張智杰 沈琴琴
摘 要:文章介紹了GM(1,1)模型幾種常見的初值改進(jìn)方法,并運(yùn)用于短時(shí)交通流的預(yù)測(cè)中。通過誤差分析,我們發(fā)現(xiàn)初值優(yōu)化對(duì)提高GM(1,1)模型擬合與預(yù)測(cè)精度的作用不大,要進(jìn)一步提高其擬合與精度需要依靠其他的優(yōu)化手段或提出更好的初值優(yōu)化方法。
關(guān)鍵詞:GM(1,1)模型;初值改進(jìn);交通流
中圖分類號(hào):N941 文獻(xiàn)標(biāo)志碼:A 文章編號(hào):2095-2945(2019)03-0011-04
Abstract: This paper introduces several common initial value improvement methods of GM(1,1) model, and applies them to short-term traffic flow prediction. Through error analysis, we find that the initial value optimization has little effect on improving the fitting and prediction accuracy of GM(1,1) model. To further improve the fitting and prediction accuracy of GM(1,1) model, we need to rely on other optimization methods or propose a better initial value optimization method.
Keywords: GM(1,1) model; initial value improvement; traffic flow