Search

Improving protein stability is important for industrial applications, and one promising method for achieving this is backbone circularization. As connector length affects stability, predicting and elucidating a more stable connector length is necessary for development of the backbone circularization method. However, the relationship between connector length and protein stability has not been completely elucidated. Here, we determined the most stable connector length for granulocyte colony‐stimulating factor by changing one residue at a time to produce connector length variants and then measuring their thermal stability. Analysis of the local structures obtained from the predicted structures of the circularized variants revealed that an approach using helix length, dihedral backbone angle, and number of unbonded hydrogen bond donors and acceptors is suitable for identifying connector lengths with higher stability. [ABSTRACT FROM AUTHOR]

A cyber physical energy system (CPES) involves a combination of processing, network, and physical processes. The smart grid plays a vital role in the CPES model where information technology (IT) can be related to the physical system. At the same time, the machine learning (ML) models find useful for the smart grids integrated into the CPES for effective decision making. Also, the smart grids using ML and deep learning (DL) models are anticipated to lessen the requirement of placing many power plants for electricity utilization. In this aspect, this study designs optimal multi-head attention based bidirectional long short term memory (OMHA-MBLSTM) technique for smart grid stability prediction in CPES. The proposed OMHA-MBLSTM technique involves three subprocesses such as pre-processing, prediction, and hyperparameter optimization. The OMHA-MBLSTM technique employs min-max normalization as a pre-processing step. Besides, the MBLSTM model is applied for the prediction of stability level of the smart grids in CPES. At the same time, the moth swarm algorithm (MHA) is utilized for optimally modifying the hyperparameters involved in the MBLSTM model. To ensure the enhanced outcomes of the OMHA-MBLSTM technique, a series of simulations were carried out and the results are inspected under several aspects. The experimental results pointed out the better outcomes of the OMHA-MBLSTM technique over the recent models. [ABSTRACT FROM AUTHOR]

The stability of steel columns is difficult to predict accurately due to uncertain initial defects such as geometric imperfections and residual stress. To address this issue, we propose a probabilistic model that uses variational autoencoder (VAE) and transfer learning to estimate the loading capacities of steel columns. Our model can predict the confidence intervals of buckling loads without knowing the exact distribution of initial defects, providing more comprehensive information for engineering applications than traditional deterministic strength index. We establish a dataset of 1500 load-displacement curves of steel columns using four data augmentation approaches, and analyze the data distribution to validate the model's assumptions. Various criterions, including the mean squared error (MSE), the prediction interval coverage probability (PICP), and the prediction interval normalized average width (PINAW), are adopted to comprehensively measure the performance of confidence interval prediction. The numerical experiment validates that the trained model accurately predicts the confidence intervals for load-displacement responses, which perfectly cover the true curves with reasonable PINAW. Finally, we conduct a case study with a practical experiment to illustrate the model's potential application in failure probability calculation and reliability design. Our proposed model provides a promising probabilistic solution for quantifying the impact of uncertain parameters on structural analysis and significantly simplifies probability-based reliability design and optimal design processes. • A Bayesian deep learning-based probabilistic model for predicting steel column stability is proposed. • Uncertainties of initial defects are considered through a VAE-based two-stage algorithm. • A transfer learning method is proposed to convert VAE generative model to a predictive model. • The model accurately predict the probability intervals that cover the true stability curves. • The model can provide more detailed information than deterministic prediction of stability strength. [ABSTRACT FROM AUTHOR]

Abstract Stability prediction is an efficient way to avoid milling chatter which is one of the major limitations in increasing efficiency of milling operations. Since Insperger et al. [1] presented the semi-discretization methods (SDM) and demonstrated that the local discretization error is o (h 3), in order to increase prediction accuracy, full-discretization methods (FDMs) [2,3], Simpson method [4] and spectral methods [5,6] etc. are proposed. These methods use polynomial interpolation to approximate the original system. The prediction errors are always analyzed based on Taylor expansions. Mathematically, the error analysis based on Taylor's theorem requires that the original function has continuous derivatives of equivalent orders, that is, if a k -order polynomial interpolation is adopted, the original function must have continuous k -order derivatives [7]. However, milling operation is a highly interrupted cutting process, and the accelerations at the moments where cutter tooth enters and exits of the cut are obviously not continuous. Thus, even the order of local error at somewhere is rather high, the global error is not improved yet, and it may cause nonuniform convergence. In this paper, problems are revealed through simulations with specific milling parameters. It is concluded that the accuracy of the representative time-domain methods can hardly reach as high as o (h 3). Moreover, the discretization scheme to overcome the problems and improve applicability of the existing methods is also discussed. [ABSTRACT FROM AUTHOR]

Highlights • A hybrid method is proposed for predicting construction projects safety. • This method combines individual ML algorithms, data imputation, and the ensemble. • Ensemble classifiers are built on seven optimum ML models. • The proposed method is validated using an underground construction dataset. • The results show such method has great potential for improved safety prediction. Abstract Artificial intelligence (AI) approaches have proliferated in stability prediction of construction projects in the past decade. However, the application of AI approaches did not reach the peak of its potential due to the inappropriate handling of missing data and the omission of state-of-the-art techniques. In the present contribution, we proposed a hybrid method for the improved stability prediction of construction projects based on individual machine learning (ML) algorithms, input missing data imputation, semi-supervised learning and the classifier ensemble. Seven ML algorithms were selected to build individual classifiers for the classifier ensemble. 5-fold cross validation was used as the validation method and the performance measures were chosen to be the confusion matrix, the receiver operating characteristic (ROC) curve and the area under ROC curve (AUC). Exhaustive grid search and firefly algorithm were used for hyper-parameters and weights tuning respectively. The capability of the proposed method was verified using an underground construction dataset, the stope hangingwall (HW) dataset. The case study shows that the input missing data imputation and semi-supervised learning improved the predictive performance of ML algorithms on HW stability prediction. The highest and average AUC values on the testing set were increased to 0.954 and 0.923 respectively on the expanded dataset, compared with 0.879 and 0.860 on the original complete dataset. Further improvement was obtained through the classifier ensemble, with the AUC value being increased to 0.976. Harnessing such method extends recent efforts for stability prediction in construction projects, and can significantly accelerate the project design and stability management. [ABSTRACT FROM AUTHOR]

Process simulations are increasingly applied to analyze machining processes regarding process stability and the resulting surface quality of the workpiece. Due to their computational time, these simulations are inappropriate for real-time applications. Using Machine Learning approaches, monitoring systems for milling processes can be realized. Unfortunately, a huge amount of experimental data is necessary to train such models. A novel Machine Learning framework, which generates reliable predictions of the process stability, is presented in this paper. The model is designed based on results of a geometric physically-based simulation with varied process parameter values and refined using an active learning approach. [ABSTRACT FROM AUTHOR]

Productivity in robotic machining processes can be limited by instability phenomena resulting from the interaction between robot dynamics and cutting conditions. The robot dynamic behavior and consequently the machining stability vary within the workspace, due to the changes of the robot posture. Therefore, the chatter in machining robot depends on cutting parameters and as well as the robot configuration. Hence, each posture of the robot has its own stable cutting conditions. Moreover, along a machining trajectory, the robot can follow an infinite number of postures in its configuration space, due to the redundancies offered by the overall kinematic chain of the robotic cell. This paper deals with the control of the multi degrees functional redundancies of the ABB IRB6660 robotic cell for optimization of the machining stability. [ABSTRACT FROM AUTHOR]

Abstract: Prediction of Process Machine Interactions (PMI) can be achieved using models for conventional milling processes in the time and frequency domain. However, instabilities such as regenerative chatter can also be observed in micro milling processes using filigree cemented carbide end mills. The dominant chatter frequencies are basically related to the end mills eigenfrequencies. Nevertheless, the dynamic characteristics of the machine tool structure and the work piece must not be neglected. In this paper a comprehensive time domain process model is presented. The machine tool dynamics are considered as non-coupled oscillators based on measured frequency response functions at the tool holder. The end mill is modeled as rotating Euler-Bernoulli beam with variable boundary conditions at the tool clamping. At first, the parameter identification for a geometric cutting force model is described. It contains the cutting edge radius as a time-depended parameter. Thereafter, the modeling and parameter identification of the structural parts are presented. A stability criterion is defined with special regard to the tool deflection at the TCP. Finally, simulation results at different stable and unstable operating points for full immersion cutting are discussed in detail and compared to experimental tests. [Copyright &y& Elsevier]