基于改进的GM(1,1)-BP组合模型的人口预测研究  

Population Forecast Research Based on Improved GM(1,1)-BP Combination Model

作　　者：胡珍 HU Zhen(School of Science,Hubei University of Technology,Wuhan 430068,China)

机构地区：[1]湖北工业大学理学院,湖北武汉430000

出　　处：《数学的实践与认识》2023年第1期184-191,共8页Mathematics in Practice and Theory

摘　　要：准确地预测人口总量发展趋势,对我国社会稳定发展具有重要意义.通过分析GM(1,1)模型背景值的构造理论,利用Newton插值公式和线性分段函数优化GM(1,1)模型的背景值,得到新的GM(1,1)模型,并结合BP神经网络模型,再利用遗传算法优化GM(1,1)-BP组合模型的权重系数,并将组合模型应用到新疆人口预测中.最后,分别应用不同的模型,以及改进的GM(1,1)-BP组合模型进行计算和平均相对误差对比,结果表明,改进的GM(1,1)-BP组合模型有效地提高了预测精度.It is of great significance to predict the development trend of total population accurately for the stable development of Chinese society.By analyzing the construction theory of GM(1,1)model background value,a new GM(1,1)model was obtained by using Newton interpolation formula and linear piecewise function to optimize the background value of GM(1,1)model,and combined with BP neural network model,using genetic algorithm to optimize the weight coefficient of GM(1,1)-BP combined model.The combined model is applied to Xinjiang population prediction.Finally,different models and the improved GM(1,1)-BP combined model are used to calculate and compare the average relative errors.The results show that the improved GM(1,1)BP combined model can effectively improve the prediction accuracy.

关 键 词：GM(1 1)模型 背景值 组合插值 BP神经网络模型 组合模型 

分 类 号：N941.5[自然科学总论—系统科学] C924.2[社会学—人口学]