Multi- Objective Evolutionary Algorithms for a Preventive Health care Fac ility Network Design

Preventive healthcare aims at red ucing the li kelihood an d severity of potentially life-threate ning illnesses by mean s of protection and early detection. In this p aper, a bi-objective mathematic al model is proposed to design a network of preventive healthcare facilities in which each facility ac ts as M/M/1 queuing system so as to minimize total trave l and waiting time as well as establishment and staffing cost. The number of facilities to be established, the loc tion of each facility, a nd the leve l of techno ogy for each prospect facility are provided s the main determinan ts of a healthcare facility networ . Since the developed m odel of the problem is of an NP-hard type, tr - meta-heuristic algor thms are proposed to solve t he problem . Initially, Pareto-based meta-he ristic algorithm, whi ch is called multi-objective simulated anneali ng (MOSA), is propose d to solve the problem. Subsequently, obtained results are validated by means o f two popular algorit hms, namely non-dom inated sorting genetic algorithm (NSGA-II) and non-do minated ranking genetic algorith m(NRGA). Considerin g that solu tion-qualit of all meta-heuristic algorithm heavily depends on t eir parameters, Taguchi method is used to fine tune parameters of the em loyed algo rithms. The computati onal results, obtained by implementing the a lgorithms on several problems of different sizes, demonstrate the reliability of the proposed methodolog y. It effici ently minim izes establishment and staffing c sts, as we l as travel and waitin g time for the servic , something which is di rectly related to the ultimate goal of manageria l strategies for maximum preventive he althcare participation achievement.

Keywords:

Multi-objective Preve ntive healthcare problems (MOPHPs) , Queuing s stem , Multi-objective simulated annealing ( MOSA) , NSGA-II , NRGA , Taguchi method