Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.757203
Title: Integrated performance-reliability optimisation of systems with multi-level redundancies
Author: Ikegwuru, Okachi
ISNI:       0000 0004 7430 0217
Awarding Body: Northumbria University
Current Institution: Northumbria University
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Redundancy allocation, in the context of reliability driven design, is the process of multi-objective optimisation of system configuration with reliability and cost related objectives. Large systems, of any type and discipline, can be divided into several subsystems comprising modules and components. Such a hierarchical form of system arrangement is regarded as multilevel configuration. These systems have the performance capability beyond traditional binary reliability framework of either completely working or totally failed. Large systems normally have redundancies at different levels. In current practice, multi-level redundancy allocation takes place sequentially. This is mainly due to lack of a robust optimisation method capable of delivering large scale redundancy allocation problems. Development of such methods leads to design of enhanced systems with better performance in terms of cost and reliability. The overall aim of this project is to develop a method for multi-state reliability optimisation of large real-world systems. To achieve the overall goal, firstly, a genetic algorithm (GA) suitable for analysis of systems with multi-level redundancies is developed. For this GA, new multi-level chromosome, new crossover and mutation operators capable of combining building blocks at different level and mutation of solutions at various levels are designed. Whilst the GA chromosome and regeneration operators are specially designed for handling multi-level systems, in the second step a Non-dominated sorted genetic algorithm (NSGA-II) is developed for multi-dimensional search towards finding Pareto frontier solutions with respect to a number of cost-related, performance-related and reliability-rated objectives including cost, size, weight, availability and failure rate. In the final stage, the developed search and optimisation methods are implemented in a software tool written in MATLAB. Employing the optimisation tool for benchmark problems with multi-level redundancies, heating, ventilation and air conditioning (HVAC) systems, it has been shown how an integrated multi-level redundancy allocation, as opposed to sequential redundancy allocation, can lead to superior solutions.
Supervisor: Maheri, Alireza Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.757203  DOI: Not available
Keywords: H900 Others in Engineering
Share: