Intelligent control system of quality work processes of construction and road machines (CRM)

Abstract

The study of the intellectualization of control system of quality work processes of construction and road machines was carried out. The analysis of existing researches and publications has been made, in which the main problem is highlighted, namely that the intellectualization concept of control system of quality work processes of CRM at this time is not enough. As a result of the analysis of existing researches and publications, the purpose of research is set, namely: analytical researches, the result of which will allow to increase of functioning efficiency arrangements of CRM with working environment using neural network and adaptation algorithm in a limited time decision. The concept of monitoring work parameters using artificial intelligence which is based on the neural network and is able to predict the work of CRM actuators in real time have been substantiated. The result of the research is selection of network learning algorithm, and also a scheme of analyzer work processes has been developed. This algorithm provides an iterative procedure for determining the minimum of a multidimensional function. The practical value lies in the fact that the method of back error propagation allows to calculate gradient components of loss functions regarding model parameters. The originality is in the fact that the results obtained prove using dual layer neural network for continuous monitoring quality of CRM workflow.