прогнозуючої моделі технологічного процесу в реальному часі. Рішення завдання полягає в розробці інформаційної технології синтезу оптимального регулятора за функціоналом узагальненої роботи на основі прогнозуючої моделі процесу, яка формується в реальному часі шляхом структурно-параметричної ідентифікації процесу з використанням інтелектуальних базисних функцій (нейронних мереж і гібридних мереж з нечіткою логікою) по структурному (мінімуму зсуву) і параметричного (мінімуму помилки) критеріям з використанням глобальних (прямий випадковий пошук і генетичний алгоритм) і локальних (градієнтних) методів оптимізації.Запропонована інформаційна технологія дозволяє в 1,85 раз знизити похибки управління і підвищити на 3,8% продуктивність подальшого процесу самоздрібнювання по вхідний руді. Ключові слова: крупнокускове дроблення, інформаційна технологія, оптимальний регулятор, функціонал узагальненої роботи, ідентифікація, прогнозуюча модель, нейронна мережа, нечітка логіка, генетичний алгоритм.^UThis thesis is submitted to obtain a Ph. D (Candidate of Technical Sciences) in specialty 05.13.06 – information Technology – Dnipro University of Technology, Dnipro, 2021.This dissertation addresses the current scientific problem of improving the operating quality and management processes for large-scale ore crushing. This is achieved during uncertain conditions and information about its state. The use and development of information technology seeks to optimize the process with management based on the identification of predictive process models in real time. The solution lies in the development of information technology for the optimal controller synthesis for the functionality of generalized work. This is based on a predictive process model. This model is formed in real time by structural and parametric identification of the process using intelligent basic functions (neural networks and hybrid networks with fuzzy logic). The foregoing applies to structural and parametric (minimum error) criteria using global (direct random search and genetic algorithm) and local (gradient) optimization methods.The method of the optimal controller synthesis for the process of large-scale ore crushing has been improved. This method consists of finding the minimum functionality of the generalized work by taking into account the required depth of the forecast. This is done by summing up the components of the function and the cost function for quality control. This ensures the correct synthesis of the optimal controller for delayed processes.The complex method of structural and parametric identification of a nonlinear dynamic process was further developed. This method includes determining the required sampling interval of the process while taking into account its stochastic and dynamic properties. This reduced the error of the process model to the value of 0.0357. The proposed information technology allows control errors to be reduced by1.85 times.A modified method of structural and parametric identification has been developed. This includes the choice of the discrediting interval during the process when determining its characteristics. It also encompasses the required amount of data which allows the dynamic and stochastic properties of the process to be taken into account.The functional structure of the automated system by the process of large-scale ore crushing has been developed as well as the software application of the proposed techniques and algorithms of identification. The optimum controller that allows expenses for research and design of control systems to be reduced has been developed as well. For SCADA Siemens, an interface of a large crusher-operator- technologist has been developed, which allows modelling by way of a real technological process.Keywords: large-scale crushing, information technology, optimal regulator, generalized work functionality, neural network, fuzzy logic, genetic algorithm.