NEURAL NETWORK MODELING FOR DETERMINING PATTERNS OF CHANGE IN THE PROPERTIES OF ADDITIVE POLYMER COMPOSITES AFTER EXPOSURE TO A MICROWAVE FIELD

Problems in determining rational hardening process modes are identified, related to the multifactorial nature of the processes occurring in materials consisting of components with different dielectric, thermal, and mechanical properties under microwave radiation exposure, which complicates the creation of an adequate mathematical description of the fundamental patterns. It is shown that the main parameter influencing the efficiency and quality of microwave hardening is the magnitude of the microwave power absorbed by the material. It is proposed to use artificial neural networks to generate models of the main processes that create conditions for improving the physical and mechanical properties of polymer composite materials (PCM). The absorbed microwave radiation power, along with the exposure time, was chosen as the object for modeling, as it has a major influence on the microwave heating temperature, which determines the conditions for the occurrence of structural changes in the composite. Based on neural network modeling, the dependence of microwave power absorption by cured PCMs with different binders and fillers on the dimensions of the part, the dielectric properties of the incoming materials, the reinforcement coefficient, the wavelength, and the radiation energy flux density was obtained. The initial data for constructing and training the model were a priori values ​​of the dielectric properties of the polymer composite components, specified values ​​for the dimensions of the impacted object, and the reinforcement ratio. It was shown that the model's performance metrics on the training and test samples were, respectively: R² = 0.998315 и R² = 0.997468. Using this model, a relationship was established between the content of components in PCM and their properties, the size and shape of the object of impact, and the radiation operating parameters (emitted power and wavelength) with the absorbed power, which makes it possible to predict the microwave heating temperature, which determines the strengthening effect of the treatment.