Significance analysis of influencing factors of highway freight transportation in China and multi-variable grey prediction for its development

Abstract

With the prosperity of national economy and the development of highway construction, highway freight transportation plays an increasingly important role in the market economy. Due to its great flexible characteristic, highway freight transportation has been the main mode of transportation in China. On the macro level, there are many factors affecting the development of highway freight transportation especially under the background of the new era. Based on the historical data of the development of highway freight transportation, grey entropy analysis method is applied to analyze the significance of influencing factors for the development of highway freight transportation whose key indicator is highway freight turnover. Then GM (1, N) model is established to predict the development trend of highway freight turnover and its significant influencing factors. Finally, main problems existing in highway freight transportation and development prospect were discussed and analyzed. The research results show that the three most significant factors affecting the development of road freight turnover in China are the total state revenue, GDP and average distance of highway freight. The established GM (1, N) model can conduct high precision prediction for the development of highway freight transportation. Opportunities and challenges of highway freight transportation industry coexist and its development prospect is promising. It is expected to provide beneficial references for the development strategy and decision-making of highway freight transportation in China.

Keywords

Highway freight transportation significance analysis grey entropy analysis method GM (1 N) prediction model

1 Introduction

Nowadays, China has embarked on a new journey of development. The pace of economic restructuring is accelerating and the logistics industry (especially the highway freight field) is undergoing great changes [1 –4]. Compared with other modes of transportation, highway freight transportation is more flexible and can realize door-to-door service, so it becomes the main mode of freight transportation. In fact, more than half of the country’s freight transportation is carried by highways, which can carry tens of billions of tons of freight each year, more than 10 million freight vehicles and 30 million truck drivers. Meanwhile, China’s logistics, a basic and strategic industry, has entered the deepening stage of supply-side structural reform under the new normal of the economy, which also imposes on the highway freight industry new opportunities and challenges [5 –7].

In general, highway freight turnover is one of the most important indexes to measure the development of a country’s highway freight. It is of great importance to the decision-making departments at all levels of highway freight transportation to formulate the development strategy and development plan of freight transportation. Therefore, a systematic analysis of the factors affecting the development of highway freight transportation is of great significance to the formulation of economic development policy of highway freight transport in China. The relationship between the highway cargo transportation system and other social and economic systems is complicated, affecting and restricting each other. The randomness and uncertainty of the interaction between them are very obvious, with typical grey characteristics. In view of the obvious timeliness of the development of highway freight transportation, the statistical data of recent years should be adopted to investigate the main impact factors of highway freight transportation, as a small sample problem and also a typical “grey system”. In order to better research the information with partly clear, partly unclear or uncertainty phenomenon, grey system theory came into existence and now plays an increasingly widespread application role.

At present, aiming at the development trend of highway freight transportation, it has the important practical significance to discuss the development prospects of the highway transportation industry under specific conditions [8 –12]. Forecasting methods can be roughly divided into two groups. One is qualitative prediction, which is based on the historical data and visual materials, personal experience and analysis judgment ability applied to analyze the future development of things in the nature and predict the direction along with the major turning point. Such as the expert meeting method, Delphi method and so on, the qualitative prediction is highly subjective, and the prediction result completely depends on the experience and ability of the participants. The other is quantitative prediction including neural network, fuzzy mathematics, probability statistics, regression analysis and grey prediction, etc. Among them, grey prediction is characterized by less original data, simple calculation and no need for accurate model of the controlled object, which can be used as an applicable method in many fields. Based on the grey entropy analysis of the influencing factors of the development of highway freight transportation, this paper uses the grey system theory to establish the GM (1, N) model of highway freight turnover and its significant influencing factors, compares the predicted data with the historical statistical data, and analyzes the error results so as to provide a basis and support for the relevant departments to make the decision of highway freight industry.

2 Methodology

2.1 Grey entropy method

Grey system theory is a new field of control theory. It takes the grey system as the research object and the whitening, desalination, quantification, modeling and optimization of the grey system as the core, and takes the prediction and control of the development of various grey systems as the purpose. The grey entropy method originates from the grey system theory proposed by Prof. Deng Julong in 1982 [13]. It is a new method to study the uncertainty of small sample and poor information, which has been widely used in many fields such as agriculture, industry, meteorology, etc. Through certain data processing, it can find the correlation among random factor sequences and extract the main factors and characteristics that affect the system in order to make positive and effective guidance for the development [14]. Therefore, this paper applies this method to analyze the significance of each influencing factor to the development of highway freight transportation in China. The analysis process is as follows:

• Grey correlation coefficient : X is the grey correlation factor set. x₀(x₀ (1), x₀ (2),...,x₀ (n)) is the reference column and x_i (x_i (1), x_i (2),...,x_i (n)) (i = 1, 2,..., m) is the comparison column, then the grey correlation coefficient can be determined as follows:

$\begin{matrix} ξ_{i} [x_{0} (k) x_{i} (k)] = \\ | \frac{min_{i = 1, m} \begin{matrix} \min_{k = 1, n} Δ_{i} (k) + ρ max_{i = 1, m} \begin{matrix} \max_{k = 1, n} Δ_{i} (k) \end{matrix} \end{matrix}}{Δ_{i} (k) + ρ max_{i = 1, m} \begin{matrix} \max_{k = 1, n} Δ_{i} (k) \end{matrix}} | \end{matrix}$ (1)

Where $min_{i = 1, m}$ $\min_{k = 1, n} Δ_{i} (k)$ is the minimum difference between the two poles and $\max_{i = 1, m}$ $\max_{k = 1, n} Δ_{i} (k)$ is the maximum difference between the two poles. ρ is the resolution coefficient, which is generally set as 0.5.

• Grey relational entropy and grey entropy correlation degree : X is a discrete sequence and x_i ∈ x (i = 1,2,...,m) is a comparison column. $P_{h} \overset{Δ}{=} \frac{ξ [x_{0} (h) x_{i} (h)]}{\sum_{k = 1}^{n} ξ [x_{0} (h) x_{i} (h)]} \begin{matrix} P_{h} \in \end{matrix} P_{i} (h = 1, 2, . . ., n)$ (2)

Where P_h is called as the density value of the distribution.

The grey relational entropy of x_i is expressed as: $H (R_{i}) \overset{Δ}{=} - \sum_{k = 1}^{n} P_{h} ln P_{h} (3)$

The grey entropy correlation of the sequence x_i is expressed as: $E (x_{i}) \overset{Δ}{=} H (R_{i}) / H_{max} (4)$

Where E(x_i) represents the maximum value of the difference information column composed of n elements.

• Analysis of key influencing factors : The rank result of grey entropy correlation degree shows the significance of influencing factors and the key influencing factors can be gained.

2.2 GM (1,N) Grey Prediction Model

Grey prediction model is a prediction method which establishes mathematical model and makes prediction through a small amount of incomplete information. There are many grey prediction models including GM(1,1), GM(1,N) and some improved models applied in the forecast of industrial production, electricity consumption and so on [15 –17]. The classical GM(1,N) model roots in the grey theory and can be applied widely in the research of prediction, decision-making and control [18 –20]. The analysis process is as follows:

Pairs of observations { $x_{1}^{(0)}$ (1), $x_{1}^{(0)}$ (2), ... , $x_{1}^{(0)}$ (n)} from a certain dynamic system with n-1 inputs { $X_{2}^{(0)}$ , $X_{3}^{(0)}$ ... , $X_{n}^{(0)}$ } and an output $X_{1}^{(0)}$ are supposed to be available at equispaced interval of time. Consider the original series, it can be obtained that $X_{1}^{(0)} = {x_{1}^{(0)} (1), x_{1}^{(0)} (2), \dots, x_{1}^{(0)} (n)}$ (3)

and $X_{2}^{(0)} = {x_{2}^{(0)} (1), x_{2}^{(0)} (2), \dots, x_{2}^{(0)} (n)}$ (4) ... ... . $X_{N}^{(0)} = {x_{N}^{(0)} (1), x_{N}^{(0)} (2), \dots, x_{N}^{(0)} (n)}$ (5)

$X_{i}^{(1)}$ are the 1-AGO series of $X_{i}^{(0)}$ ,i = 1, 2, . . . , N, respectively. In addition, $Z_{1}^{(1)}$ is the adjacent generated serial of $X_{1}^{(1)}$ , and then $x_{1}^{(0)} (k) + {az}_{1}^{(1)} (k) = \sum_{i = 2}^{N} b_{i} x_{i}^{(1)} (k)$ (6)

where

1 stands for first-order derivative of 1-AGO series of $X_{1}^{(0)}$ , and n stands for there being n-1 relative series of the system.

Define $\hat{α} = {[\begin{matrix} a & b_{2} & \cdot \cdot \cdot & b_{N} \end{matrix}]}^{T}$ as the parameter series of grey differential equation of GM(1,N) model and then according to the least square method(LSM), it can be obtained that $\hat{α} = (B^{T} B)^{- 1} B^{T} Y$ (7)

Where $B = [\begin{matrix} - z_{1}^{(1)} (2) & x_{2}^{(1)} (2) \cdot \cdot \cdot x_{N}^{(1)} (2) \\ - z_{1}^{(1)} (3) & x_{2}^{(1)} (3) \cdot \cdot \cdot x_{N}^{(1)} (3) \\ \cdot \cdot \cdot & \cdot \cdot \cdot \\ - z_{1}^{(1)} (n) & x_{2}^{(1)} (n) \cdot \cdot \cdot x_{N}^{(1)} (n) \end{matrix}]$ (8)

and $Y = {[\begin{matrix} x_{1}^{(0)} (2) & x_{1}^{(0)} (3) & \cdot \cdot \cdot & x_{1}^{(0)} (n) \end{matrix}]}^{T}$ (9)

Then $\frac{{dx}_{1}^{(1)}}{dt} + {ax}_{1}^{(1)} = b_{2} x_{2}^{(1)} + b_{3} x_{3}^{(1)} + . . . + b_{N} x_{N}^{(1)}$ (10) is the albinism differential equation or silhouette equation of Equation (4). From Equation (8), the modeling value can be derived as $\begin{matrix} x_{1}^{(1)} (t) = e^{- at} [\sum_{i = 2}^{N} \int b_{i} x_{i}^{(1)} (t) e^{at} dt + x_{1}^{(1)} (0) - \sum_{i = 2}^{N} \int b_{i} x_{i}^{(1)} (0) dt] \\ = e^{- at} [x_{1}^{(1)} (0) - t \sum_{t = 2}^{N} b_{i} x_{i}^{(1)} (0) + \sum_{i = 2}^{N} \int b_{i} x_{i}^{(1)} (t) e^{at} dt] \end{matrix}$ (11)

When the rangeability of $X_{i}^{(1)} (i = 1, 2, . . ., N)$ is small, the value of $\sum_{i = 2}^{N} b_{i} x_{i}^{(1)} (k)$ can be deemed as a grey constant. Then

$\begin{matrix} {\hat{x}}_{1}^{(1)} (k + 1) = [x_{1}^{(1)} (0) - \frac{1}{a} \sum_{i = 2}^{N} b_{i} x_{i}^{(1)} (k + 1)] \\ e^{- ak} + \frac{1}{a} \sum_{i = 2}^{N} b_{i} x_{i}^{(1)} (k + 1) \end{matrix}$ (12)

Where $x_{1}^{(1)} (0)$ is equal to $x_{1}^{(0)} (1)$ . Then ${\hat{x}}_{1}^{(0)} (k + 1) = {\hat{x}}_{1}^{(1)} (k + 1) - {\hat{x}}_{1}^{(1)} (k)$ (13)

According to equation (5)∼(15), the implementation key procedure applied in Matlab software is designed as follows:

% Grey prediction

% X0 is m*n matrix with m as number of variables and n as number of sets of data.

clear all

clc

% Input k to be predicted time and original sequence X0

k = predicted time;

X₀ = [data1; data2; ... ;data N];

% X0 means adding up the original sequence X0 to produce the sequence X1;

% X2 = cumsum(X0);

X1 = cumsum(X0’)’;

% —Calculate the data matrix L

[m n] = size(X1);

for i = 1:m

for j = 1:n-1

L(j,i) = (X1(i,j)+X1(i,j+1))/2;

end

L = [L ones(n-1,1)];

% —Calculate Y and parameter estimates

Y = X0(:,2:n)’;

D = inv(L’*L)*L’*Y;

D = D’;

A = D(1:end,1:end-1);

B = D(1:end,end);

% —Calculate the fitting value or prediction value of the model

S = X1(:,1);

if k = =1

Z = S;

Elseif k > 1 Z = expm(A*(k-1))*S+inv(A)* (expm(A*(k-1))-eye(size(expm(A*(k-1)))))*B-(expm(A*(k-2))*S+inv(A)*(expm(A*(k-2))-eye(size(expm(A*(k-2)))))*B)

else

disp(’ Input wrong! k≥1′)

end

3 Grey entropy analysis of factors affecting the development of highway freight transportation in china

X₀, X₁, X₂, X₃, X₄, X₅, X₆, X₇, X₈, X₉ stand for highway freight turnover (0.1billion ton-km), transportation, warehousing and postal services ($100 million), length of highway (10,000 km), average distance of highway freight (km), consumer price index, GDP ($100 million), Population (ten thousand), average salary of employees (yuan), total investment in fixed assets (100 million yuan), total state revenue (100 million yuan), respectively. The original data can be seen in Table 1.

Table 1
Original data sequence of China’s highway freight turnover^*

Year X₀ X₁ X₂ X₃ X₄ X₅ X₆ X₇ X₈ X₉

2001 6330.4 6871.3 169.8 60.0 100.7 110863.1 127627.0 10870.0 37213.5 16386.0

2002 6782.5 7494.3 176.5 61.0 99.2 121717.4 128453.0 12422.0 43499.9 18903.6

2003 7099.5 7914.8 181.0 61.0 101.2 137422.8 129227.0 14040.0 55566.6 21715.3

2004 7840.9 9306.5 187.1 63.0 103.9 161840.2 129988.0 16024.0 70477.4 26396.5

2005 8693.2 10668.8 334.5 65.0 101.8 187318.9 130756.0 18364.0 88773.6 31649.3

2006 9754.2 12186.3 345.7 67.0 101.5 219438.5 131448.0 21001.0 109998.2 38760.2

2007 11354.7 14605.1 358.4 69.0 104.8 270092.3 132129.0 24932.0 137323.9 51321.8

2008 32868.2 16367.6 373.0 171.0 105.9 319244.6 132802.0 29229.0 172828.4 61330.4

2009 37188.8 16522.4 386.1 175.0 99.3 348517.7 133450.0 32736.0 224598.8 68518.3

2010 43389.7 18783.6 400.8 177.0 103.3 412119.3 134091.0 37147.0 251683.8 83101.5

2011 51374.7 21842.0 410.6 182.0 105.4 487940.2 134735.0 41799.0 311485.1 103874.4

2012 59534.9 23763.2 423.8 187.0 102.6 538580.0 135404.0 46769.0 374694.7 117253.5

2013 55738.1 26042.7 435.6 181.0 102.6 592963.2 136072.0 51483.0 446294.0 129209.6

2014 56846.9 28500.9 446.4 183.0 102.0 641280.6 136782.0 56360.0 512021.0 140370.0

2015 57955.7 30487.8 457.7 184.0 101.4 685992.9 137462.0 62029.0 562000.0 152269.2

2016 61080.1 33058.8 469.6 183.0 102.0 740060.8 138271.0 67569.0 606466.0 159605.0

2017 66771.5 37172.6 477.4 181.0 101.6 820754.3 139008.0 74318.0 641238.0 172592.8

2018 71249.2 40550.2 484.7 180.0 102.1 900309.5 139538.0 82413.0 645675.0 183359.8

Year	X₀	X₁	X₂	X₃	X₄	X₅	X₆	X₇	X₈	X₉
2001	6330.4	6871.3	169.8	60.0	100.7	110863.1	127627.0	10870.0	37213.5	16386.0
2002	6782.5	7494.3	176.5	61.0	99.2	121717.4	128453.0	12422.0	43499.9	18903.6
2003	7099.5	7914.8	181.0	61.0	101.2	137422.8	129227.0	14040.0	55566.6	21715.3
2004	7840.9	9306.5	187.1	63.0	103.9	161840.2	129988.0	16024.0	70477.4	26396.5
2005	8693.2	10668.8	334.5	65.0	101.8	187318.9	130756.0	18364.0	88773.6	31649.3
2006	9754.2	12186.3	345.7	67.0	101.5	219438.5	131448.0	21001.0	109998.2	38760.2
2007	11354.7	14605.1	358.4	69.0	104.8	270092.3	132129.0	24932.0	137323.9	51321.8
2008	32868.2	16367.6	373.0	171.0	105.9	319244.6	132802.0	29229.0	172828.4	61330.4
2009	37188.8	16522.4	386.1	175.0	99.3	348517.7	133450.0	32736.0	224598.8	68518.3
2010	43389.7	18783.6	400.8	177.0	103.3	412119.3	134091.0	37147.0	251683.8	83101.5
2011	51374.7	21842.0	410.6	182.0	105.4	487940.2	134735.0	41799.0	311485.1	103874.4
2012	59534.9	23763.2	423.8	187.0	102.6	538580.0	135404.0	46769.0	374694.7	117253.5
2013	55738.1	26042.7	435.6	181.0	102.6	592963.2	136072.0	51483.0	446294.0	129209.6
2014	56846.9	28500.9	446.4	183.0	102.0	641280.6	136782.0	56360.0	512021.0	140370.0
2015	57955.7	30487.8	457.7	184.0	101.4	685992.9	137462.0	62029.0	562000.0	152269.2
2016	61080.1	33058.8	469.6	183.0	102.0	740060.8	138271.0	67569.0	606466.0	159605.0
2017	66771.5	37172.6	477.4	181.0	101.6	820754.3	139008.0	74318.0	641238.0	172592.8
2018	71249.2	40550.2	484.7	180.0	102.1	900309.5	139538.0	82413.0	645675.0	183359.8

^*All data originates from CHINA STATISTICAL YEARBOOK of National Bureau of Statistics.

According to the formula (1)∼(4), the main calculation processes of Grey entropy analysis are as follows (shown in the Table2∼4 and Fig. 1):

Fig. 1

Calculation results of grey entropy correlation degree–E(x).

From Fig. 1, it can be seen that the significance degree of each influencing factor is ranked from the largest to the smallest is X9, X5, X3, X7, X1, X8, X2, X4 and X6. Namely, the total state revenue, GDP and average distance of highway freight are the three most significant factors affecting the development of highway freight turnover in China, which can be explained as follows.

At present, China’s total state revenue has grown steadily. On one hand, under the new normal of economy, China’s industrial structure adjustment is constantly going on, light industry and the third industry are constantly developing. The optimization and upgrading of industrial structure has also led to the adjustment of the structure of highway freight transportation, and the category, quality and volume of highway transportation have changed. The survey shows that the proportion of the transportation of industrial manufactured goods, consumer goods and high value-added goods in the total volume of highway freight transportation is increasing, and the logistics demand is increasing year by year, which has become an important supporting force for the development of road freight transportation industry. On the other hand, with the support of the national finance, the structural adjustment of the highway freight transportation industry has been continuously upgraded. Through the strong alliance and resource sharing, the resources have been further integrated, the development depth and comprehensive utilization level of resources have been strengthened. The cross industry and regional freight information exchange has been continuously strengthened, and the spatial scheduling of freight has been realized by using modern information technology, which has greatly improved the efficiency of highway freight transportation. In addition, the overall domestic purchasing power has been improved, and more and more people in China tend to purchase goods that are not available locally or have obvious price advantages through the network, which also brings great benefits to the development of road transportation.

Meanwhile, the gross national product (GDP) is often regarded as the best indicator of a country’s economic condition and can reflect the development level of productivity, and profoundly affect the development of highway freight. The highway freight transportation is a barometer reflecting the national economic operation situation, which is widely concerned by the society. In the new economic situation, the increase of freight volume and the change of transportation structure, especially the increase of the proportion of highway freight volume, will further promote the rapid economic growth. That is to say, the coupling and coordination of highway freight turnover and GDP is high, and the vertical development trend is good [21].

The average distance of highway freight refers to the average kilometers per ton of goods transported by trucks in a certain period of time. The basic factors that affect the average distance are the total freight turnover and the total freight index. By the end of 2018, the total mileage of China’s highways has reached 4.8465 million kilometers, and the expressway has reached 142600 kilometers, ranking the first in the world. The pace of infrastructure construction in domestic provinces and cities is gradually accelerated, and the road construction market has great potential in the future. The road transportation network extending in all directions supports the busy logistics from south to north in China. In addition, according to “Implementation Plan for Deepening the Reform of Toll Road System and Canceling Toll Stations Within Provincial Boundaries” issued by the general office of the State Council, 487 provincial toll stations in 29 interconnected provinces have been cancelled since 0 : 00 on January 1, 2020, greatly improving the traffic efficiency of the whole road network, reducing energy consumption, emissions, etc., decreasing the average distance of highway freight transportation, and realizing the total volume and turnover volume of highway freight to rise continuously, which promotes the rapid development of highway freight transportation.

4 GM (1, N) model for the development of highway freight transportation in china

Based on the above results of grey entropy analysis, this paper intends to establish GM (1, N) grey prediction model according to the formula (5)∼(15) and the designed procedure, and analyze the error results of the model. Considering the timeliness, the data from 2010 to 2018 are selected for prediction. The prediction results and relative errors are shown in Table 5.

Table 2
Sequence averaging based on Table 1

Year X₀ X₁ X₂ X₃ X₄ X₅ X₆ X₇ X₈ X₉

2001 0.1748 0.3415 0.4689 0.4444 0.9844 0.2593 0.9543 0.2797 0.1266 0.1871

2002 0.1873 0.3725 0.4874 0.4519 0.9697 0.2847 0.9605 0.3196 0.1480 0.2158

2003 0.1960 0.3934 0.4998 0.4519 0.9893 0.3214 0.9663 0.3613 0.1890 0.2479

2004 0.2165 0.4626 0.5166 0.4667 1.0157 0.3785 0.9720 0.4123 0.2397 0.3014

2005 0.2401 0.5303 0.9237 0.4815 0.9952 0.4381 0.9777 0.4726 0.3020 0.3613

2006 0.2693 0.6057 0.9546 0.4963 0.9922 0.5132 0.9829 0.5404 0.3742 0.4425

2007 0.3135 0.7259 0.9896 0.5111 1.0245 0.6317 0.9880 0.6416 0.4671 0.5859

2008 0.9076 0.8135 1.0300 1.2667 1.0352 0.7466 0.9930 0.7521 0.5879 0.7002

2009 1.0269 0.8212 1.0661 1.2963 0.9707 0.8151 0.9979 0.8424 0.7640 0.7823

2010 1.1981 0.9336 1.1067 1.3111 1.0098 0.9638 1.0027 0.9559 0.8561 0.9488

2011 1.4186 1.0856 1.1338 1.3481 1.0304 1.1412 1.0075 1.0756 1.0595 1.1859

2012 1.6440 1.1811 1.1702 1.3852 1.0030 1.2596 1.0125 1.2035 1.2745 1.3387

2013 1.5391 1.2944 1.2028 1.3407 1.0030 1.3868 1.0175 1.3248 1.5181 1.4752

2014 1.5697 1.4166 1.2326 1.3556 0.9971 1.4998 1.0228 1.4503 1.7416 1.6026

2015 1.6004 1.5154 1.2638 1.3630 0.9913 1.6044 1.0279 1.5962 1.9116 1.7384

2016 1.6866 1.6432 1.2967 1.3556 0.9971 1.7308 1.0339 1.7387 2.0629 1.8222

2017 1.8438 1.8477 1.3182 1.3407 0.9932 1.9195 1.0394 1.9124 2.1811 1.9705

2018 1.9674 2.0155 1.3384 1.3333 0.9981 2.1056 1.0434 2.1207 2.1962 2.0934

Year	X₀	X₁	X₂	X₃	X₄	X₅	X₆	X₇	X₈	X₉
2001	0.1748	0.3415	0.4689	0.4444	0.9844	0.2593	0.9543	0.2797	0.1266	0.1871
2002	0.1873	0.3725	0.4874	0.4519	0.9697	0.2847	0.9605	0.3196	0.1480	0.2158
2003	0.1960	0.3934	0.4998	0.4519	0.9893	0.3214	0.9663	0.3613	0.1890	0.2479
2004	0.2165	0.4626	0.5166	0.4667	1.0157	0.3785	0.9720	0.4123	0.2397	0.3014
2005	0.2401	0.5303	0.9237	0.4815	0.9952	0.4381	0.9777	0.4726	0.3020	0.3613
2006	0.2693	0.6057	0.9546	0.4963	0.9922	0.5132	0.9829	0.5404	0.3742	0.4425
2007	0.3135	0.7259	0.9896	0.5111	1.0245	0.6317	0.9880	0.6416	0.4671	0.5859
2008	0.9076	0.8135	1.0300	1.2667	1.0352	0.7466	0.9930	0.7521	0.5879	0.7002
2009	1.0269	0.8212	1.0661	1.2963	0.9707	0.8151	0.9979	0.8424	0.7640	0.7823
2010	1.1981	0.9336	1.1067	1.3111	1.0098	0.9638	1.0027	0.9559	0.8561	0.9488
2011	1.4186	1.0856	1.1338	1.3481	1.0304	1.1412	1.0075	1.0756	1.0595	1.1859
2012	1.6440	1.1811	1.1702	1.3852	1.0030	1.2596	1.0125	1.2035	1.2745	1.3387
2013	1.5391	1.2944	1.2028	1.3407	1.0030	1.3868	1.0175	1.3248	1.5181	1.4752
2014	1.5697	1.4166	1.2326	1.3556	0.9971	1.4998	1.0228	1.4503	1.7416	1.6026
2015	1.6004	1.5154	1.2638	1.3630	0.9913	1.6044	1.0279	1.5962	1.9116	1.7384
2016	1.6866	1.6432	1.2967	1.3556	0.9971	1.7308	1.0339	1.7387	2.0629	1.8222
2017	1.8438	1.8477	1.3182	1.3407	0.9932	1.9195	1.0394	1.9124	2.1811	1.9705
2018	1.9674	2.0155	1.3384	1.3333	0.9981	2.1056	1.0434	2.1207	2.1962	2.0934

Table 3

Calculation of grey correlation coefficient

Year	X₁	X₂	X₃	X₄	X₅	X₆	X₇	X₈	X₉
2001	0.7500	0.6273	0.6476	0.3774	0.8583	0.3864	0.8286	0.9167	0.9831
2002	0.7293	0.6225	0.6520	0.3855	0.8393	0.3884	0.7917	0.9323	0.9519
2003	0.7163	0.6196	0.6597	0.3823	0.8008	0.3893	0.7517	0.9935	0.9105
2004	0.6685	0.6225	0.6648	0.3805	0.7555	0.3939	0.7179	0.9619	0.8578
2005	0.6304	0.4182	0.6728	0.3940	0.7156	0.3997	0.6812	0.8938	0.8062
2006	0.5950	0.4176	0.6865	0.4046	0.6706	0.4077	0.6464	0.8287	0.7426
2007	0.5446	0.4209	0.7161	0.4086	0.6085	0.4215	0.6011	0.7654	0.6453
2008	0.8441	0.8048	0.5790	0.7978	0.7566	0.8569	0.7631	0.6073	0.7059
2009	0.7077	0.9325	0.6479	0.9032	0.7014	0.9509	0.7300	0.6534	0.6698
2010	0.6521	0.8480	0.8174	0.7259	0.6795	0.7182	0.6720	0.5909	0.6655
2011	0.5975	0.6348	0.8800	0.5596	0.6410	0.5453	0.5902	0.5790	0.6810
2012	0.5156	0.5097	0.6571	0.4340	0.5621	0.4377	0.5280	0.5720	0.6184
2013	0.6698	0.5950	0.7152	0.4786	0.7669	0.4855	0.6989	0.9659	0.8904
2014	0.7660	0.5945	0.6990	0.4621	0.8808	0.4735	0.8086	0.7441	0.9440
2015	0.8576	0.5949	0.6766	0.4466	0.9997	0.4621	0.9993	0.6138	0.7845
2016	0.9250	0.5586	0.5989	0.4161	0.9238	0.4295	0.9102	0.5675	0.7877
2017	1.0000	0.4836	0.4946	0.3659	0.8717	0.3790	0.8830	0.5943	0.7991
2018	0.9170	0.4386	0.4367	0.3360	0.7844	0.3468	0.7658	0.6847	0.8000

Table 4

Grey correlation density calculation

Year	X₁	X₂	X₃	X₄	X₅	X₆	X₇	X₈	X₉
2001	0.0573	0.0584	0.0544	0.0436	0.0621	0.0436	0.0620	0.0681	0.0690
2002	0.0557	0.0579	0.0548	0.0445	0.0607	0.0438	0.0592	0.0692	0.0668
2003	0.0547	0.0577	0.0554	0.0441	0.0580	0.0439	0.0562	0.0738	0.0639
2004	0.0511	0.0579	0.0559	0.0439	0.0547	0.0444	0.0537	0.0714	0.0602
2005	0.0482	0.0389	0.0565	0.0455	0.0518	0.0450	0.0510	0.0664	0.0566
2006	0.0455	0.0389	0.0577	0.0467	0.0485	0.0460	0.0484	0.0615	0.0521
2007	0.0416	0.0392	0.0602	0.0472	0.0440	0.0475	0.0450	0.0568	0.0453
2008	0.0645	0.0749	0.0486	0.0921	0.0548	0.0966	0.0571	0.0451	0.0496
2009	0.0541	0.0868	0.0544	0.1043	0.0508	0.1072	0.0546	0.0485	0.0470
2010	0.0498	0.0789	0.0687	0.0838	0.0492	0.0810	0.0503	0.0439	0.0467
2011	0.0457	0.0591	0.0739	0.0646	0.0464	0.0615	0.0442	0.0430	0.0478
2012	0.0394	0.0474s	0.0552	0.0501	0.0407	0.0493	0.0395	0.0425	0.0434
2013	0.0512	0.0554	0.0601	0.0553	0.0555	0.0547	0.0523	0.0717	0.0625
2014	0.0585	0.0553	0.0587	0.0534	0.0638	0.0534	0.0605	0.0553	0.0663
2015	0.0655	0.0554	0.0568	0.0516	0.0724	0.0521	0.0748	0.0456	0.0551
2016	0.0707	0.0520	0.0503	0.0481	0.0669	0.0484	0.0681	0.0421	0.0553
2017	0.0764	0.0450	0.0416	0.0423	0.0631	0.0427	0.0661	0.0441	0.0561
2018	0.0701	0.0408	0.0367	0.0388	0.0568	0.0391	0.0573	0.0509	0.0562

Table 5

Comparison between the calculation results of grey model and the original data

Year	X₀ Highway freight turnover (0.1billion ton-km)			X₃ Average distance of highway freight (km)			X₅ GDP ($100 million)			X₉ Total state revenue (100 million yuan)
	Original data	Prediction data	Relative error/%	Original data	Prediction data	Relative error/%	Original data	Prediction data	Relative error/%	Original data	Prediction data	Relative error/%
2010	43389.7	43389.7	0.00	177.0	177.0	0.00	412119.3	412119.3	0.00	83101.5	83101.5	0.00
2011	51374.7	53460.0	4.06	182.0	180.0	–1.10	487940.2	490220.0	0.47	103874.4	104010.0	0.13
2012	59534.9	56470.0	–5.15	187.0	180.0	–3.74	538580.0	539440.0	0.16	117253.5	117480.0	0.19
2013	55738.1	56410.0	1.21	181.0	180.0	–0.55	592963.2	589520.0	–0.58	129209.6	129450.0	0.19
2014	56846.9	56970.0	0.22	183.0	180.0	–1.64	641280.6	638490.0	–0.44	140370.0	140440.0	0.05
2015	57955.7	58390.0	0.75	184.0	180.0	–2.17	685992.9	688900.0	0.42	152269.2	150840.0	–0.94
2016	61080.1	60950.0	–0.21	183.0	180.0	–1.64	740060.8	744400.0	0.59	159605.0	161050.0	0.91
2017	66771.5	65120.0	–2.47	181.0	180.0	–0.55	820754.3	810270.0	–1.28	172592.8	171560.0	–0.60
2018	71249.2	70710.0	–0.76	180.0	180.0	–1.10	900309.5	894530.0	–0.64	183359.8	183040.0	–0.17
2019		75326.0			180.0			1000800.0			196500.0
2020		79262.0			181.0			1124400.0			203570.0
2021		84057.0			183.0			1275500.0			221760.0

From Table 5, it can be seen that compared with the data of CHINA STATISTICAL YEARBOOK of National Bureau of Statistics, the predicted results calculated by GM (1, N) have the same relative error less than 5%, meeting the accuracy requirements, so there is no need for residual correction. The research results show that the GM (1, N) can better predict the development trend of highway freight turnover and its three most significant influencing factors. The prediction results from 2019 to 2021 calculated by the model are reasonable in theory and can provide reference for the development strategy of highway freight transportation in China.

5 Main problems existing in highway freight transportation and development prospect

5.1 Main problems

With the progress of information technology and the further improvement of freight service quality requirements, the problems in highway freight transportation are increasingly prominent, which limits the development of road freight industry. As a whole, the problems such as the backward technical level of highway freight transportation and the backward management concept are increasingly prominent, which affect the improvement of its service quality, transportation efficiency and economic benefits.

First, the rising operating costs affect the overall profits of the highway freight transportation industry. With the continuous rise of employee compensation, and highway transportation is a labor-intensive industry, which leads to the increasing operating costs of enterprises. In addition, the fluctuation of fuel price will affect the transportation cost. The frequent fluctuation of the prices of various factors has a certain impact on the operating costs of freight transportation enterprises.

Second, price competition in the industry still exists. There are a lot of small and medium-sized enterprises or even micro enterprises in the highway freight transportation, and the homogenization of products and services is more serious. They usually rely on price factor to attract supply. In the future, enterprises must improve service quality and adjust product differentiation to win more market share.

Third, there is a lack of high-end talents. In terms of enterprise management and operation technology, there is a relatively lack of high-quality talents and compound talents in route design, route planning and other aspects. Due to the rapid update of management concept and technical level and the need for training time of professional talents, the whole logistics industry is still facing the situation of talent shortage.

Fourth, under the background of continuous improvement of regulatory standards, safety laws and policies, the number of road freight accidents in China has been effectively controlled in recent years, but the overall accident risk is still high. According to the statistics, the number of accidents per million kilometers in China was 3.7 in 2019. However, according to the U.S. Department of Transportation, as early as 2014, the number of road freight traffic accidents per million kilometers has dropped to about 0.1. Besides, the mortality rate of truck drivers in China is around 1‰ all the year round. This shows that there is still a long way to go to improve the safety management of China’s road freight industry.

Fifth, the technical innovation ability of the road freight transportation industry in China is insufficient. The utilization rate of Internet technology and information technology is not high, the grasp of market information trends is not strong. Although many road freight enterprises have begun to realize the importance of information technology in transportation services, and also began to introduce related electronic office equipment and software, the problem of low level of science and technology in highway freight transportation in China has not been fundamentally solved.

Sixth, the environmental load of highway traffic and transportation is aggravated and the requirements for green freight transportation will be higher and higher. Enterprises should respond to the national call and actively promote the development of green freight transportation. At the same time, it is hoped that the government will give full consideration to the actual bearing capacity of the freight industry and the rigid demand of the society for the normal operation of freight transportation, and provide policy support and financial guidance on the macro level.

5.2 Development prospect

China’s macro-economic growth drives the sustainable development of highway transportation industry. According to the economic and social development goals set out in the 13th Five Year Plan, by 2020, the income of urban and rural residents will double that of 2010. The macro-economy will continue to maintain stable development, which will effectively ensure the growth of customer business volume dominated by small and medium-sized enterprises and urban and rural residents.

There is a large growth space for highway freight industry in the future. With the improvement of highway construction in China, national highway network layout has been formed, which provides favorable basic conditions for highway freight transportation.

The market demand for product types and service quality will be diversified, and the proportion of business volume will continue to increase in the future. With the continuous adjustment of China’s economic structure in the future, the structure of freight products will also change. It is predicted that the share of bulk goods and primary products in highway freight transportation will decrease in the future, but the transportation demand of all kinds of consumer goods and high-tech products in manufacturing industry will increase. At the same time, the requirements for handling, packaging, monitoring and tracking are increased.

The vehicle-freight matching platform has been born and gradually enters the “Network+”. Highway freight transportation enterprises will pay more attention to the application of information technology and establish a perfect information system, give full play to the role of “big data” in the operation of enterprises, and timely obtain the information of market changes, reasonably plan relevant transportation routes, so as to make correct decisions for enterprises and improve highway freight transportation efficiency.

Highway freight is increasingly integrated into the modern logistics industry. As a new economic operation mode, logistics has become an important service sector of the national economy. In the future, it will further integrate the modern logistics system, realize the information integration of people, vehicles and goods, as well as the integration of downstream storage, loading and unloading markets. In addition, in the face of new development opportunities and challenges, except for doing a good job in various services, the healthy development of the highway freight transportation industry also requires the government to play a regulatory role, and enterprises to strengthen their own management, so as to reduce unfair market competition and unsafe highway transport behavior. That will make the development of the highway freight transportation industry more orderly and efficiently [22, 23].

It’s worth noting that the motorcade/freight enterprises should strengthen the construction of safety culture, further improve the fine management system, and realize the overall cost reduction and efficiency increase including risk reduction. Let the safety management awareness permeate in every link of fine operation, and achieve comprehensive cost reduction and efficiency increase, helping ensure the safe development of road freight transportation.

Have the whole world in view, it is quite obvious that “The Belt and Road Initiative” will promote China’s deepening interconnection with the countries along the route. Connectivity and increased bilateral trade with relevant countries will drive international freight demand increase. At the same time, the realization of the goal of a well-off society in an all-round way should pay more attention to people based on the development model of benefiting the people. The process of urban-rural integration is accelerating and E-commerce will increase the demand of rural scattered transportation. As the large-scale domestic demand is constantly released and the distribution of productivity is further adjusted in China, the demand for freight transportation is increasing. On the whole, the countries along the One Belt And One Road route, rural and central & western regions in China will become new important growth pole for the development of highway freight transportation industry.

6 Conclusions

Based on the grey entropy analysis and GM (1, N) prediction of the factors influencing the development of highway freight transportation, the following conclusions are obtained:

The development of highway freight transportation studied in this paper is a “grey body”, and there are many influencing factors. Through the analysis of grey entropy, it is concluded that the total state revenue, GDP and average distance of highway freight are the three most significant factors that affect the development of highway freight turnover in China.

The established GM (1,N) model can predict the development trend of highway freight turnover and its three most significant influencing factors, and the prediction results are in good agreement with the statistical data.

It is feasible and applicable to introduce the grey system theory into the research of highway freight transportation development. Through the grey entropy analysis and GM (1,N) grey prediction model of the significance of each influencing factor, it can provide the basis for decision-making departments at all levels to formulate freight transportation development strategy and development planning.

This paper discusses the main problems existing in the highway freight transportation of China under the new situation and analyzes its development prospect. The Belt and Road Initiative brings new opportunities for the development of freight transport in China and it is expected to provide a useful reference for the development of highway freight transport industry in future.

Footnotes

Acknowledgments

The author wishes to appreciate the financial support from the Scientific Research Project of Shaanxi Provincial Department of Education (20JK0062).

References

Challenges and advances in sustainable transportation systems : plan, design, build, manage, and maintain: Proceedings of the 10th Asia Pacific Transportation Development Conference: Beijing, China, 2014.

H.Q.

, Lu

, Zhang

and Wang

T.Y.

, Trends in road freight transportation carbon dioxide emissions and policies in China, Energy Policy 57 (2013), 99–106.

Wang

, Wu

Q.Q.

and Xiong

, Analysis and prediction of factors influencing the development of highway freight transportation, Statistics and Decision 20 (2015), 83–85.

Song

M.L.

, Zheng

W.P.

and Wang

Z.Y.

, Environmental efficiency and energy consumption of highway transportation systems in China, Int J Production Economics 181 (2016), 441–449.

Tian

J.H.

, Forecast on energy Consumption and environmental emissions of transportation sector in China, Dalian: Dalian University of Techonology 2008.

Wang

Z.H.

and Lu

M.L.

, An empirical study of direct rebound effect for road freight transport in China, Applied Energy 133 (2014), 274–281.

Chung

, Zhou

G.H.

and Yeung

I.M.H.

, A study of energy efficiency of transport sector in China from 2003 to 2009, Appl Energy 112 (2013), 1066–1077.

de Campos

R.S.

, Simon

A.T.

and de Campos Martins

, Assessing the impacts of road freight transport on sustainability: A case study in the sugar-energy sector, Journal of Cleaner Production 220 (2019), 995–1004.

Bretzke

W.R.

, Sustainable logistics: in search of solutions for a challenging new problem, Logistics Research 3(4) (2011), 179–198.

10.

Cottrill

C.D.

and Derrible

, Leveraging big data for the development of transport sustainability indicators, J Urban Technol 22(1) (2015), 45–64.

11.

Fürst

and Oberhofer

, Greening road freight transport: evidence from an empirical project in Austria, J Clean Prod 33 (2012), 67–73.

12.

Sinay

M.C.F.D.

, Cruz

, Ribeiro

S.K.

and Sinay

, Neurofuzzy technology to evaluate the environmental performance of the cargo transportation mode, Int J Logist Syst Manag 25(1) (2016), 44–60.

13.

Deng

J.L.

, Grey ystem theory. Wuhan: Huazhong University of Science and Technology press, 1990.

14.

G.L.

and Lv

Y.J.

, Method for multiple attribute group decision making based on grey incidence, Computer Engineering and Applications 48(9) (2012), 235–237.

15.

Liu

, Lao

, Wu

W.Z.

, et al., Application of optimized fractional grey model-based variable background value to predict electricity consumption[J], Fractals, 2020.

16.

Liu

, Wu

W.Z.

, Xie

W.L.

, et al., Application of a novel fractional grey prediction model with time power term to predict the electricity consumption of India and China-ScienceDirect[J], Chaos, Solitons & Fractals, 141.

17.

Zeng

, Analysing the high-tech industry with a multivariable grey forecasting model based on fractional order accumulation[J], Kybernetes 48(2), 2018.

18.

Tien

, A research on the grey prediction model GM(1,n), Applied Mathematics and Computation 218 (2012), 4903–4916.

19.

Deng

, Control Problems of Grey Systems, System Control Letters 1982(1), 288–294.

20.

Zhang

and Luo

, Multi-factored grey GM(1, N) model of experimentation data processing and its application, Journal of Machine Design 20(3) (2003), 23–24.

21.

Zhang

Z.J.

and Zuo

Q.L.

, Analysis on the coupling coordination degree of highway freight transport volume and GDP, Statistics and Decision Making 22 (2016), 92–94.

22.

Wang

Y.P.

, Relationship between road freight transport demand and national economy development based on co-integration analysis, J Jilin Univ (Engineering and Technology Edition) 41 (2011), 56–61.

23.

D.C.

, Road freight: industry development characteristics, problems and future opportunities, China Logistics & Purchasing 1 (2020), 25–26.