THE RESULTS OF THE STUDY OF A DIGITAL SYSTEM AND A MULTIFUNCTIONAL MACHINE FOR AUTOMATIC SORTING OF EGGS INTO CATEGORIES

In industrial poultry farming, improving the accuracy, objectivity, and reproducibility of egg quality control is a pressing issue, as traditional grading methods, based primarily on product weight, do not take into account a range of morphological and physical characteristics, such as shape, volume, and density. This leads to reduced classification reliability, especially in the boundary zones of categories, and limits the capabilities of process control, negatively impacting the efficiency of subsequent production processes. This study utilized non-destructive testing methods, including weight measurement, machine vision systems, and digital image processing algorithms to determine egg geometric parameters. Based on the obtained data, volume and density indicators were additionally calculated, allowing for expanded quality assessment criteria. Fuzzy logic algorithms were used for classification, accounting for measurement uncertainty and biological sample variability. Experimental studies were conducted on a sample of 760 eggs of various categories. It was found that the developed system ensures stable operation in continuous mode with a throughput of up to 3,600 eggs per hour while maintaining the required measurement accuracy. The use of fuzzy logic increased sorting accuracy to 95.7 %, a 15-20 % increase over traditional methods. It was shown that taking geometric parameters into account allows for the detection of shape and density deviations not detected by standard weight sorting. These results confirm the effectiveness of the multi-criteria intelligent approach and can be used in the development of automated systems for process control and egg sorting, as well as adapted for other types of agricultural products in real-world production conditions.