Air pollution control with semi-infinite programming

Abstract

Air pollution control problems can be formulated as a semi-infinite programming (SIP) problem and we describe three main approaches. The first consists in optimizing an objective function while the pollution level in a given region is kept bellow a given threshold. In the second approach the maximum pollution level in a given region is computed and in the third an air pollution abatement problem is considered. These formulation allow to obtain the best control parameters and the maxima pollution positions, where the sampling stations should be placed. To illustrate this idea, the (SIP)AMPL modeling language was used to code three academic problems. The SIPAMPL software package includes an interface to connect AMPL to any SIP solver, in particular to the NSIPS solver. Numerical results are shown with the discretization method, implemented in the NSIPS solver and it proved to be efficient in solving the proposed problems.