A genetic-algorithm-based chance-constrained model for air quality management

Abstract

Air quality management problem has always been in exist, yet it is attracting more and more attention in the past hundred years since the negative consequence due to air pollution are becoming more and more significant. Different strategies have been implemented trying to tackle such a problem. Various methods and models have been proposed by scientist and engineers along the way. However, due to large uncertainties associated with air pollutants transfer and dispersion process, the problem remains to be very challenging. In this report, a Genetic-Algorithm-Aided Stochastic Optimization (GASO) model was formulated. The model incorporated genetic algorithm and Monte Carlo Simulation into the traditional chance constrained programming framework. Genetic algorithm was used to search for optimums and Monte Carlo Simulation was used to check the accuracy of solutions. A hypothesized case study was introduced to demonstrate the applicability of the proposed model. Parameters were analyzed and the problem was solved in MATLAB by translating the parameters and constraints into MATLAB program codes. The results demonstrated that the GASO model was applicable and effective in solving air quality management problems under uncertainty. Alternatives in term of tradeoffs between system cost and risk level were also made available for decision makers.