41 important scientific results were obtained in 2023 for the research works performed at the Institute, 11 of them were approved by the Science Council.

1. Application of artificial intelligence methods in the investigation of recognition, diagnosis, prognosis and decision-making problems.
Summary: A model of low-resource languages based on deep neural networks with the use of artificial intelligence technologies has been developed, an Azerbaijani-language virtual assistant has been created and introduced.
Implemented by: Full Member of ANAS Ali Abbasov, Associate Professor Abulfat Fatullayev, PhD in Engineering, Associate Professor Samir Rustamov, PhD in Engineering

2. Development and implementation of theoretical foundations of intelligent monitoring and control systems, which allow to provide identification of the beginning of change in the technical condition of objects and the dynamics of their development.
Summary: A high-order probabilistic model of the useful component of the noisy signal as a diagnostic attribute of inadmissible changes in the technological process has been developed, possible values of parameters and corresponding degrees of danger have been determined. The possibilities of using the developed algorithms for the objects of the transportation complex have been considered.
Implemented by: Full Member of ANAS, Prof. Telman Aliev, DSc in Engineering, Prof. Naila Musaeva, DSc in Engineering

3. Development and implementation of theoretical foundations of intelligent monitoring and control systems, which allow to ensure the identification of the beginning of changes in the technical condition of objects and the dynamics of their development.
Summary: Technologies for analyzing noise and useful signals of strategically important industrial objects, wattmeter charts, dynamometer charts and other technological parameters that can give a great economic effect have been proposed, SCADA control and management systems for oil wells operated by sucker rod pumps, submersible pumps, drilling rigs, reservoir irrigation stations have been created and implemented.
Implemented by: Assoc. Prof. Gambar Guluyev, DSc in Engineering, Assoc.Prof. Asif Rzayev, DSc in Engineering, Assoc.Prof. Farhad Pashayev, DSc in Engineering

4. Development of mathematical models, methods and algorithms for controlling special-purpose objects and systems in conditions of information deficit. Object recognition in aerospace images.
Summary: A recognition algorithm, an appropriate neural network and a learning mechanism based on the properties of aerospace photographs and/or objects in them, which remain unchanged regardless of their orientation, have been developed and implemented. Images in photographs are treated as planar figures and their static features are computed. These features include the "center of gravity" and "principal axes of inertia" of geometric figures. The features considered do not depend on the apparent orientation of the object from space, but are determined only in terms of its own shape. When the capture angle of the image changes little, these features also change little. By calculating these features of objects in aerospace images, it becomes possible to determine their orientation. Thus, it becomes possible to bring them to a common standard and identify them.
Implemented by: Full Member of ANAS, Prof. Ali Abbasov, DSc in Engineering, Assoc.Prof. Adalat Pashayev, PhD in Engineering, Assoc.Prof. Elkhan Sabziev, DSc in Engineering, junior researcher Ulviyye Goyushova

5. Development of identification methods, mathematical models and technologies for solving control problems in emergency situations.
Summary: Based on the principle of variable plan, i.e., prediction of optimal control through periodic processing of current data, new models, methods and optimization tools have been developed to improve flexible and organizational control. The main attention is paid to the method of dynamic programming, its modifications, and dimension reduction methods. The complexity principle is used to reduce computational difficulties, to compare methods with each other and to maximize the use of all computational resources.
Implemented by: Corresponding Member of ANAS, Prof. Aminaga Sadigov, DSc in Engineering

6. Development of methods, algorithms and software for mathematical modeling, parametric identification and optimal feedback control of objects with lumped and distributed parameters, computer experiments.
Summary: The problem of determining the parameters of a system consisting of a large number of dynamic subobjects has been solved. These subobjects are related to each other in an arbitrary order only by their states at the initial and/or final points, i.e., their boundary conditions are nonlocal. The state of each object is described by a system of linear non-autonomous ordinary differential equations independent of each other. Unknown parameters of external sources participate both in the differential equations and in the nonlocal conditions. Necessary conditions of optimality taking into account the unknown parameters have been obtained, and a scheme of numerical solution of the problem has been developed. Computer experiments on model problems have been conducted.
Implemented by: Corresponding Member of ANAS, Prof. Kamil Ayda-zade, DSc in Mathematics, Prof. Vagif Abdullayev, DSc in Mathematics, Assoc.Prof. Yegana Ashrafova, DSc in Mathematics

7. The optimization of the first mixed initial boundary value problem for hyperbolic differential inclusions (DFI) given by the Laplace operator has been considered. For this purpose, using the method of discretization of hyperbolic DFIs and the optimality conditions already obtained for discrete inclusions, the optimality conditions of the discrete-approximation problem in the form of Euler-Lagrange type inclusions have been formulated. In turn, by means of equivalence theorems, coupled Euler-Lagrange inclusions have been constructed and adequate optimality conditions for hyperbolic DFIs have been obtained. Furthermore, a method of extending the obtained results to a multivariate problem is shown. To demonstrate the above approach, their applications to some specific problems have been given.
Implemented by: Prof. Elmkhan Mahmudov, DSc in Mathematics

8. Development of methods for calculating and optimizing the performance of queuing-inventory systems (QIS).
Summary: Mathematical models of queuing-inventory systems whose stocks can be increased from different sources and with finite queues, taking into account the possibility of destructive requests to the warehouses have been constructed. The systems have been studied using different replenishment strategies, and exact and approximate numerical methods for calculating and optimizing their characteristics have been created.
Implemented by: Corresponding Member of ANAS, Prof. Agasi Melikov, DSc in Engineering, R. Mirzoyev (National Academy of Aviation), J. Sztrik (Hungary, Debrecen University)

9. Identification of problems arising in the implementation of Internet of Things (IoT) technologies for decision support in the process of smart city management and ways to solve them.
Summary: The possibilities of applying IoT decision support technologies in the process of smart city management have been investigated, and recommendations for solving problems and security issues have been developed.
Implemented by: Prof. Elchin Aliyev, DSc in Engineering, Abulfat Rahmanov

10. Compared with classical sets, intuitionistic fuzzy sets express information through membership, non-membership and oscillation functions. From this point of view, intuitionistic fuzzy methods are convenient in converting imprecise information into information and solving multi-criteria problems. Based on the proposed model, macroeconomic stability index has been calculated. In order to verify the adequacy of the model, the problem of model fit using macroeconomic indicators of Azerbaijan for 2016-2020 has been solved using the proposed algorithm.
Implemented by: Corresponding Member of ANAS, Prof. Gorkhmaz Imanov, DSc in Economics, Asif Aliyev, PhD in Economics

11. System analysis of the state of the art of methods and systems used in the field of oil production has been carried out, new intelligent methods of measurement, diagnostics, control and systems for oil production processes have been created. These systems make it possible to simplify the process of oil production equipment control as much as possible, to determine reservoir fluid leakage on the basis of the proposed algorithm, to automatically determine the density of oil-gas-water mixture, the degree of oil water cut and the percentage of cylinder filling without the use of a force transducer with the measurement of temperature at the wellhead of the tubing, pressure at the discharge line of the well and in 3 selected points of the tubing.
Implemented by: Prof. Abbas Rzayev, DSc in Engineering, Assoc.Prof. Rena Asadova, PhD in Engineering, Assoc.Prof. Mazahir Isayev, DSc in Engineering