• Nie Znaleziono Wyników

Repository - Scientific Journals of the Maritime University of Szczecin - Fuzzy sets application for ship...

N/A
N/A
Protected

Academic year: 2021

Share "Repository - Scientific Journals of the Maritime University of Szczecin - Fuzzy sets application for ship..."

Copied!
7
0
0

Pełen tekst

(1)

Maritime University of Szczecin

Akademia Morska w Szczecinie

2011, 28(100) z. 1 pp. 60–66 2011, 28(100) z. 1 s. 60–66

Fuzzy sets application for ship power plant operator safety

analysis during design process

Wykorzystanie zbiorów rozmytych do analizy bezpieczeństwa

operatora siłowni okrętowej podczas projektowania

Tomasz Kowalewski

Gdynia Maritme University Akademia Morska w Gdyni

81-225 Gdynia, ul. Morska 81–87, e-mail: tomkow@am.gdynia.pl Key words: safety, design process, ship power plant, hazard, fuzzy inference Abstract

Complex technical objects such as ship power plant are the source of many hazards to their operators. Identification and elimination of these hazards on the finished objects is very labor-intensive task and involves significant financial outlays. Therefore, it would be advisable to carry out these activities much earlier – at the design stage. However, this entails to some difficulties. Depending on the design phase the designer has a limited amount of information from which the operator’s safety can be assessed. Additionally, this information is associated with considerable uncertainty. These difficulties can be overcome by using subjective estimates of persons having practical knowledge in the field of our interest – the experts. Such knowledge can be formulated most easily in linguistic categories, that is fuzzy logic language. In the paper the basis of construction of ship power plant operator risk assessment system at the design stage are presented. This system was based on a fuzzy inference mechanism. For this purpose linguistic variables are defined. Based on these variables, fuzzy risk assessment would be carried out. Variables were related to each other within the fuzzy rule base. Fuzzy inference in the system is based on the Mamdani model. Use of this system will enable the identification of hazards for the operators of ship power plant and will indicate the necessary corrective actions.

Słowa kluczowe: bezpieczeństwo, projektowanie, siłownia okrętowa, ryzyko, wnioskowanie rozmyte Abstrakt

Złożone obiekty techniczne, takie jak siłownia okrętowa, stanowią źródło wielu zagrożeń dla ich operatorów. Przeprowadzanie identyfikacji i eliminacji tych zagrożeń na gotowym obiekcie jest bardzo pracochłonne i wiąże się ze znacznymi nakładami finansowymi. Z tego też względu celowym wydaje się przeprowadzenie tych czynności znacznie wcześniej – już na etapie projektu. Wiążą się z tym jednak pewne problemy. W za-leżności od etapu projektowego projektant dysponuje ograniczoną ilością informacji, na podstawie których można wnioskować o bezpieczeństwie operatora. Dodatkowo informacje te wiążą się z dużą niepewnością. Trudności te można pokonać, wykorzystując subiektywne oszacowania osób posiadających praktyczną wie-dzę w interesującej nas dziedzinie, czyli ekspertów. Najłatwiej taką wiewie-dzę formułować w kategoriach lin-gwistycznych, czyli języku logiki rozmytej. W pracy przedstawione zostały podstawy budowy systemu oceny zagrożeń operatora siłowni okrętowej na etapie jej projektowania. System ten oparty został na mechanizmie wnioskowania rozmytego. W tym celu określono zmienne lingwistyczne, na podstawie których dokonywana będzie rozmyta ocena zagrożeń. Zmienne te powiązane zostały ze sobą w ramach bazy reguł rozmytych. Wnioskowanie rozmyte w systemie odbywa się na podstawie modelu Mamdaniego. Wykorzystanie tego sys-temu umożliwi identyfikację czynników niebezpiecznych dla operatorów siłowni okrętowej oraz pozwoli wskazać niezbędne działania korekcyjne.

Introduction

Identification and assessment of hazards in the ship power plant is a very important for the safety

of its operators. Ship power plant is a technical object that has a complex structure. Many inter related machines and equipment works in it in

(2)

a limited space. The appropriate location of these elements in the ship power plant is very important, because it can significantly affect the safety of ope-rators. During the operator’s safety analysis should be considered not only hazards generated by ma-chinery and equipment currently operated by the operator, but also the impact of the other compo-nents of power plant located in the surroundings. On the activities carried out by the operator greatly affects a construction of surroundings. This may be manifested by limiting of the range of motion or forcing the adoption of determined positions.

The process of risk analysis can be performed most easily on the finished object that is directly at the workplace. By observing the operator during the execution of his duties may be assessed on what risks he is exposed and whether there is still pos-sibility to improve the current state. Corrective actions at this stage are associated with a large amount of work and expenses.

An alternative solution might be to carry out these activities that are identification and the risk assessment, before it comes to produce a physical object that is at the design stage. Such a solution does not require so significant amount of work and greatly reduces incurred costs. The disadvantage of this solution, however, is limited resource of know-ledge that can be used in the analysis of hazards [1, 2]. This inconvenience can be overcome with the help of experts of the field that is ship power plant experienced operators. The knowledge they possess, however, can most easily be imparted in common language using linguistic categories. Due to the complexity of risk analysis, the best solution would be to prepare for that purpose a computer system. The knowledge of experts in this case should be converted in an appropriate way to lan-guage understandable for computer. According to the author, therefore, it is advisable to use fuzzy logic. The use of such a pattern of inference allows to save a subjective knowledge which experts have acquired [3, 4, 5]. Then, using a computer the infer-ence is possible in a manner very similar to human reasoning [6].

Design stages of ship power plant and their impact on hazard assessment of operator

In the engine room, design process can be speci-fied as the following stages [7]:

 possible conceptual design,  offer project,

 preliminary design (contract),  technical-classification design,

 design workshop (working),  passing documentation (operating).

With the progress of project, the range of avail-able information on the construction of the ship power plant is gradually widening. Too early phas-es of the project make it impossible to gather enough information to be able to identify the real hazards of the operator. In the final stages of the project, when they are already known all details, making changes is very time consuming and re-quires considerable work. Therefore, it was as-sumed that impact on operator’s safety will take place at the preliminary and technical-classification design stages [7].

Preliminary design, among other things, speci-fies the exact location of the power plant on the ship and determines more important mechanisms and power plant equipment with their technical characteristics.

Technical-classification design includes dia-grams of pipelines of particular power plant instal-lations, solutions of foundations construction and fixings of the main of machinery and equipment, construction of shaft lines together with the neces-sary strength calculations, vibrations, etc., the dis-mantling plan of machinery and equipment in the power plant, plans of workshops and storage areas.

Because of these differences, in the available knowledge system will operate in two stages. Based on the preliminary design stage, information about the hazards arising from the work of different tech-nical devices will be collected. This makes it pos-sible to initial determine power plant areas which may constitute a potential risk to the operator. Dur-ing the next phase – the technical-classification design – these areas will be subject to further as-sessment, taking into account new information about their construction.

This method of evaluation allows identifying the factors posing a risk to the operator and enables corrective action to be taken.

Risk assessment in the preliminary phase of design

On the basis of the analysis carried out in [7], dangerous and harmful factors causing risk to the operator conventionally have been divided into two groups:

• functional factors – dependent on the function performed by a technical device,

• operational factors – related to the operation of technical devices.

The group of functional factors includes the fol-lowing hazards:

(3)

 chemical hazard;  thermal hazard;  pressure hazard;

 hazard due to work environment factors (noise, vibrations, air parameters and composition, etc.);

 mechanical energy hazard;  electric energy hazard.

The set of operational factors include risks re-sulting from the implementation of the procedures for:

 operation,  maintenance,  supply,  safety control.

The system of engine room operator’s risk as-sessment was built based on functional factors. For each of these factors sets of their manifestations

have been assigned. Functional factors have been considered in relation to:

 the possibility of operator’s contact with the given type of hazard,

 level of risk for the operator in case of direct contact with the given type of hazard.

Creation of an assessment system required to develop a fuzzy membership function shapes of individual fuzzy sets. This task was carried out by a research questionnaire specially created for this purpose [8]. Based on expert opinion for each type of hazard, sets determining the possibility of con-tact with this hazard in the engine room and the operator’s level of risk exposed to direct contact with the given type of hazard were developed. Ex-amples of fuzzy sets for the mechanical energy hazards are shown in figures 1 and 2.

During the assessment of selected elements of engine room each of functional factors generates

Fig. 1. Fuzzy sets presenting a mechanical energy hazard in the case of direct contact

Rys. 1. Zbiory rozmyte przedstawiające zagrożenie energią mechaniczną w przypadku bezpośredniego kontaktu

Fig. 2. Fuzzy sets representing the possibility of contact with the particular manifestations of mechanical energy hazard Rys. 2. Zbiory rozmyte przedstawiające możliwość kontaktu z poszczególnymi przejawami zagrożenia energią mechaniczną

0.6 0.5 0.4 0.3 0.2 0.1 0 0.8 0.6 0.4 0.2 0

(4)

a certain risk level of the operator. Its value is de-termined on fuzzy sets presented in figure 3.

Obtaining the resulting value of risk level for the given functional factor is performed using fuzzy inference mechanism based on the Mamdani model [2, 6, 9, 10]. The overall level of risk resulting from the impact of all functional factors is obtained after the aggregation, using the same model of inference [11, 12]. Using this method of the operator risk assessment allows to initial determination of areas in which performed actions may pose a potential risk to operators.

Risk assessment in the later stages of design

At the technical classification design stage available knowledge enables a more detailed defini-tion of risk level of the operator. It can be assumed that it will be a function of factors derived from [13]:

 operation of machinery and equipment,

 accessibility to the place where the operational activity is carried out,

 operator’s position when performing certain operational activities,

 type of performed operational activity.

Risks associated with the operation of machin-ery and equipment are dependent on their function. According to the author, it would also be advisable to consider the impact to safety of the operator from the machinery or equipment (general structural unit1) at:

 identified functional hazards associated directly with operated structural unit Hu and unit in

sur-roundings of an operator Hs (preliminary design

of power plant),

1 Structural unit is a technical object very widely

inter-preted namely as any item, part, device, machinery and equipment, installation, which can be considered sepa-rately (based on [9]).

 remedial measures already provided in the pro-ject RM.

Assumed that the accessibility to the place of carried out operational activities (APA), together with the location of structural unit at the correct height (LSU), significantly affect the course of the work performed by the operator and the position adopted by him when performing maintenance. These values will define the (ON) operational nuisance.

In the case of the impact of their operations on the operator’s risk, should be considered such factors as:

• degree of activities differentiation (AD) – num-ber of different elementary operations performed at the specified structural unit,

• maximum range of performed movements (RPM) – way of performing activities, for ex-ample by hand, arm, with or without the use of tools, etc.,

• variability of adopted position (VP) – infor-mation that specifies the dynamics of the movements made by the operator.

These factors will determine the complexity of operations (CO). Relationships between factors in the risk assessment system of the operator are shown in figure 4.

Building a fuzzy risk assessment system of the operator requires the creation of fuzzy representa-tion of individual factors and their relarepresenta-tionships by the relevant rules. Initially, for purposes of repre-sentation of individual linguistic variables, fuzzy sets have been adopted by the triangle and trape-zoid shapes. Examples of fuzzy sets describing the factors that make up the complexity of operations (CO) are shown in figures 5–8. Under the numeri-cal values on the abscissa different cases difficulties in accessing to the structural unit are included. For example, for a value of 10 was adopted access to a unit using only the fingers with the simultaneous need to lean the operator. Under the particular Fig. 3. Resulting fuzzy sets – risk level of the operator

(5)

Fig. 4. Diagram showing the operator's risk assessment system Rys. 4. Schemat obrazujący system oceny zagrożenia operatora

Fig. 5. Fuzzy sets describing the activities differentiation Rys. 5. Zbiory rozmyte opisujące zróżnicowanie czynności

Fig. 6. Fuzzy sets describing the range of morion

Rys. 6. Zbiory rozmyte opisujące zakres wykonywanych ruchów

Fig. 7. Fuzzy sets describing the variability of operator’s positions Rys. 7. Zbiory rozmyte opisujące zmienność przyjmowanych pozycji

(6)

numerical values on the axes of abscissa manifesta-tions of individual variables are hidden. For exam-ple, the range of performed movements (RPM) at figure 6 includes activities carried out by using fingers (1), hands (2), ..., both hands and tools (10). In this case, the maximum range of movements performed by the operator for specific activities is considered.

Variability of adopted position (VP) at figure 7 takes into account whether the operator performs actions in the forced positions and which positions must be taken during operations.

Fuzzy rules are created in the form of IF-THEN expressions. If more premises the AND operator is used. Created fuzzy rules have the following form:  IF activities differentiation AD is high, AND

range of performed movements RPM is high

AND variability of adopted position VP is high THEN the complexity of operations CO is high.

 IF complexity of operations CO is high, AND the hazard level Hu is high AND operational

nuisance ON is substantial THEN risk level RL is high.

Information needed for the system operation are collected based on the designer dialogue with this system. Depending on the needed information sys-tem will collect data by properly prepared windows that require making a specific choice (Fig. 9), or for example situational schemes which require com-pleting them by values needed to evaluate the level of risk.

The obtained information is converted into corresponding fuzzy sets based on the knowledge accumulated in the system base. For example, from Fig. 8. Fuzzy sets describing the complexity of operations

Rys. 8. Zbiory rozmyte opisujące stopień skomplikowania czynności

Fig. 9. Sample screen of the operator risk assessment system Rys. 9. Przykładowy ekran systemu oceny zagrożeń operatora

(7)

the window shown in figure 9 details about the type of performed activities and their quantity are ob-tained. Thanks to this, fuzzy sets concerning differ-entiation of activities are defined. Knowing also how the operations will be carried out (e.g.: by using hand and tools), can be determined the range of movements performed by the operator RPM. After gathering information concerning the varia-bility of positions VP adopted during operation, on the basis of relevant fuzzy rules, information about the complexity of operations CO is obtained. This information, in connection with other data (ob-tained in a similar manner) concerning the opera-tional nuisance and the hazards generated by the structural units, allows for the final risk assessment of the operator.

The effect of the risk assessment process of the operator performing specific operational activities is a value in the range 0, 10. This value allows for the classification of level of this risk as low, medium, or high. Depending on the result, it will be possible to take appropriate action to improve the safety of operator. Data collected by the system will allow for the identification of factors that signifi-cantly jeopardize the operator.

Conclusions

The risk assessment carried out on the stage of preliminary design of the engine room due to the limited amount of information provides an initial opportunity to identify areas of risk. It is very im-portant that at this early stage of the project can be provided specific preventive measures. However, with the development of the project there are new, not included in this assessment, potentially hazard-ous for the operator factors. Then, it becomes nec-essary further consideration of hazardous areas, taking into account just these factors. In addition, it is necessary to verify measures already introduced after the preliminary design stage.

Application of fuzzy inference allows inter-operability of assessment systems for the various phases of the project. Moreover, the use of linguis-tic variables in this inference model allows reflect-ing the imprecision of natural language used by experts. Fuzzy inference is very similar to human reasoning, so it becomes easier to process the rele-vant knowledge into the language understood by the computer.

References

1. ELOFF J.H.P., DE RU W.G.: Risk analysis modeling with the use of fuzzy logic. Computers & Security, Vol. 15, No. 3, Elsevier Science Limited, 1996, 239–248.

2. TAREŁKO W.: Metodologia projektowania właściwości eksploatacyjnych złożonych obiektów technicznych. Bi-blioteka Problemów Eksploatacji, Gdynia 2011.

3. MURE S.,DEMICHELA M.,PICCININI N.: Assessment of the risk of occupational accidents using a ‘‘fuzzy’’ approach. Cogn Tech Work, 2006, 8, 103–112.

4. PEREEZDCHIKOV I.V.: Methodology of complicated system hazard analysis by the use of sharp and fuzzy sets. Safety, diagnosis, and repair. Chemical and Petroleum Engineer-ing, Vol. 41, Nos. 7–8, 2005.

5. SII H.S., RUXTON T.,WANG J.: A fuzzy-logic-based ap-proach to qualitative safety modeling for marine systems. Reliability Engineering & System Safety, 73 (2001), 19– 34.

6. ŁACHWA A.: Rozmyty świat zbiorów, liczb, relacji, faktów, reguł i decyzji. Akademicka Oficyna Wydawnicza EXIT, Warszawa 2001.

7. TAREŁKO W.i in.: Antropotechniczne założenia do projek-towania bezpiecznych obiektów technicznych – sprawoz-danie merytoryczne. Sprawozsprawoz-danie z realizacji projektu badawczego nr 8T07C 011 20 finansowanego przez KBN, Zeszyt nr 1, KPT, Akademia Morska w Gdyni, Gdynia 2004.

8. TAREŁKO W. i in.: Metoda projektowania bezpieczeństwa operatorów złożonych obiektów technicznych – sprawoz-danie merytoryczne. Sprawozsprawoz-danie z realizacji projektu badawczego nr 4T07B02529 finansowanego przez KBN, Zeszyt nr 1, KPT, Akademia Morska w Gdyni, Gdynia 2009.

9. KOWALEWSKI T.,TAREŁKO W.: Propozycja zastosowania wnioskowania rozmytego do oceny stopnia zagrożenia ope-ratora obiektu technicznego. Zeszyty Naukowe nr 56, Aka-demia Morska w Gdyni, Gdynia 2006.

10. PIEGAT A.: Modelowanie i sterowanie rozmyte. Akademic-ka Oficyna Wydawnicza EXIT, Warszawa 1999.

11. KOWALEWSKI T.,PODSIADŁO A.,TAREŁKO W.: Application of fuzzy inference to assessment of degree of hazard to ship power plant operator. Polish Maritime Research, No. 3(53), Vol. 14, Gdańsk 2007.

12. KOWALEWSKI T.,PODSIADŁO A.,TAREŁKO W.: Analysis of hazard to operator during design process of safe ship power plant. Polish Maritime Research, No. 4(67), Vol. 17, Gdańsk 2010.

13. WINKLER T.: Komputerowo wspomagane projektowanie układów antropotechnicznych. WNT, Warszawa 2005.

The paper was published by financial supporting of West Pomeranian Province

Cytaty

Powiązane dokumenty

Choć początki cywilizacji chińskiej są bardzo odległe, długie też są dzieje państwowości chińskiej, do przybliżenia wyglądu i symboliki flag i herbów Chińskiej

funkcji produkcji nale˝y wprowadziç zmiennà zarzàdzania Z, która wyra˝ona jest przez nast´pujàce parametry: wskaênik rotacji aktywów z, stratnoÊç aktywów s, poziom

Założono występowanie silnej kompensacji cech ocenianych wyrobów: pomimo różnych wartości poszczególnych cech diagnostycznych wszystkie towary otrzymały taką samą – w

Wpływ postępowania wszczętego wskutek zaskarżenia uchwały o przekształceniu na bieg postępowania rejestrowego uzależniony jest od szeregu okoliczności, wśród których

Plan dostępu do zasobów węgla kamiennego w latach 2004–2006 oraz Plan zamknięcia kopalń w latach 2004–2007, Ministerstwo Gospodarki i Pracy, Warszawa, 7 września 2004.. Polska

Mam świadomość, że aktualna sytuacja na Węgrzech jest ciągle przedmiotem debaty prowa- dzonej przez krytycznych analityków krajowych i zagranicznych: czy Węgry mogą być nazywane

Pomimo obserwowanych w ostatnich kilku latach spadków wydatków, sumy wydawane obecnie na promocję są nadal znacznie wyższe w porównaniu do okresu sprzed dekady (jak

civil relations between economic subjects involved in economic activity basing on the principles defined in separate provisions. 2 section 2) includes the