Multiple criteria decision-making for hospital location-allocation based on improved genetic algorithm

Hospital site selection is an essential problem in all societies, and the allocation of the population to these centers are included as the important optimization issues to be considered in urban planning. The present paper aims at improving the genetic algorithm by using the effective and affectability rates for the combination of chromosomes. For evaluating the performance of the new algorithm, it was attempted to use the comparison of improved genetic algorithm (IGA) with GA and particle swarm optimization (PSO). For limiting the searching space, it was attempted to use the analysis capabilities of geospatial information systems (GIS) as well as the analytic hierarchy process (AHP) for selecting the candidate sites. Then, the abovementioned algorithms were implemented to determine six optimized sites and allocating the peer blocks in real data. The findings of this paper indicate that selecting proper combinations of chromosomes with high fitness function and chromosomes with low fitness function will result in an increased algorithm's exploitation and exploration ability. As a result, the algorithm will not be involved with local minimums, the convergence process of the algorithm will improve, and the algorithm will indicate a high level of stability in different performances. Given the findings of this paper, IGA has a better performance in comparison to the other algorithms. The convergence velocity of IGA is higher than that of the GA and PSO. All algorithms showed a high level of repeatability. However, in comparison to the other algorithms, IGA has a higher level of stability. Moreover, the run time of IGA is much shorter than the other algorithms.