Integral Dynamic Models in Identification and Diagnostics Problems

• Дата: 2020
• Автори: Верлань, Дмитро, Іванюк, Віталій, Фомін, Олександр
• Формат: Стаття
• Мова: Ukrainian
• Опубліковано: Кам'янець-Подільський національний університет імені Івана Огієнка 2020
• Онлайн доступ: http://mcm-math.kpnu.edu.ua/article/view/224858
• Назва журналу: Mathematical and computer modelling. Series: Physical and mathematical sciences

The paper considers a model approach to solving the problem of diagnostics and control of dynamic objects based on the application of integral equations. The growing complexity of energy facilities, taking into account their dynamic properties, increasing requirements for accuracy and objectivity of decisions lead to the need to develop new effective algorithms for mathematical software for diagnostic information processing systems to meet these requirements and automate the process of controlling power plants. Currently, the direction based on the restoration of the model (operator) for diagnostic purposes is being developed in technical diagnostics. It is generally assumed that faults change only the parameters of the model of objects of control, which are evaluated by diagnostic methods of parametric identification, but there may be cases when you need to take into account changes in the structure of the object. Real objects, as a rule, have both nonlinear and dynamic properties. Therefore, as an informative description of the objects of control of unknown structure, it is expedient to use nonlinear nonparametric dynamical models based on integer-degree Volterra series, which characterize the properties of objects of control (its state) as a sequence of multidimensional weight functions — Volterra kernels. The use of models based on Volterra kernels allows you to fully and accurately take into account the nonlinear and inertial properties of objects of control, makes the procedure of model diagnostics more universal, increases the reliability of the diagnosis.