Fitness-oriented multi-objective optimisation for infrastructures rehabilitations

A new multi-objective decision support system (MODSS) is developed for rehabilitation planning of public infrastructures. The method is generic and provides decision-makers a set of optimal rehabilitation tradeoffs over a desired analysis period. Two main objective functions are handled simultaneously, namely cost and performance, in addition to a set of bounding constraints. The method is based on a new fitness-oriented technique where problem knowledge is taken into account. In order to analyse cost and performance together, a normalisation technique of both objectives is achieved through an innovative time-value concept for both cost and condition states. The proposed method is based on life-cycle costing (LCC) methodology using a dynamic Markov chain to represent the deterioration mechanism and genetic algorithm is used to find the optimal rehabilitation profile. A case study is presented with a comparison between the traditional Markov decision process (MDP) and the newly developed method. The MODSS results in a lower LCC and is found practical in providing a complete maintenance and rehabilitation plan over a required study period. It is proven that the developed multi-objective optimisation is an effective tool in analysing real-life situations involving conflicting goals. Also, weighted sum method could be easily implemented and its outcome is sufficient given that many external factors might alter the decision-makers choice irrespectively of the optimisation method that is used. Furthermore, genetic algorithm is proven useful in the optimisation process in overcoming the computational difficulties associated with large combinatorial problems. The new method is beneficial to researchers and practitioners as it provides a major step towards a broad infrastructure management system.