NOANIQLIKLAR SHAROITIDA QURITISH JARAYONINI INTELLEKTUAL BOSHQARISHNI TAHLIL QILISH VA OPTIMALLASHTIRISH

This paper examines the analysis and optimization of intelligent control for the wheat drying process under conditions of uncertainty. A multi-criteria vector objective function is formulated, simultaneously accounting for quality, energy, and time indicators, on the basis of which an algorithm for generating Pareto-optimal solutions is proposed. The decision-making process is organized according to H. Simon's four-stage scheme (Intelligence–Design–Choice–Implementation), integrated with ANFIS and the Pareto-optimization algorithm. The developed ANFIS controller was implemented in the MATLAB/Simulink environment using the Fuzzy Logic Toolbox and trained on a dataset of 1200 observations (split 70/15/15%). Training results showed RMSE stabilizing at 0.14% (train) and 0.18% (validation) over 150 epochs, confirming the absence of overfitting or underfitting. A comparative analysis of PID, static Fuzzy (Mamdani), ANN, and the proposed ANFIS controllers demonstrated the ANFIS controller's superiority in terms of moisture content, control signal, grain temperature, and specific energy consumption (34 kWh/t).