Assessment of Groundwater Resources and Environment Carrying Capacity Based on Coupled Model of PSO and Projection Pursuit

It is a multi-objective, multi-criteria, multi-level and high-dimensional comprehensive issue to assess groundwater resources and environment carrying capacity affected many factors. Projection pursuit (PP) technology was adopted to reduce dimension of the above non-linear and high-dimensional problem. Particle swarm optimization (PSO) algorithm was introduced to optimize projection index function, which is a complex nonlinear optimal problem and is difficult to deal with by traditional methods. The coupled model based on PSO and PP was applied to assess groundwater resources and environment carrying capacity. Carrying capacity is strong in 1995, while it is weak both in 2000 and in 2005 in People´s Victory Canal Irrigation District. There is a trend of declining of carrying capacity with economy and society development, which shows that it is a fact that eco-environment of groundwater has become increasingly weaker. Compared with the Compared fuzzy comprehensive evaluation method and catastrophe theory model, the above result is reasonable. In addition, the assessment model based on projection pursuit is driven the sample data directly and doesn´t pre-given weight of each assessment index, so the result is more objective and less subjective than fuzzy comprehensive evaluation model. In comparison with both Genetic Algorithm Toolbox and Pattern Search Algorithm Toolbox in MATLAB, PSO algorithm is quick to converge and obtain the global optimal solution to the projective index function, and is easy to be programmed.