Vladimír Gašpar

Intelligent Supervisory System for Small Turbojet Engines

The paper deals with design of an intelligent supervisory system usable in the area of engineering applications of small turbojet engines. Such supervisory system consists of diagnostic and control system applied in different phases of the engine’s operation. In order to design it we need to know the object and its models that will be realized by means of situational modeling concept and neural networks. Diagnostic system will be based on these types of models applied into the majority voting method of online diagnostic. Intelligent supervisory system will also include off-line pre-start diagnostics for such engine with perspective health monitor database.

On scientific research information system

The presented paper proposes a design of a scientific research information system (SRIS), intended for real world use according to existing requirements and processes in the Laboratory of Intelligent Control Systems of Aircraft Engines. The novelty of the proposed SRIS lies in universal and narrow scope of the system and digitalization of existing manual processes, with the addition of algorithms and methods for research efficiency evaluation. The key idea is the possibility to design own quantitative key performance indicators (KPIs) for the efficiency evaluation. It is possible to generalize the presented approach for any research environment, either for technical, financial, time related or information efficiency evaluation. In case of a complex technical systems, the SRIS can also be considered a decision support system (DSS), when using various data mining algorithms e.g. clustering, classification, association rules, etc.

Expansion of working envelope of a small turbojet engine MPM-20

Development of aircraft engines is aimed on increase of reliability, efficiency and ecology of operation. The presently still used hydro-mechanic control systems are being replaced by electronic ones characterized by high speed, precision of control and ability to optimize all operating regimes of an engine. One of the aims of research in the Laboratory of intelligent control systems of aircraft engines lies in new possibilities of expansion of working envelope of turbojet engines by implementation of variable exhaust geometry and new approaches in its control strategies and construction solutions. Such expansion can be beneficial for small turbojet engines and increase their possibilities of real-world applications. New control methodologies and approaches can be further expanded to normal sized engines. All these problems are described in the presented article.

Digital start-up control of the small turbojet engine MPM-20

The main thesis of this paper considers the possibilities and difficulties of switching from hydromechanical engine control to FADEC (digital) engine control. Regarding different specifications and elements in system, possible errors, advantages and disadvantages, each method has its specific properties of behavior in specific environments and applications. When considering our controlled system, the MPM-20 small turbojet engine, it is a challenge to propose functional control system with backups and with slight possibility of errors during or before operation. In this paper we propose methods of backups for specific elements, prestart diagnostics and on-line diagnostics.

On practical constraints of approximation using neural networks on current digital computers

Goal of this paper is to highlight the most common problems and constraints which accompany the implementation of artificial neural networks on current digital computers. We focus on feed-forward multilayer neural networks, i.e. multilayer perceptrons, in role of function approximators. Multiple constraints of approximation by neural networks are discussed within the paper, taking into account research from the previous two decades. We address the issues of structural construction of feed-forward neural networks, learning and data pretreatment. Conclusions stated by universal approximation theorem cannot be blindly applied to implementations on real hardware without considering the limitations such as finite accuracy of floating point operations and data type overflow issues. This fact is emphasized in the paper.

Scientific Research Information System as a Solution for Assessing the Efficiency of Applied Research

The aim of this paper is to discuss the possibilities of evaluating qualitative parameters (mostly physical and technical efficiency) in research of a certain complex system using an information system with intelligent elements. Our main objective is to create the possibility to test the quality of each research element or task. The research of a complex thermodynamic system, in our case a small turbojet engine MPM–20, requires assessing each element with respect to its nature. Nature of these elements may be physical (e.g. reaching limiting temperatures, fuel consumption), technical (e.g. quality of control algorithms, quality of diagnostics), financial (e.g. research funding quality), time related (e.g. latencies, dynamics) or informational (quality of research papers and research reports). A basis for quality assessment of each research element is provided by key performance indicators (KPIs), which are measureable and robust. Such an information system creates a new overview of the research process and progress, while serving as an integrating element of the whole research as well as an aid in the experimental identification of the particular complex system. Intelligent elements such as automatic generation of log reports and research schedules, parameter and measurement classification and clustering of atypical situations, which are discussed later in this article, add up to the simplification of research work and create an environment that is beneficial for improving the quality of research. In the future, we are planning to generalize the proposed concepts and modules to serve any kind of applied research.

Redundant backup and diagnostic system of MPM-20 engine

Progressive development of turbojet engine requires focus not only on improving performance, but also high reliability and safety are very important. Thats why we put more and more emphasis on concepts like diagnostics and backup, which represent the basic techniques and methods that contribute to increased safety of any complex system. The article deals with the design of diagnostics/backup system, which uses majority method and backup through computational models that take advantage of experimental identification methods and neural networks. The proposed system for jet engines was subsequently tested in laboratory conditions (LIRS LM-Laboratory of Intelligent Control Systems of Jet Engines) on a small turbojet engine MPM-20, where confirmed its functionality and reliability.

Thrust vectoring nozzle for small turbojet engine in laboratory conditions

The article deals with description of thermodynamic processes in a class of small turbojet engines with description of experimental and analytic modeling approaches to create a mathematical model. The model will describe a small turbojet engine as a system with one degree of freedom (fast exhaust nozzle), two degrees of freedom controlled by fuel supply and exhaust nozzle diameter. The article then deals with extension of such model with the third degree of freedom in the form of thrust vector. Such system developed for a small turbojet engine can be beneficial in applications with highly maneuverable small unmanned aerial vehicles.

Advanced approaches in diagnostics of the turbojet engine MPM-20

This paper deals with modeling and diagnostics of complex systems and with a prestart diagnostics expert system design and implementation. In the area of modeling, the authors describe dynamic models of the speed dependence upon the fuel flow supply of a small turbojet engine MPM-20. The first model is done with use of neural networks and the second is a linear model. Both models are then methodologically incorporated into the diagnostic and backup system of the optical sensor measuring the engine speed.

Design and implementation of a client warehouse application over an enterprise resource planning system for mobile devices

This paper proposes a client warehouse application over Microsoft Dynamics Navision 5.0 enterprise resource planning system. The application has been created to meet criteria and constraints stated by real world company, or in terms of application itself, the future user. By means of effectiveness, created application shortens or eliminates some warehouse activities. Employees can easily roam in warehouse and create physical inventory, transfer, intake or expend goods using PDAs with WiFi connection to database server. Furthermore, if WiFi signal is not strong enough to cover the whole working area of the warehouse, there is a possibility to synchronize essential data between source server database and local device compact database. Usage of barcode scanner makes every operation even more effective in terms of activity time consumption.