Skip to Main content Skip to Navigation

Optimization of the availability of multi-states systems under uncertainty

Abstract : Dependability has become a necessity in the industrial world during the twentieth century. Dependability is an activity domain that proposes means to increase the attributes of the system in a reasonable time and with a less cost. In systems engineering, dependability is defined as the property that enables system users to place a justified confidence in the service it delivers to them and it is a measure of a system’s availability, reliability, and its maintainability, and maintenance support performance, and, in some cases, other characteristics such as durability, safety and security. The key concept that our work is based on is the availability. The availability A(t) is the ability of a system to be operational at a specific moment. The cost of some system with high availability is very expensive. The designer must compromise between the availability and the economic costs. Users can reject systems that are unsafe, unreliable or insecure. Therefore, any user (or industry) will ask this questionbefore getting any product: "What is the optimal product in the market?" To answer to this question, we must combine the following two points : - The best availability of the system : the user wants a product that lasts as long as possible. - The best cost of the system : the user wants a product without costing him a fortune. Availability calculation is based primarily on knowledge of failure rates and repairs of system components. Availability analysis helps to calculate the ability of a system to provide a required level of performance depending on the level of degradation. Several methods have been used to calculate the availability of a system, amongst which we find the Universal Generating Function (UGF), Inclusion-Exclusion technique, Markov models, etc. These methods employ different probabilistic techniques to evaluate this criterion, but these proposed approaches remain effective only for very specific cases, for example the cases of binary systems. A binary system is a system where only two cases are possible : perfect functioning and total failure. While the transition to multi-state systems (MSS) drastically restricts the application of most of these methods. In real life, the systems corresponds to MSS. In such scenarios, systems and their components can operate at different performance levels between working and failure states. However, the evaluation of the availability of the MSSs is more difficult than in the binary case, because we have to take into account the different combinations of the component failure modes. Throughout this thesis, we search for a method that helps us to compute and to optimize the availability of MSS.
Complete list of metadatas

Cited literature [135 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Tuesday, October 13, 2020 - 10:45:07 AM
Last modification on : Tuesday, October 20, 2020 - 8:57:59 AM
Long-term archiving on: : Thursday, January 14, 2021 - 6:27:53 PM


Version validated by the jury (STAR)


  • HAL Id : tel-02965302, version 1



Joanna Akrouche. Optimization of the availability of multi-states systems under uncertainty. Other [cs.OH]. Université de Technologie de Compiègne; Université libanaise, 2020. English. ⟨NNT : 2020COMP2545⟩. ⟨tel-02965302⟩



Record views


Files downloads