RELIABILITY BEHAVIOR ANALYSIS OF SOME INDUSTRIAL SYSTEMS USING IFS THEORY

Abstract

In an industrial system, great attention is to be paid to fulfill various requirements of consumers of its products as well as to improve the system performance in terms of quality, reliability and safety with the advancement of the technologies and complexity of the systems. To avoid the high risks of process failure and safety and to analyze the relation between customers satisfaction, system profit and resources consumption, it is required to have proper knowledge and information about the performance behavior of systems with respect to the random failures and different operating conditions. The behavior analysis includes the study of main reliability attributes such as system reliability, availability, maintainability and risk and safety analysis of the system as well as of its components/units. These attributes basically provide the phenomenon to characterize and measure the performance of systems with some likelihood of failures so that efforts are made to increase the system design life and its operational availability by incorporating suitable maintenance strategies. The modeling of systems in terms of the states is desirable to analyze their operating behavior with the use of various quantitative and qualitative tools. To analyze the failure pattern as well as corresponding performance of the systems, the data regarding the component failures and repairs are collected either from past records or experiments or expertise decisions, assumed to be precisely known. These data may not necessarily be precise as they have various types of errors such as inability to predict future failure, manhandling error, diverse operating conditions, environmental error and lack of knowledge of expertise in the area. This uncertain data gives rise to inadequate behavior of system performance which affects the optimum performance having the desired goals. It becomes tough job for system analysts to construct a mathematical model that corresponds to the real situation and demonstrates the uncertain behavior of industrial systems properly. The main objective of the research work is to analyze the behavior of some repairable industrial systems such as Sugar industry, TAB industry, Butter-Oil Processing plant and some ii series-parallel configurations such as five-component series system, complex-bridge system, overspeed protection system of gas turbine, life support system in a space capsule and mixed series-parallel system. Markov process has been applied to model the different states of the Sugar industry and its availability optimization is developed using the obtained differential equations. To find the optimal parameters of component failure rates and repair times of each component for the Feeding and Refining systems of Sugar industry the Genetic Algorithm (GA) has been applied. As the data taken in the study are imprecise, the Intuitionistic Fuzzy set theory has been incorporated to deal with the uncertain behavior of TAB industry and Butter-Oil Processing plant. System reliability indices such as system failure rate, mean time between failures (MTBF), mean time to repair (MTTR), expected number of failures (ENOF), reliability and availability are calculated which help the decision makers to choose the appropriate decision and maintenance policy for the improvement of system operating performance. Apart from single objective problems, multi-objective reliability optimization problems are also considered during the research with fuzzy environment. In the reliability optimization problems with reliability and cost as the objectives, the component reliabilities are assumed as fuzzy numbers and then a fuzzy optimization problem is generated which then is converted into the crisp one by using the Intuitionistic fuzzy programming technique with the help of membership and non-membership functions. The final optimization problem is solved by Particle Swarm Optimization technique (PSO) and GA and then the resulted values are compared with each other. Another multi-objective reliability optimization problem i.e. multi-objective reliability-redundancy allocation problem (RRAP) is discussed by the concept of fuzzy theory to analyze the uncertainty in the system data with the formation of a single objective RRAP with the decision maker’s preferences. The problem is again solved by using the same optimization techniques PSO and GA with the statistical simulations. This thesis is divided into seven chapters. Chapter 1 provides the introduction of work related to evaluation of system reliability/availability, performance analysis. Chapter 2 covers the preliminaries and basics associated with the analysis of system reliability, modelling of systems and optimization techniques which have importance in the subsequent chapters. Chapter 3 describes the performance analysis of a repairable system, Sugar industry by using application of Markov process. Two of its subsystems i.e. Feeding and Refining systems are studied through the process. Similar analysis may be done for other systems. The optimization of steady state availability problem is done by GAs to find the optimal MTBF and MTTR. iii These parameters may assist the system analyst in achieving optimum system performance with suitable maintenance actions. Chapter 4 analyzes the optimization of bi-objective reliability-redundancy allocation problem for complex bridge system and overspeed protection system using the fuzzy set theory. The final results obtained from this analysis are compared with each other and statistical simulations are performed. Chapter 5 describes a methodology for solving multi-objective reliability optimization problems which is mainly based on Intuitionistic fuzzy programming technique. Two objectives as reliability and system cost are taken for the analysis with the condition that the component reliabilities are triangular fuzzy numbers. The five-component series system, life support system in a capsule, complex bridge system and series-parallel system are considered here for the illustration of the approach. In Chapter 6 the vagueness involved in the parameter values of TAB industry and Butter oil processing plant (BOPP) is analysed by using the concept of Intuitionistic fuzzy set theory (IFS). This analysis enables a more useful way of dealing vague systems with the chances of having better performance. Chapter 7 gives the conclusion of the entire study and the scope for future research.