ПОДХОД К ПЛАНИРОВАНИЮ РЕСУРСОВ ПРОИЗВОДСТВА МАШИНОСТРОИТЕЛЬНОГО ПРЕДПРИЯТИЯ С ИСПОЛЬЗОВАНИЕМ НЕЙРОННЫХ СЕТЕЙ

Abstract

Generally, this is the experience of production workers, economists, designers, etc. that used to get a preliminary assessment of feasibility of hardware manufacturing at an enterprise. For a preliminary assessment, it is enough to understand the complexity of the product and available analogues. At the same time, it is not always possible to calculate an accurate production time for the product due to the lack of a complete set of design and technological documentation. The article presents an approach to calculating the production time of hardware using neural networks based on existing data for previous periods and types of hardware using. This approach allows estimating the production time without using accurate data on the design and manufacturing technology. The article also describes the structure of neural networks and defines the training sample. Some experiments were conducted based on the sample data, which allowed determining the initial weight coefficients of the neural network. The software implementation is made in the form of an additional module for an interactive web resource and uses T-SQL.