摘要
GM(1,1)模型是灰色系统理论中的核心,已经得到广泛应用。一种改进GM(1,1)模型无论用于拟合或预测,其结果都明显优于常规GM(1,1)模型。结合遗传算法和最小二乘法获得该模型的待定参数,对改进的GM(1,1)模型给出了一种新的求解方法。将此改进GM(1,1)模型用于交叉口交通量的预测,预测结果较好。将等维递推和自适应的思想引入改进GM(1,1)模型,可进一步提高该模型的预测精度和实用性。
GM(1,1) model is the kernel of Grey system theory, and it is applied widely in practice. An improved GM(1,1) model can achieve much better results than that of conventional one in fitting or prediction. A novel method to solve this model is proposed by Combined Genetic Algorithm and Least Square method to get its indeterminate coefficients. This model is employed to predict traffic flow with better results. Prediction accuracy and practicability can be further improved by use of equal-dimension recurrence and self-adapting in GM(1,1) model.
出处
《公路交通科技》
CAS
CSCD
北大核心
2004年第2期80-83,共4页
Journal of Highway and Transportation Research and Development
关键词
灰色模型
遗传算法
交通量预测
Gray model
Genetic algorithm
Traffic flow prediction