The quality prediction of machining processes is essential for maintaining process stability and improving component quality. The prediction accuracy of conventional methods relies on a significant amount of process s...The quality prediction of machining processes is essential for maintaining process stability and improving component quality. The prediction accuracy of conventional methods relies on a significant amount of process signals under the same operating conditions. However, obtaining sufficient data during the machining process is difficult under most operating conditions, and conventional prediction methods require a certain amount of training data. Herein, a new multiconditional machining quality prediction model based on a deep transfer learning network is proposed. A process quality prediction model is built under multiple operating conditions. A deep convolutional neural network (CNN) is used to investigate the connections between multidimensional process signals and quality under source operating conditions. Three strategies, namely structure transfer, parameter transfer, and weight transfer, are used to transfer the trained CNN network to the target operating conditions. The machining quality prediction model predicts the machining quality of the target operating conditions using limited data. A multiconditional forging process is designed to validate the effectiveness of the proposed method. Compared with other data-driven methods, the proposed deep transfer learning network offers enhanced performance in terms of prediction accuracy under different conditions.展开更多
High product quality is one of key demands of customers in the field of manufacturing such as computer numerical control(CNC)machining.Quality monitoring and prediction is of great importance to assure high-quality or...High product quality is one of key demands of customers in the field of manufacturing such as computer numerical control(CNC)machining.Quality monitoring and prediction is of great importance to assure high-quality or zero defect production.In this work,we consider roughness parameter Ra,profile deviation Pt and roundness deviation RONt of the machined products by a lathe.Intrinsically,these three parameters are much related to the machine spindle parameters of preload,temperature,and rotations per minute(RPMs),while in this paper,spindle vibration and cutting force are taken as inputs and used to predict the three quality parameters.Power spectral density(PSD)based feature extraction,the method to generate compact and well-correlated features,is proposed in details in this paper.Using the efficient features,neural network based machine learning technique turns out to be able to result in high prediction accuracy with R2 score of 0.92 for roughness,0.86 for profile,and 0.95 for roundness.展开更多
Product variation reduction is critical to improve process efficiency and product quality, especially for multistage machining process（MMP）. However, due to the variation accumulation and propagation, it becomes qui...Product variation reduction is critical to improve process efficiency and product quality, especially for multistage machining process（MMP）. However, due to the variation accumulation and propagation, it becomes quite difficult to predict and reduce product variation for MMP. While the method of statistical process control can be used to control product quality, it is used mainly to monitor the process change rather than to analyze the cause of product variation. In this paper, based on a differential description of the contact kinematics of locators and part surfaces, and the geometric constraints equation defined by the locating scheme, an improved analytical variation propagation model for MMP is presented. In which the influence of both locator position and machining error on part quality is considered while, in traditional model, it usually focuses on datum error and fixture error. Coordinate transformation theory is used to reflect the generation and transmission laws of error in the establishment of the model. The concept of deviation matrix is heavily applied to establish an explicit mapping between the geometric deviation of part and the process error sources. In each machining stage, the part deviation is formulized as three separated components corresponding to three different kinds of error sources, which can be further applied to fault identification and design optimization for complicated machining process. An example part for MMP is given out to validate the effectiveness of the methodology. The experiment results show that the model prediction and the actual measurement match well. This paper provides a method to predict part deviation under the influence of fixture error, datum error and machining error, and it enriches the way of quality prediction for MMP.展开更多
Although many works have been done to construct prediction models on yarn processing quality,the relation between spinning variables and yarn properties has not been established conclusively so far.Support vector mach...Although many works have been done to construct prediction models on yarn processing quality,the relation between spinning variables and yarn properties has not been established conclusively so far.Support vector machines(SVMs),based on statistical learning theory,are gaining applications in the areas of machine learning and pattern recognition because of the high accuracy and good generalization capability.This study briefly introduces the SVM regression algorithms,and presents the SVM based system architecture for predicting yarn properties.Model selection which amounts to search in hyper-parameter space is performed for study of suitable parameters with grid-research method.Experimental results have been compared with those of artificial neural network(ANN)models.The investigation indicates that in the small data sets and real-life production,SVM models are capable of remaining the stability of predictive accuracy,and more suitable for noisy and dynamic spinning process.展开更多
基金supported by the National Natural Science Foundation of China (Grant No.51675418).
文摘The quality prediction of machining processes is essential for maintaining process stability and improving component quality. The prediction accuracy of conventional methods relies on a significant amount of process signals under the same operating conditions. However, obtaining sufficient data during the machining process is difficult under most operating conditions, and conventional prediction methods require a certain amount of training data. Herein, a new multiconditional machining quality prediction model based on a deep transfer learning network is proposed. A process quality prediction model is built under multiple operating conditions. A deep convolutional neural network (CNN) is used to investigate the connections between multidimensional process signals and quality under source operating conditions. Three strategies, namely structure transfer, parameter transfer, and weight transfer, are used to transfer the trained CNN network to the target operating conditions. The machining quality prediction model predicts the machining quality of the target operating conditions using limited data. A multiconditional forging process is designed to validate the effectiveness of the proposed method. Compared with other data-driven methods, the proposed deep transfer learning network offers enhanced performance in terms of prediction accuracy under different conditions.
文摘High product quality is one of key demands of customers in the field of manufacturing such as computer numerical control(CNC)machining.Quality monitoring and prediction is of great importance to assure high-quality or zero defect production.In this work,we consider roughness parameter Ra,profile deviation Pt and roundness deviation RONt of the machined products by a lathe.Intrinsically,these three parameters are much related to the machine spindle parameters of preload,temperature,and rotations per minute(RPMs),while in this paper,spindle vibration and cutting force are taken as inputs and used to predict the three quality parameters.Power spectral density(PSD)based feature extraction,the method to generate compact and well-correlated features,is proposed in details in this paper.Using the efficient features,neural network based machine learning technique turns out to be able to result in high prediction accuracy with R2 score of 0.92 for roughness,0.86 for profile,and 0.95 for roundness.
基金Supported by National Natural Science Foundation of China(Grant Nos.51205286,51275348)
文摘Product variation reduction is critical to improve process efficiency and product quality, especially for multistage machining process（MMP）. However, due to the variation accumulation and propagation, it becomes quite difficult to predict and reduce product variation for MMP. While the method of statistical process control can be used to control product quality, it is used mainly to monitor the process change rather than to analyze the cause of product variation. In this paper, based on a differential description of the contact kinematics of locators and part surfaces, and the geometric constraints equation defined by the locating scheme, an improved analytical variation propagation model for MMP is presented. In which the influence of both locator position and machining error on part quality is considered while, in traditional model, it usually focuses on datum error and fixture error. Coordinate transformation theory is used to reflect the generation and transmission laws of error in the establishment of the model. The concept of deviation matrix is heavily applied to establish an explicit mapping between the geometric deviation of part and the process error sources. In each machining stage, the part deviation is formulized as three separated components corresponding to three different kinds of error sources, which can be further applied to fault identification and design optimization for complicated machining process. An example part for MMP is given out to validate the effectiveness of the methodology. The experiment results show that the model prediction and the actual measurement match well. This paper provides a method to predict part deviation under the influence of fixture error, datum error and machining error, and it enriches the way of quality prediction for MMP.
基金National Science Foundation and Technology Innovation Fund of P.R.China(No.70371040and02LJ-14-05-01)
文摘Although many works have been done to construct prediction models on yarn processing quality,the relation between spinning variables and yarn properties has not been established conclusively so far.Support vector machines(SVMs),based on statistical learning theory,are gaining applications in the areas of machine learning and pattern recognition because of the high accuracy and good generalization capability.This study briefly introduces the SVM regression algorithms,and presents the SVM based system architecture for predicting yarn properties.Model selection which amounts to search in hyper-parameter space is performed for study of suitable parameters with grid-research method.Experimental results have been compared with those of artificial neural network(ANN)models.The investigation indicates that in the small data sets and real-life production,SVM models are capable of remaining the stability of predictive accuracy,and more suitable for noisy and dynamic spinning process.
