Improved simplex method and its application

To improve the performance and efficiency of parameter identification in complicated environmental models,an improved simplex method(IMSM) is presented in this paper.The IMSM integrates a random global search algorithm and a simplex local search algorithm,and greatly enhances the capability of wide and deep searches.For instance,in the parameter identification for the time-space coupled predicting model,which predicts surface subsidence due to water pumping based on stochastic medium theory,the IMSM is applied to identify four parameters for the model.The results show that the global optimized solutions for these parameters can be reliably and effectively obtained under either the condition of non-disturbed data or the condition of disturbed data.This demonstrates that the IMSM is suitable for the parameter identification in complicated environmental models.Furthermore,the comparison among different optimized methods also shows that the IMSM has advantages in performance and efficiency over other methods.