A5047024716- Nuclear reactor physics and engineering
- Neural Networks and Applications
- Model Reduction and Neural Networks
- Particle accelerators and beam dynamics
- Nuclear Physics and Applications
- Fault Detection and Control Systems
- Nuclear and radioactivity studies
- Nuclear Materials and Properties
- Graphite, nuclear technology, radiation studies
- Advanced Control Systems Optimization
- Neural Networks and Reservoir Computing
- Industrial Vision Systems and Defect Detection
- Magnetic confinement fusion research
- Radiation Effects in Electronics
- Magnetic Bearings and Levitation Dynamics
- Advanced Data Processing Techniques
- Particle Detector Development and Performance
- Control Systems and Identification
- Digital Media Forensic Detection
- Muon and positron interactions and applications
- Modeling and Simulation Systems
- Color Science and Applications
- Iron and Steelmaking Processes
- Power Systems and Renewable Energy
- Advanced battery technologies research
St Petersburg University
2014-2024
Here, we propose a novel approach to the classification of blue ballpoint pen inks based on combination selective extraction coloring components from paper carrier, digital color analysis (DCA) remaining traces, and hierarchical cluster DCA results. Since most documents high importance are still produced in hard copies, proposed method, being highly time- cost-efficient, could be significant contribution forensic science field authenticating handwritten documents. Several commonly used...
Long-lived actinides transmutation in accelerator driven system (ADS) is considered this paper. The objective to improve the performance from overall radiotoxity of nuclear waste point view by means an optimization approach based on control theory. Parameters subjected determination are initial values concentration loaded fuel. change time actinides’ atomic described a ordinary differential equations with conditions, that can be given some inaccuracy. For such dynamical gradient-based...
he rpid development of informtion tehnology leds to the ft tht numer users in teleommunition networks is onstntly growing long with need improve qulity serE vieF xetwork tr0 predition n importnt tsk this reD s it underlies network dignostis nd e0ient use its resouresF rious liner nonE methods suh neurl re tively used nlyze temporl sptil reltionships forest tr0F roweverD most them nnot urtely desrie dynmis networkD importne di'erent nodes hnges over timeD whih omplites topologyF his rtile...
Optimization approach to effective radioactive waste management in advanced fuel cycle is proposed the paper. This based on optimal control theory. The considered controlled system contains a of ordinary differential equations, describing isotopes concentration change time and number switching points which state can be changed. values variables are optimizing parameters, subjected determination. Moreover, because recycling nuclear supposed fulfill subcritical reactor, constraints...
The paper investigates the newly designed zinc-airflow battery cells and aims to build a mathematical model for such energy storage systems. work builds on electro-chemical characterization performed in previous study move focus towards modelling framework suitable management/control/supervision. In this endeavour, internal process is abstracted order obtain behavioural which concentrates parameters representative input-state-output model. Ideally, result will replicate behaviour allow...
The connection of Taylor maps and polynomial neural networks (PNN) to solve ordinary differential equations (ODEs) numerically is considered. Having the system ODEs, it possible calculate weights PNN that simulates dynamics these equations. It shown proposed architecture can provide better accuracy with less computational time in comparison traditional numerical solvers. Moreover, network derived from ODEs be used for simulation different initial conditions, but without training procedure....
Dynamics with feedbacks of Accelerator Driven Subcritical Reactor (ADSR) is considered. Different ways ADS control by accelerator are investigated. The response ADSR to charged particle beam characteristic changes obtained.
The production and integration of renewable energy sources into microgrid systems have recently demonstrated significant growth due to their ability meet growing electricity needs while having minimal impact on environmental pollution. Combined cooling, heating, power (CCHP) systems, also known as trigeneration are the most efficient stable way use energy, which has a wide range applications. However, increase efficiency reduce overall operating costs such development mathematical model...
The present paper deals with an optimization approach for long-lived transuranic isotopes transmutation carrying out in Accelerator Driven System (ADS). proposed methods consider the distributions of actinides nuclear concentration charged fuel as arguments. time-evolution fits a system ordinary differential equations (ODE), that is considered dynamical constraints. A gradient-based algorithm constructed this so overall reactivity minimized. Moreover, constraints on robustness and value...
The control problems concerning the manipulation of system trajectories ensembles have received increased attention in recent years. This paper considers a class nonlinear discrete-time systems with additive and develops systematic method to design optimal controls that steer an ensemble from initial state terminal one minimizing cost functional estimates dynamics average. Necessary optimality condition as well variation are constructed. These allow building different iterative or...
The stable manifold method is applied to construct a nonlinear real-time feedback optimal control system for the roll motion and vertical position of certain maglev platform. chosen platform uses combined electromagnetic suspensions consisting permanent magnets upper lower electromagnets. Within given technical gaps between guideway, magnetic forces provide highly effects. This makes this object multi-input multi-output (MIMO) system. an stabilizing controller. benefit in comparison with...
Predicting the behavior of a certain process in time is an important task that arises many applied areas, and information about system generated this can either be completely absent or partially limited. The only available knowledge accumulated data on past states parameters. Such successfully solved using machine learning methods, but when it comes to modeling physical experiments areas where ability model generalize interpretability predictions are important, then most methods do not fully...