No abstract

No abstract

Structure of convolutional neural network (CNN) applied for image recognition requires large numbers of tuning for designated datasets in practice. It is a time-consuming process to finally come up with a feasible structure for specific requirement. This paper proposes a method based on Taguchi method which can efficiently determine the optimal structure of hyperparameters combination. Five hyperparameters with four levels are defined as control factors and two indicators are chosen to measure the performance of CNN structure. L16 (45) orthogonal array is used to arrange the experiment. S/N ratio and main effect plot are used to identify the optimal structure (hyperparameter combination) of CNN. The classic case of MNIST is employed to verify the practicability of the proposed method. Results show that the proposed method can identify the optimal CNN structure efficiently and also rank the significance priority of hyperparameters.

Building information modeling (BIM) not only can be a medium for 3D display and simulation but also can be embedded in the knowledge ontology to connect more comprehensive information. Despite the advantages of the BIM‐embedded knowledge ontology that has been explored in engineering practices, it is still less discussed in engineering education. Therefore, this paper aims to establish a BIM‐embedded knowledge‐sharing platform and its corresponding learning community model to improve learning outcomes. The BIM‐embedded knowledge‐sharing platform is established by the MediaWiki to connect the knowledge entries with the related BIM components to expand the BIM‐embedded knowledge ontology. The learning community model is developed on problem‐based learning and project‐based learning to encourage students to build their own knowledge by creating, sharing, and modifying knowledge entries. In the experimental teaching, twenty students are invited to participate in the learning community for a class on prefabricated building design. Students are guided to build and share their knowledge entries and the related BIM components by following the learning community model. Besides this, they also use the BIM‐embedded knowledge‐sharing platform to complete a real design project. Results show the proposed BIM‐embedded knowledge‐sharing platform and the learning community model are beneficial for the learning outcomes.

In petrochemical industry, the execution of construction involves three main issues, namely, design planning, construction, and job safety. Three-dimensional (3D) models are increasingly applied to design and construction. However, the improper concept of 3D design has bred potential unsatisfactory behaviors and the lack of vigilance among workers. Besides, many employees are not fully aware of the safety in 3D design and construction planning. Therefore, our goal is to improve the safety and health of construction workers through design practices in the upstream of the construction phase, and verify the applicability of the combination of 3D models and safety knowledge. Specifically, a questionnaire survey was carried out among 124 employees in the construction-related fields of the petrochemical industry. The collected data were processed, and statically analyzed on SPSS. The results show that safety knowledge was acceptable in 3D model design from the perspective of project executors, and the integration of safety knowledge into the design helps to improve the safety environment of the construction site.