Analysis of source contributions to the ambient aerosol sample by simulated annealing

Simulated annealing (SA) is an optimization technique for locating the global optimum in a multiparameter problem. In this paper, SA is explored as a receptor modelling method for solving the air pollution source apportionment problem by considering the problem as an optimization process in a multi-dimensional space. The simulated aerosol composition data generated by the then National Bureau of Standards (NBS) for the EPA workshop on mathematical and empirical receptor modeling held at Quail Roost, NC in 1982 have been examined. Based on the concept of mixing fractions of individual sources contributing to the ambient aerosol sample, the search space for the SA algorithm is constrained. Satisfactory results have been obtained for the NBS data sets I and II.