Development and verification of digital prototypes of construction machines using deep learning technologies: shortening the design cycle

Abstract

Digital prototyping of construction machinery is an innovative approach to optimizing design processes in the construction industry. This study is devoted to the development of a comprehensive methodology for creating and verifying digital prototypes of construction equipment using deep learning algorithms. The research is based on the integration of machine learning methods, computer modeling and multi-criteria optimization to reduce time and material costs for the development of new models of construction equipment. The work includes a comparative analysis of 387 digital models developed using traditional methods and the proposed approach based on neural network algorithms. Empirical validation of the methodology was carried out using the example of designing hydraulic excavators, tower cranes and bulldozers in real production conditions. The results demonstrate a 43,7% reduction in the design cycle, a 62,3% reduction in physical prototyping costs, and a 27,9% increase in the structural reliability of the final products. The developed neural network models ensure the accuracy of predicting operational characteristics at the level of 94,2±1,8%. The proposed methodology provides construction machinery designers with tools to effectively optimize the parameters of the designed machines and reduce the time it takes for new models to enter the market, which helps to increase the competitiveness of construction machinery enterprises.