Design and Evaluation of Virtual Compression Testing Machine Based on Multilayer Perceptron for Vocational-Education Virtual Laboratory

Abstract

The weaknesses of Conventional Compression Testing Machines (CCTM) for student practicum are the high costs of procurement, maintenance, electricity, concrete materials, and human/equipment error factors. This research proposed of Virtual Compression Testing Machine (VCTM) based on Multilayer Perceptron that uses 2 hidden layers, 60 neurons with bagging, and other parameters of a Deep Neural Network with RMSE value of 4.738. The application of VCTM has been successfully carried out and there was no significant difference with CCTM. VCTM can be used for various types of concrete with high testing intensity. Lecturers and researchers can use VCTM to conduct research.