Abstract
Content
- Introduction
- 1. Actuality
- 2. The purpose and objectives of the study, the expected results
- 3. The total content
- 3.1 A brief overview of the features of the subject area
- 3.2 Description of the approach to classification of type of accident in the mine
- 3.3 The structure of the classification system
- 4. Conclusions
- Conclusions
- References
Introduction
Mining production is a complex and dangerous industry. In the course of the work can be various hazardous events of natural and technological nature. Therefore, the problems of safety at mining operations are of particular importance. Analysis and assessment of the risk of possible accidents is one of the key issues of industrial safety. For comprehensive coverage of production conditions apply control systems, such as UTAS. [1]
Assessment of emergency exercises mountain manager. If you receive information about possible emergency or pre-crash situations, mining manager need to quickly, objectively and comprehensively evaluate the productive environment for the adoption of the current management solutions. [2]
The analysis is based on sensor readings issued by the control system in use or oral notification of the employees.
At this stage, there are often difficulties. Since automatically measured only a small number of indicators, and the available information has a high proportion of ambiguity. As a result, there is no unambiguous rules, criteria and methods for assessing emergency situations. Also, there is a high dependence on the human factor. Remedy or account of this uncertainty is now often based on the experience of the supervisor or other experts and are insufficient. This leads to erroneous decisions, which can lead to serious consequences.
1. Actuality
Workers' livesdependof production processes control level and the environment in the workings. In the process control of the mine is very important to the timeliness and accuracy of obtaining the necessary information on the status of the production environment. Therefore, since the 1970 of the last century, the coal industry is widely implemented automation equipment, automated process control system (APCS) system and of the supervisory control (SSC). [3, 4]
However, a purely mechanical fixation is often not enough. Lack of funds a comprehensive analysis of the data creates difficulties in decision-making, by fixing the manufacturing violations. Unambiguous interpretation of the problem is identified key indicators of deviations from the norm to the form of an emergency. This creates difficulties for accurate and objective classification of the overall type of accident. But the development and application of appropriate software can improve, streamline and simplify the job.
Thus, it was found that the problem of creating intelligent systems for determining the type of emergency, is relevant for coal mines.
2. The purpose and objectives of the research, the expected results
The ultimate goal of this
work – to realize intelligent
decision support system for mining manager with the elements of fuzzy
logic and methods of classification and artificial intelligence to
solve the problem of determining the type of emergency situation in
real time. [5]
Object
of study:
a description of the methods and techniques and identify conditions of
the production process in the mine.
Subject
of research: methods of
classification of emergencies at
the mine.
The main objectives of the study:
- To analyze the problem of classification of emergencies at the mine.
- To analyze the ways to determine the status of the production process.
- Choose a set of indicators to assess the current state of the situation of the working environment of the production process.
- Describe the characteristics of key performance indicators in terms of the classification of accidents.
- To study the characteristics of accidents and describe them using a dedicated set of indicators.
- To develop methods of classification using the elements of fuzzy logic and artificial intelligence.
- Develop a system of classification of the form of an emergency.
- Rate the quality of the classification
3. The total content
According to the theme, the work discusses the features of accidents at
coal mines in Ukraine. In order to simplify the system, we consider
only the most dangerous emergency situations of natural
character.
[4]
Assessment of emergency exercises mountain manager. In the event of an emergency its main task is to control the optimal course of liquidation of emergency situations to save lives and minimize the consequences of the accident. [2]
If the type of accident is known, the next steps are regulated by the appropriate instruction Manager Emergency Response Plan (ERP). ERP consists of instructions on how to eliminate each dedicated to the mine plan and position for each type of accidents. This guide is designed for all operational, under construction or reconstructed mine. [4] But type of accident, however, are often initially unknown or not obvious.
Generalized qualitative characteristics of the accident is shown in Figure 1. [6]
Figure 1 denotes:
tno – the normal operation of the facility;
tw – worsening state of the object;
tра – the period of the accident;
tlа – the period of liquidation of the accident;
А – the point of the critical state of the object.
It is important to note that often the decision on the form of the measures and emergency response is performed at the stage of having a clear state of emergency. That corresponds to the interval of time after point A in Figure 1.
This paper describes the development of classification system for emergencies on the. To develop such a system used by the capabilities of existing systems for monitoring and control of rock [1, 3] production combined with the use of modified methods of fuzzy logic and artificial intelligence. [5]
3.1 Brief overview of the features of the subject area
As part of the subject area is selected, that the coal mines there are four main types of underground industrial accidents are not of a technical nature, such as: fire, explosion, collapse and flooding or water breakthrough. [4, 7 ] .
Emergency situations characterized by analyzing a small amount available to measure such parameters as temperature T, the level of CH4, CO carbon dioxide level, the direction (+ / -) of air V; presence / absence (+ / -) of the current I in the base equipment and a common power supply lines. There are normative data on safety, describing the testimony of the measured parameters automatically. [5]
These data are recorded by an automated process control system or of the supervisory control. The paper discusses these two popular and highly efficient system as ADCS and UTAS. [1, 2]
The same type of emergency confirmed by analysis of key phrases extracted from a set of possible standard voice notifications. [5]
3.2 Description of the approach to classification of type of accident in the mine
According to the safety standards and regulations, established the possible ranges of readings automatically measured parameters. Ranges scaled to appropriate terms identified by the experts of linguistic variables "low", "medium", "high", "very high" risk . [8] In the developed system of classification, a qualitative analysis of the situation begins when any of the critical parameters measured before pre-alarm limit, which corresponds to the "low" or "medium" level of risk.
Also very effective kind of accident can be categorized by key phrases communication experts with a high probability can be correlated with a certain type of accident. A few key phrases can be complementary and to make certain the chain, classifying sequences. A set of these rules may, as a self-classify the situation and used to complement the basic rules of the system.
As a result of this work is designed Expert System for the Classification of accidents at the mine. It can be used as a support system to support decision-making in the activities of mining dispatcher.
3.3 Structure of the classification system
The developed system consists of three logical units that form stepwise classified crash. The classification is based on the ratio of the current state of the measured parameters and classifying facts to prospectively possible types of situations using established methods.
Block number 1 conducts initial analysis of the criticality of the situation on the basis of data received from the sensors. поступающих с датчиков.
Block number 2 sets of analyzes of the facts that help classify the characteristics and type of emergency (see section 3.3). The results of the block number 1 and number 2 are checked for inconsistency in the module «V».
In block number 3 is the final analysis and classification of the accident. [5]
Logical-formal model designed multi-level classification system crashes is presented in Figure 2.
Figure 2 – Logical formal model of multi-level classification system crashes
Where:
p1
- рn
– the numerical value of the measured parameters (level of
methane
– CH4, gaz
– CO, air temperature
– T);
f1
- fn – revealed facts
that describe the voice of
experts from the mine;
k1,
k2, k3 –code is
generated at the output of block №
1, № 2 and number 3, respectively;
REZ – interpret the results of the final classification (the message of the proposed form of the situation and the level of confidence in the accuracy of the result).
4. Future work
The developed approach to the classification of an emergency on mine is the base. Further improvements are possible at the expense of use of statistics and the expansion of the rule base. This system will be further developed by the classification rules for other types of accidents.
However, the compilation of the rules and their optimization is a very complex process. This requires close interaction with experts. A rating system performance is very difficult due to the lack of available statistical analysis in this field.
Conclusions
The paper analyzes are used in the mine control systems. the available indicators of basic emergencies are selected, structured and categorized. The model of the system of classification of mine emergencies is presented and described its basic principles.
The developed classification system is an expert system that uses artificial intelligence techniques and elements of fuzzy logic.
This system is designed for use in mine mountain manager. The use of such a system will speed up the process of determining the type of accident, and reduce the number of failed or untimely decisions and reduce human error, due to the continuous automatic analysis of all the available options.
Important
Note: When writing this master's
work is not completed. Final completion: January 2014. Full text of the
materials can be obtained from the author or his manager after that
date.
↑ Тор