REGRESSION MODEL DEVELOPMENT TO PREDICT THE COOKING TIME OF SWEET PEPPER PUREE, ENSURING FINISHED PRODUCT OPTIMAL ORGANOLEPTIC PROPERTIES ACHIEVEMENT

The aim of the study is to develop a regression model for predicting the boiling time of sweet pepper puree used in the subsequent production of marshmallows, which ensures the achievement of optimal organoleptic properties of the finished product, based on the content of dry matter, sugars, and titratable acidity in the original raw materials. Objectives: to analyze the variability of the qualitative composition of raw materials; to establish the optimal duration of heat treatment for each experimental sample; to construct and verify a mathematical model of the dependence; to propose practical recommendations for the technological process. Research was conducted in the Department of Storage and Complex Processing of Agricultural Raw Materials of the Krasnodar Research Institute for Storage and Processing of Agricultural Products - a branch of the Federal State Budgetary Scientific Institution North Caucasus Federal Scientific Center of Horticulture, Viticulture, and Winemaking. The object of the study is puree prepared from sweet pepper of the Krepysh variety; samples of boiled sweet pepper puree, selected at various stages of heat treatment. A regression model has been developed to predict the cooking time of sweet pepper puree, ensuring the achievement of optimal organoleptic properties of the finished product for further use in marshmallow production. An analysis of the variability of the quality composition of sweet peppers was conducted. The optimal heat treatment duration for each experimental sample was determined, along with the boiling time (topt) corresponding to the maximum integrated quality assessment (the sum of scores for color, taste, odor, and consistency). For all experimental samples, the optimal boiling time was determined to be between 25 and 80 minutes. A mathematical model of the relationship was constructed and verified. Practical recommendations for the technological process were developed, the limits of the model's applicability were determined, and the potential economic impact of its implementation (reduced energy costs, standardized product quality) was determined.