Non-Equidistant Non-Homogeneous Multivariate Grey Model with Fractional Order Accumulation and Its Application
Special Issues Editor (Nottingham Tent University, United Kingdom (Great Britain))
We consider fitting approximations to non-homogenous series to build non-equidistant multivariate grey model NFMGM (1, n) with fractional order accumulation by fitting data with a homogenous exponential function. There are many approximations to non-homogenous series. Based on the modeling principle of non-equidistant multivariate grey model NFMGM (1, n) with the fractional order accumulation, we propose a non-equidistant non-homogeneous multivariate grey model NNFMGM (1, n) with fractional order accumulation. The parameters of the proposed model are estimated by least square method and the time response function is given. We consider: i) the number of fractional order, ii) the coefficient of the background value, iii) the modified values of response function initial values as design variables, and the iv) minimum average relative error as object function. We then establish the optimal model and write the solution program using Matlab. The resulting model with high precision and adaptability is not only suitable for equidistant modeling, but also for non-equidistant modeling. Examples show that the model is practical and reliable, and has wide potential applications in engineering and related fields.
Journal: International Journal of Simulation: Systems, Science & Technology, IJSSST V17
Published: Jul 14, 2016