Improving the Quality of Production Management Processes Based on Neural Network and Neuro-Fuzzy Models and Tools

The article describes a way to improve the quality of product control processes in food production by means of neural network and neuro-fuzzy methods, models, and tools. It is proposed to use feature extraction using convolutional networks with further postprocessing in a fuzzy inference system. During the operation of the proposed system, a high percentage of correct recognitions was obtained (91.9%) and customer returns of products due to defects decreased by 63% compared to the same period last year. The results obtained show that defect identification using an adaptive neuro-fuzzy inference system is a suitable tool for solving defect analysis problems.