Search results

In recent years, the convolutional neural network (CNN) based deep learning approach has succeeded in data-mining the relationship between microstructures and macroscopic properties of materials. However, such CNN models usually rely heavily on a large set of labeled images to ensure the accuracy and generalization ability of the predictive models. Unfortunately, in many fields, acquiring image data is expensive and inconvenient. This study aims to propose a data augmentation technique to enhance the performance of the CNN models for linking microstructural images to the macroscopic properties of composites.

Microstructures of composites are synthesized using discrete element simulations and Potts kinetic Monte Carlo simulations. Macroscopic properties such as the elastic modulus, Poisson's ratio, shear modulus, coefficient of thermal expansion, and triple-phase boundary length density are extracted on representative volume elements. The CNN model is trained using the 3D microstructural images as inputs and corresponding macroscopic properties as the labels. The comparison of the predictive performance of the CNN models with and without data augmentation treatment are compared.

The comparison between the prediction performance of CNN models with and without data augmentation showed that the former reduced the weighted mean absolute percentage error (WMAPE) for the prediction from 5.1627% to 1.7014%. This significant reduction signifies that the proposed data augmentation method can effectively enhance the generalization ability and robustness of CNN models.

This study demonstrates that data augmentation is beneficial for solving the problems of model overfitting, data scarcity, and sample imbalance for CNN-based deep learning tasks at a low cost. By developing more and advanced data augmentation techniques, deep learning accelerated homogenization will boost the multi-scale computational mechanics and materials.

This paper aims to study the impact of individual relationships on tacit knowledge sharing in the company setting of compulsory bond, expressive bond, instrumental bond and self-monitoring by empirical explorations.

The paper raises seven hypotheses that focus on the impact of employees’ relationship with tacit knowledge sharing in knowledge-intensive industries and positions based on relationship theory. Before distributing the formal questionnaires, a pre-research was done in a college by collecting comments and suggestions so as to correct and modify the questionnaires. A four-page questionnaire based on the Likert scale with 45 questions was used for data collection, and 210 valid questionnaires were collected from a research institute, a software company and an educational institute. Finally, SPSS17.0 was used to analyze these data, including reliability analysis, validity analysis, correlation analysis and regression analysis, etc.

The findings include: there is a positive correlation between employees’ compulsory bond and the efficiency of tacit knowledge sharing; there is a positive correlation between employees’ expressive bond and the efficiency of tacit knowledge sharing; there is a negative correlation between employees’ instrumental bond and the efficiency of tacit knowledge sharing; the more apt employees are at self-monitoring, the more effectively they will share tacit knowledge; the interaction between compulsory bonds and self-monitoring has a positive and stimulating impact on tacit knowledge sharing; the interaction between expressive bonds and self-monitoring has a positive and stimulating impact on tacit knowledge sharing; and the interaction between instrumental bonds and self-monitoring has a certain impact on tacit knowledge sharing.

However, the efficiency of tacit knowledge sharing cannot be measured easily and how to share the tacit knowledge based on employees’ relationships should be further concerned by knowledge industries.

This paper illustrates multiple, in-depth approaches to research on knowledge sharing. It shows why it is important to pay attention to employees’ relationships during the process of tacit knowledge sharing. The author argued some key factors such as compulsory bond, emotional bond and self-monitoring that may have a certain impact on the tacit knowledge sharing. The paper also further discussed the influence on the sharing of tacit knowledge as for the interaction between different relationship types and self-monitoring.

The knowledge is critical to enhance enterprises’ performance, and it will become more useful when the new knowledge is shared with others. However, tacit knowledge cannot be measured easily, and how to share the tacit knowledge based on employees’ relationships should be further concerned by knowledge industries. A series of findings are proposed in this paper.

Integrating the knowledge of different individuals, of which 90 per cent is tacit knowledge, in an organization that engages in producing products and providing service is instrumental to the sustainability and productivity of that organization. This study addressed the factors and dynamics of tacit knowledge sharing that can be used in knowledge management to effectively capture, store and disseminate tacit knowledge across an organization.

In this work, the authors aim to provide a set of convenient methods for generating training data, and then develop a deep learning method based on point clouds to estimate the pose of target for robot grasping.

This work presents a deep learning method PointSimGrasp on point clouds for robot grasping. In PointSimGrasp, a point cloud emulator is introduced to generate training data and a pose estimation algorithm, which, based on deep learning, is designed. After trained with the emulation data set, the pose estimation algorithm could estimate the pose of target.

In experiment part, an experimental platform is built, which contains a six-axis industrial robot, a binocular structured-light sensor and a base platform with adjustable inclination. A data set that contains three subsets is set up on the experimental platform. After trained with the emulation data set, the PointSimGrasp is tested on the experimental data set, and an average translation error of about 2–3 mm and an average rotation error of about 2–5 degrees are obtained.

The contributions are as follows: first, a deep learning method on point clouds is proposed to estimate 6D pose of target; second, a convenient training method for pose estimation algorithm is presented and a point cloud emulator is introduced to generate training data; finally, an experimental platform is built, and the PointSimGrasp is tested on the platform.