Demand Forecasting Method of Emergency Materials Based on Metabolic Gray Markov

doi:10.16182/j.issn1004731x.joss.21-0874

Abstract

Abstract:

In order to improve the prediction accuracy of the demand for emergency materials of people affected by the disaster, a forecasting method based on metabolism-gray Markov's is proposed. To realize the dynamic prediction of the number of people affected by the disaster, according to demand forecast ideas, the prediction model of metabolism-gray Markov fused is constructed progressively through gray, Markov and metabolism theories. A flexible demand forecasting model for emergency supplies is built through safety stock theory to complete the balance of supply and demand between people number and the materials demand. The prediction results of different models show that the relative error of the proposed prediction model is 0.002 1% smaller better than other models, and the prediction accuracy is significantly better than that of the gray model, in which the prediction of the number of people affected by disasters and the demand for emergency supplies have a higher fit degree.

Key words: Metabolism-grey Markov model, emergency materials, dynamic demand, flood disaster

CLC Number:

TP391.9

Long Ma, Baodong Qin, Na Lu, Meng Kou. Demand Forecasting Method of Emergency Materials Based on Metabolic Gray Markov[J]. Journal of System Simulation, 2023, 35(2): 229-240.

Figures/Tables 13

Fig. 1

Table 1

Table 2

Fig. 2

Table 3

Table 4

Fig. 3

Table 5

Fig. 4

Fig. 5

Table 6

Parameter value of emergency material demand forecast

变量	值	变量	值
$α 1$ /kg	2 500	$α 5$ /条	0.50
$α 2$ /kg	1.60	$L$ /d	1.00
$α 3$ /支	1.66	$∂$	0.05
$α 4$ /顶	0.25	$Z 1 - ∂$	1.65

Table 6

Fig. 6

Fig. 7

References 16

1	Guo Z, Qi M. Research on the Demand Forecast of Emergency Material Based on Fuzzy Markov Chain[C]//2010 International Conference on E-Product E-Service and E-Entertainment. Henan, China: IEEE, 2010: 1-4.
2	蔡玫, 曹杰. 应急物资需求量的二型模糊集合预测方法[J]. 中国安全科学学报, 2015, 25(9): 165-170.
	Cai Mei, Cao Jie. A Type-2 Fuzzy Set Based Approach to Predicting Emergency Material Demand[J]. China Safety Science Journal, 2015, 25(9): 165-170.
3	郭子雪, 韩瑞, 齐美然. 基于多元模糊回归的应急物资需求预测模型[J]. 河北大学学报(自然科学版), 2017, 37(4): 337-342.
	Guo Zixue, Han Rui, Qi Meiran. Predictive Method of Emergency Supplies Demand Based on Multiple Fuzzy Linear Regression Model[J]. Journal of Hebei University(Natural Science Edition), 2017, 37(4): 337-342.
4	张斌, 陈建国, 吴金生, 等. 台风灾害应急物资需求预测模型[J].清华大学学报(自然科学版), 2012, 52(7): 891-895.
	Zhang Bin, Chen Jianguo, Wu Jinsheng, et al. Emergency Materials Demand Forecast Model for Typhoon Disaster Response[J]. Journal of Tsinghua University(Science and Technology), 2012, 52(7): 891-895.
5	曾波, 孟伟, 刘思峰, 等. 面向灾害应急物资需求的灰色异构数据预测建模方法[J]. 中国管理科学, 2015, 23(8): 84-91.
	Zeng Bo, Meng Wei, Liu Sifeng, et al. Prediction Modeling Method of Grey Isomerism Data for Calamity Emergency Material Demand[J]. Chinese Journal of Management Science, 2015, 23(8): 84-91.
6	赵一兵, 高虹霓, 冯少博. 基于支持向量机回归的应急物资需求预测[J]. 计算机仿真, 2013, 30(8):408-412.
	Zhao Yibing, Gao Hongni, Feng Shaobo. Emergency Materials Demand Prediction Based on Support Vector Machine Regression[J].Computer imulation, 2013, 30(8): 408-412.
7	胡忠君, 刘艳秋, 李佳. 基于改进GM(1,1)的洪涝灾害应急物资动态需求预测[J]. 系统仿真学报, 2019, 31(4): 702-709.
	Hu Zhongjun, Liu Yanqiu, Li Jia. Dynamic Demand Forecast of Emergency Materials for Flood Disasters Based on Improved GM(1,1) Model[J]. Journal of System Simulation, 2019, 31(4): 702-709.
8	Li B, Zhu Y, Zou Y. The Optimal Allocation of Emergency Materials for Multiple Points on Water Based on Genetic Algorithm[J]. Journal of Physics Conference Series(S1617-012075), 2020, 1617: 012075.
9	王正新, 刘思峰. 基于Fourier-GM(1,1)模型的灾害应急物资需求量预测[J]. 系统工程, 2013, 31(8): 60-64.
	Wang Zhengxin, Liu Sifeng. Forecasting Demand of the Disaster Emergency Supplies Based on Fourier-GM(1,1) Model [J]. Systems Engineering, 2013, 31(8): 60-64.
10	Chen Fujiang, Chen Junying, Liu Jingang. Forecast of Flood Disaster Emergency Material Demand Based on IACO-BP Algorithm[J]. Neural Computing and Applications(S0941-0643), 2022, 34(5): 3537-3549.
11	刘建华, 张正, 吴洁明. 基于BP神经网络的城市水灾灾情预测模型[J]. 计算机工程与设计, 2005, 26(3): 699-701.
	Liu Jianhua, Zhang Zheng, Wu Jieming. City Flood Forecast Model Based on BP Network[J]. Computer Engineering and Design, 2005, 26(3): 699-701.
12	Mohammadi R, Ghomi S M T, Zeinali F. A New Hybrid Evolutionary Based REF Network Method for Forecasting Time Series: A Case Study of Forecasting Emergency Supply Demand Time Series[J]. Engineering Application Soft Artificial Intelligence(S0952-1976), 2014, 36: 204-214.
13	詹沙磊, 傅培华, 李修琳, 等. 基于马尔科夫决策的应急物资动态分配模型[J]. 控制与决策, 2018, 33(7): 1312-1318.
	Zhan Shalei, Fu Peihua, Li Xiulin, et al. Dynamic Programming Approach for Relief Goods Allocation Based on Markov Decision[J]. Control and Decision, 2018, 33(7): 1312-1318.
14	邓聚龙. 灰色系统基本方法[M]. 武汉: 华中科技大学出版社, 2005: 56-68.
	Deng Julong. Grey System Basic Method[M]. Wuhan: Huazhong University of Science and Technology Press, 2005: 56-68.
15	滕婕, 夏志杰, 罗梦莹. 基于新陈代谢GM(1, N)马尔科夫模型的动态网络舆情危机预测[J]. 情报科学, 2020, 38(8): 88-94.
	Teng Jie, Xia Zhijie, Luo Mengying. Dynamic Network Public Opinion Crisis Prediction Based on Metabolic GM(1, N) Markov Model[J].Information Science, 2020, 38(8): 88-94.
16	付中华, 郑榕娇. 时滞多变量GM(1, N)协调度模型及其应用[J]. 统计与决策, 2018, 34(13): 77-80.
	Fu Zhonghua, Zheng Rongjiao. Time Delay to Multivariable GM(1, N) Coordination Model and Application[J]. Statistics and Decision, 2018, 34(13): 77-80.

第几天	受灾人数	第几天	受灾人数
1	4.66	7	58.73
2	7.90	8	66.48
3	22.25	9	112.23
4	24.33	10	122.55
5	36.46	11	157.78
6	57.33	12	185.80

第几天	实际值/万人	基本GM(1， N) 模型		GM(1， N)残差修正
第几天	实际值/万人	预测值 /万人	相对误差/%	预测值 /万人	相对误差/%
平均相对误差			0.031 8		0.42
1	4.66	4.66	─	5.08	─
2	7.90	8.23	0.000 6	7.90	-14.00
3	22.25	22.73	0.001 3	25.94	4.82
4	24.33	28.91	0.001 9	32.65	-0.90
5	36.46	36.53	0.002 0	41.09	-0.07
6	57.33	52.21	0.089 2	51.72	5.12
7	58.73	61.85	0.053 1	65.09	-3.12
8	66.48	62.68	0.057 1	81.92	3.80
9	112.23	100.69	0.102 8	103.11	11.54
10	122.55	126.63	0.033 3	129.77	-4.08
11	157.78	157.40	0.002 4	163.33	0.38
12	185.80	184.64	0.006 2	205.57	1.16

第几天	相对误差/%	状态分布
1	─	─
2	-14	S₁
3	4.82	S₅
4	-0.9	S₁
5	-0.07	S₁
6	5.12	S₅
7	-3.12	S₁
8	3.80	S₅
9	11.54	S₅
10	-4.08	S₁
11	0.38	S₃
12	1.16	S₅

方法	第几天	实际值/万人	预测值/万人	相对误差/%
灰色模型	13	197.83	211.93	0.007 4
	14	237.16	242.69	0.007 9
	15	251.33	259.17	0.006 8
灰色马尔科夫模型	13	197.83	198.71	0.006 9
	14	237.16	238.71	0.007 0
	15	251.33	259.44	0.009 5
新陈代谢灰色马尔科夫模型	13	197.83	─	─
	14	237.16	224.16	0.008 1
	15	251.33	237.92	0.008 9

第几天	实际值	文献[7]预测值	本文模型预测值
1	4.66	─	─
2	7.90	7.90	7.90
3	22.25	25.94	23.89
4	24.33	32.65	28.17
5	36.46	41.09	40.13
6	57.33	51.72	55.81
7	58.73	65.09	57.26
8	66.48	81.92	74.37
9	112.23	103.11	110.64
10	122.55	129.77	126.05
11	157.78	163.33	160.27
12	185.80	205.57	198.91