Essay on the topic of master's thesis
Development of expert system for diagnosis of turbogenerators in the thermoelectric power station
Contents
Goals and objectives of the system
The general formulation of the problem
Introduction
The object of study is the generator on a power plant and its system of care.
Technical Diagnostics - a relatively new branch of knowledge, rapidly growing in recent years for two main reasons. First, the introduction of new methods of technical diagnostics greatly increases the effectiveness of preventive maintenance. This is especially important in connection with the backlog of personnel engaged in repair and prevention of growth of the installed equipment. The situation is exacerbated by the fact that for large substations significant part of the main equipment used outside the rated service life. Improving the effectiveness of preventive maintenance in the first place should be provided a transition from planning audits and inspections on terms of service, dependent on the state of the equipment [7].
Secondly, the use of technical diagnostics, using modern techniques, improving reliability and availability. Great importance in this case is technically sound preventative maintenance. [Up]
Goals and objectives of the system
To implement the diagnostic function of the equipment necessary to receive and build a complete and accurate information during the entire period of operation of power equipment.
The main goal of the system:
- a picture of the current status and trends of development of scientific and practical developments on the topic of master's work;
- identify gaps, unresolved issues in existing development;
- improving decision-making process;
- fixing the parameters exceed the value specified with appropriate signals or commands;
- graphically display the current settings in time;
- rationale for recommendations to address the deviations;
- determining the causes of situations of giving up.
Relevancy
Turbogenerator power plant is a major component of thermoelectric power station. It consists of a complex integrated system in which work can be a lot of unforeseen situations. To ensure continuous operational control turbine generators, as well as their support systems is the basis for reliable operation of process equipment. [Up]
Scientific novelty
At the moment did not solve the problem of determining the causes of situations with delivery errors. There is a universal tool, which is difficult to implement in the form of applied expert systems, including due to differences in software, hardware in different systems. [Up]
Similar studies
At these times, there are many diagnostic systems turbogenerator. Systems used in Ukraine, do not solve all the required tasks of diagnosis.
- An example is the system of Neptune, designed for automated monitoring and diagnosis of turbine generators and their auxiliary systems on technology, power and vibration parameters during operation. The system is used for the diagnosis of turbine generators with a hydrogen-water cooled and completely at the factory, as well as reconstruction of existing thermal power. Possible to use the system for controlling the air-cooled turbo-generators [8].
- Also used in power system monitoring and diagnosis of generators STK-ER. The identification of deviations from normal operation and replacement reports automatically or on request from the operator are printed on a printer. Input tuning parameters: the choice of channel, type converter, plug setting, measuring range, value systems, etc. performed with a panel computer and is password protected. The system allows operational control, indication and recording of the technological parameters of the generator in a database system. Also allows for periodic or on request from the operator-technologist registration generator parameters on paper or magnetic media. Carried out continuous monitoring of the generator with the provision of information in the form of mimics, parameter tables, graphs, depending on the parameters of the time, warning and emergency messages [9].
- To improve the diagnosis and the decision by the operator at the power used by a decision support system personnel. Decision Support System - a computer automated system, which aims to help people make decisions in difficult circumstances for a full and objective analysis of the substantive work. Decision support systems have a number of service functions for displaying information in various forms, including a sample of user queries. The main drawback of the above systems is the lack of mining the information received. All presented above expert systems can detect an error, but does not explain the reason for its appearance. Moreover, the number of possible situations known to the system is limited, and when an emergency situations, information can not be correctly processed, which significantly limits the ability of systems of technical diagnostics. Thus, in order to solve this problem in this paper, we propose the establishment of the diagnostic system that includes data mining. [Up]
- introduction of new methods of technical diagnostics greatly increases the effectiveness of preventive maintenance. This is especially important in connection with the backlog of personnel engaged in repair and prevention of growth of installed equipment. The situation is exacerbated by the fact that for large substations significant part of the main equipment used outside the rated service life. Improving the effectiveness of preventive maintenance in the first place should be provided a transition from planning audits and inspections on terms of service, dependent on the state of equipment;
- application of technical diagnostics, using modern techniques, improving reliability and availability. Great importance in this case is technically sound preventative maintenance.
- Справочник по ремонту турбогенераторов // Под ред. Бекова Х.А., Барило В.В.- М.: ИПКгосслужбы, ВИПКэнерго, 2006. - 724 с.
- Алексеев Б.А. Определение состояний (диагностика) крупных турбогенераторов // М.:НЦ ЭНАС, 2001. - 152 с.
- Система поддержки принятия решений – Wikipedia - свободная интернет-энциклопедия [Электронный ресурс]. Режим доступа: URL: http://ru.wikipedia.org/wiki/ Система_поддержки_принятия_решений.
- Lutz M. Learning Python: Powerful Object-Oriented Programming 2nd Edition // O'reilly press, 2009. - 726 c.
- Chandrasekaran B. Generic tasks in knowledge-based reasoning: High-level building blocks for expert system design // IEEE expert, 1986. - 25 с.
- Крупенин Н.В., Голубев А.В., Завидей В.И. Некоторые аспекты технологии проведения предремонтной и послеремонтной диагностики турбогенераторов - сайт фирмы Панатест - оборудования для неразрушающего контроля [Электронный ресурс]. Режим доступа: URL: http://www.panatest.ru/static?al=turbogen_diag
- Турбогенератор - Wikipedia - свободная интернет-энциклопедия [Электронный ресурс]. Режим доступа: URL: http://ru.wikipedia.org/wiki/турбогенератор
- Средства и системы компьютерной автоматизации [Электронный ресурс]. Режим доступа: URL: http://www.asutp.ru/?p=400015
- Системы технологического контроля параметров турбо- и гидрогенераторов на базе ПТК "СТК-ЭР" [Электронный ресурс]. Режим доступа: URL: http://www.rakurs.com/system/detail.php?ID=202
- Python Programming Language – Official Website [Электронный ресурс]. Режим доступа: URL: http://python.org
The general formulation of the problem
Technical Diagnostics - a relatively new branch of knowledge, rapidly growing in recent years for two main reasons:
For the introduction of fuzzy logic methods to make the transition from deterministic database management to the space of fuzzy sets of binary diagnostic features.This requires a logical decomposition of the system, which resulted in overlapping sets of situations we find M1, M2,…, Mn. Decomposition allows us to construct a cause-effect relationships, taking into account the intersection of subsets of coefficients of reliability possible situation, describing the state space with some decomposition. The situation is described as trees. Treetops of situations associated with sensors monitoring the real state of the object.
Many situations are presented as a graph of subordination situations. In this case, the expert system generates a sequence of local problems of implementing the global challenge of. At a certain change of conditions of the object changes the degree of importance (priorities) of the individual local problems in the formation of global problems. Changing the degree of significance of individual local problems gives rise to change in order of their decisions in the process of solving a global problem. Local tasks that make up these systems should include the possibility to address them in a different order. As a model of such problems possible functional soluble with respect to any subset of its arguments. For each subset of the local problem 
there exists a function 
, which displays a subset of 
to its complement 
in the set X. This assumption means that each subset mapped by on a subset of all the remaining elements 
from the set X relative to a selected subset . Formally, the above condition has the form:
With this approach, the family of all subsets of 
, corresponds to a family of mapping functions 
. For all 
corresponding function in (1) is defined by its model (mathematical, logical) with a set of data. If we assume that the elements of any subset are known quantities that are specified decisions of the previous network of local problems, according to (1) we can calculate the values of elements of the Supplement 
. Since this possibility holds for 
, we can conclude on the universality of the local problem.
Thus, the decomposition allows the structuring of knowledge that allows you to reduce the search area at an early stage, enables informed choice of the search provides a link in the search for causes between different sets of possible situations.
Also, the technology used in the creation of a new system, you can create a web interface to the system and the use of modern database management systems. For the implementation uses the Python programming language and based on its platform Django to create a Web-based applications. Python is a high-level multiparadigmenny programming language ideal for programming tasks related to mathematical logic and Web applications [4, 10]. Django allows you to quickly and efficiently create standards-based web applications of varying complexity. As the database used by the modern post-relational database PostgreSQL 9, allowing you to create rich data model for applications. [Up]
Conclusions
Technical diagnostics of turbo is quite a complex task that depends on a large number of input parameters. Effective implementation of such a diagnostic system involves creating a large directory of known system states and actions to take when they occur. However, in the event of contingencies in the turbine, this approach can be inefficient, since a deterministic handbook can not correctly diagnose such conditions.
After analyzing the existing methods used in diagnostic systems, has been proposed transition from a deterministic database features to the knowledge base that uses intelligent methods, in particular fuzzy logic, which will remove the limit on the number of possible situations and in case of emergency situations, will allow the program to process information correctly, thereby increasing efficiency and expanding the scope of the diagnostic system. Has developed block diagram of an expert system, which allows us to solve the problem diagnosis turbogenerator using fuzzy logic methods. The developed system allows to reduce the search area at an early stage, enables informed choice of the search provides a link in the search for causes between different sets of possible situations.[Up]
Important Note: At the time of writing this essay master's work is not finished. Final completion of the work will take place in December 2011. In the near future the development, implementation and testing of the expert system.