Nonparametric grade evaluation method for water resource system

The grade evaluation of water resource system is a hotspot and difficulty problem in water resource system engineering. The common evaluation methods require that the structure form of evaluation function should be presumed beforehand, which are very difficulty to be adapted to the complex and variable practical evaluation systems. For this reason, a new non-parametric evaluation method for grade evaluation of water resource system (NGEM) is presented. The NGEM does not require the structure form of the evaluation function, and the evaluation information can be got directly from the evaluation standard datas, the objective weights of evaluation indexes can be determined by using the accelerating genetic algorithm based projection pursuit, and the evaluation function can be established by using accelerating genetic algorithm based nonparametric regressive method. The calculation examples show that the NGEM is visual, simple, effective and general, which can also be applied to both prediction and simulation of different water resource systems.