Modern energy objects are very expensive systems. Improved operational reliability of such systems is inextricably linked to the development of modern means and systems of technical diagnostics [1–2].
Analysis of the current state of diagnostic software in power engineering in Ukraine shows a number of problems that are hindering its development. These problems relate to both the methodological and hardware, and algorithmic–software. Determination of technical condition of equipment is controlled by a number of the documents. However, in most cases they are not designed as a unified system with clear interdependent separate operations aimed at forming diagnostic conclusions and decisions.
Even new power equipment, which is entered into operation, is not equipped with means of technical control. Existing tools are not diagnostic systems, their main function is to measure, the primary processing and display of individual parameters. Diagnostic output is given by the staff.
The existing scheme of information processing has no possibility of a comprehensive quantification of the causes of lower reliability on the one hand, and assess the impact of various factors on the rise – on the other [3–11].
Most important result of the use of expert systems in power plants is the possibility of reducing the cost of electricity production.
The main differences between the control and management of complex power equipment and systems of their diagnosis are more pronounced in recent years.
Experience with expert systems has shown that the most efficient they can bring in when they use online information in the process of the equipment and when they integrate into an automated control system of TPP [12].
Aim of the masters work is the improving the shell tool and knowledge base of expert system diagnosis of the condition of electrical units TPP and the creation of applied expert systems.
Expert system diagnosis of various equipment is widely distributed in various areas, including in electrical engineering.
The main quantity of such systems, conducts regular monitoring, while others may give information not only about controlled objects, but also to ascertain something or other damage.
The novelty of the expert system of turbine generator, which we develop and improve, in that in addition to the above features, it has one significant advantage which consists in giving to the user not only to the monitoring results and the fact of failure or bad operation mode, but also solutions to the problem situation. That gives further support to the dispatcher in making decisions on the implementation of operations.
Also, this expert system can operate in a simulator, which is very important because it can be used for training personnel.
Developments on the topic held at the Department of Electrical Systems of DonNTU. Expert systems for power is actively being developed in Ukraine and abroad. In the U.S. and the EU, such systems have long been working on energy companies, but to get information on them at the moment is quite difficult. This is due to the fact that the developers do not provide information on the principles of construction of their expert systems, which in turn does not fully appreciate their advantages and disadvantages.
Scheme of expert system
The user interacts with the expert system through the user interface. This interface represents a system of menus, buttons and graphics.
Block that defines the damage consists of two parts: the Task Scheduler and inference decision. With task scheduler implemented support strategies for solving the problem. Choosing a strategy decision is based on the analysis of linkages of situations describing the cause–effect relationships.
Display unit is required to display the results on screen display solutions. For a better perception of the results there ara used block diagrams.
Information model of expert system is a base of facts, data and knowledge.
The database is the part of an automated database management system of the local objects of electricity system. The database is implemented relational data model. Database of the expert system contains the information about the object and its auxiliary systems, which is required in decision–making process in the analysis of the abnormal situation. This information includes normative data on the operation.
Sources of information for the database of facts is a system for collecting information from sensors installed or the results of testing equipment.
The knowledge base is divided into two levels. The lower level is a set of rules, and the upper level – the description of cause–effect situations.
These descriptions are a logical flow diagram of certain violations of the facility and systems to ensure its work. The interaction of personnel with the information model through the appropriate interfaces.
Maintaining the knowledge base associated with the use of decision tables, identifiers describing sensors, boards, relays, indicators and properties of objects and systems. Maintain the integrity of the knowledge base, improving database management, reducing the probability of making mistakes personnel are necessary requirements to use interface.
ES diagnosis of the condition of the generator can operate as a node of the local area network, and as part of the automated workplace (AWP), the shift power. ES can operate in two modes: automatic mode and the "operational control".
In automatic mode, the PC always included in order to control the generation of electricity. In this mode, the software manages the process of transferring data from an information collection system, controls the serviceability of measurement channels, identifies deviations of controlled parameters of the limit values, searches for the causes of accidental deviations, comparing current information on the rules entered into BR generates guidance to staff. ES uses the information from 60 sensors installed in different nodes of a turbogenerator.
By exceeding the parameter setting value on the display screen image of the sensor changes color and starts the ES for the analysis.
If the automatic entry of information into a PC is not guaranteed, the operating personnel can enter the appropriate information into ES.
In the "operational control" mode there is following functions: monitoring the current state of the generator and systems to provide search for the cause of deviation parameters for the limiting values.
Figure 1 – Screen when the automatic mode is in use (cyclic gif – animation, 9 frames, sise 108 kb)
ES allows to determine the causes of change in temperature of active parts of the generator, the hydrogen pressure in the casing of the generator, the voltage on the findings of a generator, stator and rotor current, the pressure drop "oil–hydrogen" at the seals of the generator, etc., and assess the dynamics of change in operating conditions unit.
A knowledge base developing
There are currently developed decision tables and files a knowledge base for the following situations:
Under development, an expert system gives the correct solution for situations where decision tables are developed. Thhat makes it possible to give a conclusion about its effective operation.
The use of expert systems in power is an effective tool for reducing the number of operational errors. This is due to some emotional relaxation and increased confidence in the workflow.
The preliminary results of research and development shows great practical significance of such systems, which is no doubt is an essential stimulus for further improvement and increase user data software products quality.
At writing of the abstract master's work isn't finished yet. The definitive variant of work can be received at the author or the scientific adviser after December, 2010.