Using Deterministic Models and Neural Networks in Digital Twins of Metal Cutting Processes

The use of data processing algorithms based on neural networks in combination with deterministic mathematical models is considered. The pairing of these two approaches was first proposed in constructing a uniform virtual modeling system for cutting processes in metalworking machines, which use a neural network for adaptive parametric identification of the models describing the processes. Two types of research methods are employed: experiments and Matlab modeling. On the basis of the results, a systematic new approach to the construction of digital twin systems may be proposed.