decreased in such feature subsets.Comparing four machine learning algorithms,SVR had the best performance and was chosen to modeling.The hyper-parameters of the SVR model were optimized by particle swarm optimization.The samples in validation set were classified according to minimum distance of sample to clustering centers,and then the SVR model trained by feature subset of corresponding class was used for prediction.Compared with the feature subset of original data set,the predicted values of model trained by feature subsets of classified samples by WFCM had higher correlation coefficient and lower root mean square error.It indicated that WFCM was an effective method to reduce the dispersion of features and improve the accuracy of model.
基金the National Natural Science Foundation of China,grant number U2233213.
文摘Probabilistic damage tolerance is a critical method to understand and communicate risk and safety.This paper reviews recent research on the probabilistic damage tolerance design for life-limited parts.The vision of the probabilistic damage tolerance assessment is provided.Five core parts of the probabilistic damage tolerance method are introduced separately,including the anomaly distribution,stress processing and zone definition,fatigue and fracture calculation method,probability of failure(POF)calculation method,and the combination with residual stress induced by the manufacturing process.The above currently-available risk assessment methods provide practical tools for failure risk predictions and are applied by the airworthiness regulations.However,new problems are exposed with the development of the aeroengines.The time-consuming anomaly distribution derivation process restricts the development of the anomaly distribution,especially for the developing aviation industries with little empirical data.Additionally,the strong transient characteristic is prominent because of the significant temperature differences during the take-off and climbing periods.The complex loads then challenge the fatigue and fracture calculation model.Besides,high computational efficiency is required because various variables are considered to calculate the POF.Therefore,new technologies for the probabilistic damage tolerance assessment are provided,including the efficient anomaly distribution acquisition method based on small samples,the zone definition method considering transient process,and stress intensity factor(SIF)solutions under arbitrary stress distributions combined with the machine learning method.Then,an efficient numerical integration method for calculating failure risk based on the probability density evolution theory is proposed.Meanwhile,the influence of the manufacturing process on residual stress and the failure risk of the rotors is explored.The development of the probabilistic damage tolerance method can meet the requ
