A fuzzy-TOPSIS approach to enhance emergency logistics supply chain resilience

Abstract

Unconventional emergencies are difficult to be predicted and controlled. Emergency logistics supply chains (SCs) are exposed to great natural risks in operation and prone to encounter chain breakage. Based on a large number of literature analysis and practical research, this paper proposes 8 strategies and develops a Fuzzy-topsis (Technique for Order Preference by Similarity to an Ideal Solution) approach to enhance emergency logistics SC resilience. An empirical analysis is carried out among several Emergency Logistics Experts. The results show that the 3 strategies of “increasing the number of transport links”, “improving information monitoring and warning capabilities”, “improving the accuracy of the plans” and “establishing a green passage” are most effective to enhance the resilience of the emergency logistics SC. Sensitivity analysis shows that the 8 strategies proposed play different roles in the contribution of SC resilience enhancing and can be applied in different situations.

Keywords

Emergency logistics SCs supply chain risks supply chain resilience fuzzy mathematics TOPSIS

1 Introduction

Emergency logistics SC is a dynamic supply chain alliance, which is set up to ensure the production and supply of emergency materials caused by unconventional emergencies and is commanded by emergency management center. Once an emergency occurs, the emergency logistics SC will start to operate quickly to meet the tremendous demand for materials. Emergency logistics SC is facing serious and uncertain risks. The risks come not only from the instability factors within the supply chain, but also from the interference of external emergencies. Compared with the ordinary SC, emergency logistics SC is more vulnerable, which further increases the difficulty of emergency relief. How to overcome vulnerability is an important issue in the field of emergency logistics SC research, and also a key problem to be solved in emergency relief practice. SC resilience enhancement is considered to be an effective way to overcome vulnerabilities and return to original state. Therefore, it is necessary to study the emergency logistics SCs’ resilience.

In the view of the importance of emergency logistics SCs’ resilience, this paper fills in the research gaps by adopting the Fuzzy-TOPSIS approach to prioritize resilience strategies from a SC perspective. This study reviews the related literature and highlights the research gaps first. Section 3 presents the methodology, while Section 4 proposes 8 resilience strategies builds an evaluation index system, explains the results and provides discussions. Section 5 is the last section which outlines the main conclusions and contributions of the study.

2 Literature review

2.1 Emergency logistics risks management

Sun and Qin [1] propose that emergency logistics are characteristic of suddenness, complexity, destructiveness, constancy and opportunity. Chen [2] described the risks of emergency logistics from the perspective of emergency logistics organization coordination. Kumar et al. [3, 4] systematically elaborated the risk management of emergency SC. Chen et al. [5] constructed a set of reliability evaluation index system for emergency logistics SC and evaluate it. Gong et al. [6] and Li et al. [7] adopted fuzzy mathematics to describe the uncertainties in emergency logistics risk assessment.

Some related research study transportation scheduling and node optimization. Peng et al. [8] study the robust optimization of emergency material supply network, considering of uncertainties in supply and demand. Tofighi [9], Yu [10] comprehensively consider the uncertainty of supply, demand and traffic availability after the disaster, and optimize the emergency transportation network. Jeong et al. [11] divide emergency logistics network into two stages: strategic design and operational planning. This study analyzes the robustness of emergency network by disturbing transportation cost in different damage situation of the main facilities. Chen [12] constructed an emergency logistics network with a single starting-point and a single end-point, and analyze the invulnerability of the emergency logistics network based on complex network theory. Li [13] evaluate the network reliability by calculating the reliability of network nodes.

2.2 SC resilience

Martin and Helen [14] defined SC resilience as the ability of SC system returning into its original state or a new and more ideal state from disturbance. And SC resilience is referring to contain flexibility and agility. To enhance resilience is one of the effective approaches to overcome SC vulnerabilities [15, 16]. SC resilience enables the system to respond quickly and actively to shocks, which makes the organization to take the lead over its competitors in the market [17]. The resilience of the social system can be understood from three aspects [18]: (1) To recover to an ideal state as far as possible. (2) To shorten the recovery time as much as possible. (3) To reduce the possibility of re-failure.

Some research focus on the resilience measurement and enhancement. Paton, Smith [19] applied the risk management framework to explore the impact of vulnerability and resilience on the complex relationship between risk factors and crisis outcomes. Calgaro et al. [20] describe the complex relationship framework of vulnerability and resilience in tourist destinations based on chaotic complexity theory and dynamic method, from the perspective of the sustainable development of tourist destinations. Carvalho and Barroso [21] study the cases from automobile supply chain, take “lead-time ratio” and “total cost” as performance evaluation indicators, and analyze SC resilient enhancement effect of 6 strategies. Jasmine [22] developed a quality function deployment approach to enhance resilience of maritime SC.

2.3 Research gaps

Resilience enhancement, as discussed in the above literature, is an important and effective way of preventing SC risks and accelerating SC’s recovery from destruction. However, there is little literature on measurement and enhancement of emergency logistics SC of resilience. Wu [23] explored the resilience of emergency logistics network, but the resilience measurement and mechanism haven’t been discussed in depth. Ge et al. [24] proposed three resilience enhancing strategies for a three-level emergency distribution network, and build a mixed integer programming model of multi-modal transportation and multi-objective under traffic interruption. Mera [25] proposed an objective function of vulnerability measurement for post-disaster road maintenance, which minimize the vulnerability of damage caused by natural accidents under budget constraints, and thus improve the resilience of the network. They have studied the resilience of the emergency material transportation network from the micro perspective, but they have not considered organization coordination and information transmission in the emergency logistics supply chain. Ge [26] developed the quality function deployment method to enhance the emergency logistics SC resilience. But they haven’t established the resilience evaluation index system of emergency logistics SC. To fill these research gaps, this paper is devoted to construct a set of evaluation index system of logistics SC resilience, and develop a fuzzy-Topsis evaluation method to choose the best solutions.

3 Methodology

TOPSIS is the abbreviation of “technology for order preference by similarity to ideal solution”, which means “approaching ideal solution technology”. Its basic principle is to find out the best solution and the worst solution in limited alternatives and to calculate the relative the distances between each evaluated alternative and the best solution, so as to judge which is better or worse. The information and data involved in emergency logistics is obtained from a variety of sources and consumes high acquisition cost. TOPSIS method has little requirement for data, and it can make full use of information and offer relatively objective evaluation results. These advantages make TOPSIS very suitable for choosing solutions of emergency logistics SC resilience enhancement.

However, traditional TOPSIS method has some shortcomings. During practice, the index value of many alternatives has great uncertainty. Especially when the expert survey method based on group decision-making is adopted, the uncertainty will greatly increase the evaluation errors. In this paper, the fuzzy mathematics method and TOPSIS are combined to solute the quantitative, qualitative or fuzzy uncertain factors that affect the resilience enhancement of emergency logistics SC.

Fuzzy mathematics method is first proposed by Zadeh in 1965, which is usually used to describe and model vague concepts and phenomena existing in the real world. Generally, there are triangle fuzzy numbers and trapezoid fuzzy numbers. This paper adopts triangle fuzzy number, which is recorded as $\tilde{A}$ = (a, b, c). Its membership function can be expressed as:

$u_{\tilde{A}} = {\begin{matrix} \frac{x - a}{b - a} & a \leq x \leq b \\ \frac{c - x}{c - b} & b \leq x \leq c \\ 0 & else \end{matrix}}$ (1)

In addition, for trigonometric fuzzy functions Ã = (a1, b1, c1), $\tilde{B}$ = (a2, b2, c2) and the exact number k, there are the following basic algorithms: $\tilde{A} + \tilde{B} = (a 1 + a 2, b 1 + b 2, c 1 + c 2)$ (2)

$\tilde{A} - \tilde{B} = (a 1 - a 2, b 1 - b 2, c 1 - c 2)$ (3)

$\tilde{A} * \tilde{B} = (a 1 a 2, b 1 b 2, c 1 c 2)$ (4)

$k \tilde{A} = (ka 1, kb 1, kc 1)$ (5)

Fuzzy TOPSIS increases the objective accuracy of qualitative or fuzzy uncertain factors, and is applicable for the evaluation of emergency logistics SC enhancement strategies.

Step 1- Determining the weight value of fuzzy evaluation index

In this paper, the index performance is described as: very unimportant, relatively unimportant, medium, relatively important and very important. The relationship between performance and scores is shown in Table 1.

Table 1

Relationship between index importance and scores

Importance	Scores
Very unimportant	(0.1,0.1,0.3)
Relatively unimportant	(0.1,0.3,0.5)
Medium	(0.3,0.5,0.7)
Relatively important	(0.5,0.7,0.9)
Very important	(0.7,0.9,0.9)

Formulas (2, 5 and 6) are used to calculate the index weight value.

${\tilde{ω}}_{J} = \frac{1}{K} ({\tilde{ω}}_{J}^{1} (+) {\tilde{ω}}_{J}^{2} (+) \dots (+) {\tilde{ω}}_{J}^{K})$ (6)

K represents the number of decision makers, m represents the number of alternatives, and n represents the evaluation index. $\tilde{W_{J}}$ is the weight value of index J from decision maker K, ${\tilde{ω}}_{J} = (ω_{j 1} + ω_{j 2} + ω_{j 3})$ , ∀j∈n.

Step 2- Building a decision matrix

According to the enhancement performance on each evaluation index, the alternatives are rated as low, medium low, medium, medium high and high. All qualitative indicators are converted to fuzzy values, as shown in Table 2.

Table 2

Relationship between index performance and scores

Performance	Scores
Low	(1,1,3)
Medium low	(1,3,5)
Medium	(3,5,7)
Medium high	(5,7,9)
High	(7,9,9)

By formula (5) and formula (7), the fuzzy decision matrix is calculated. ${\tilde{x}}_{IJ} = \frac{1}{K} ({\tilde{x}}_{IJ}^{1} (+) {\tilde{x}}_{IJ}^{2} (+) \dots (+) {\tilde{x}}_{IJ}^{K})$

In which, x_IJ represents the evaluation value of index j from decision maker K, and

${\tilde{x}}_{IJ} = (a_{ij} + b_{ij} + c_{j 3}), \forall i \in m, j \in n$ (7)

Step 3- Decision matrix standardization

The triangular fuzzy number ${\tilde{x}}_{IJ} = (a_{ij} + b_{ij} + c_{j 3})$ , can be divided into “benefit type” and “cost type”. The larger the “benefit type” index value of is, the better the performance is; The smaller the “cost type” index value is, the better the performance is. The standardized number ${\tilde{γ}}_{IJ} = (γ_{ij 1}, γ_{ij 2}, γ_{ij 3})$ can be expressed as:

$\begin{matrix} {\tilde{γ}}_{IJ} = (\frac{a_{ij}}{{c_{j}}^{+}}, \frac{b_{ij}}{{c_{j}}^{+}}, \frac{c_{ij}}{{c_{j}}^{+}}), {c_{j}}^{+} = \max c_{ij} or \\ {\tilde{γ}}_{IJ} = (\frac{{a_{j}}^{-}}{c_{ij}}, \frac{{b_{ij}}^{-}}{b_{ij}}, \frac{{c_{ij}}^{-}}{a_{ij}}) s, {a_{j}}^{-} = \min a_{ij} \end{matrix}$ (8)

Step 4- The weighted standardized matrix calculation

${\tilde{v}}_{IJ} = {\tilde{γ}}_{IJ} \otimes {\tilde{ω}}_{J} = (v_{ij 1}, v_{ij 2}, v_{ij 3}) = (r_{ij 1} \times ω_{j 1}, v_{ij 2} \times ω_{j 2}, v_{ij 3} \times ω_{j 3})$ (9)

Step 5- Determination fuzzy ideal solution and fuzzy negative ideal solution

$\begin{matrix} A^{+} = ({\tilde{v}}_{1}^{+}, {\tilde{v}}_{1}^{+}, \dots, {\tilde{v}}_{n}^{+}) \\ A^{-} = ({\tilde{v}}_{1}^{-}, {\tilde{v}}_{1}^{-}, \dots, {\tilde{v}}_{n}^{-}) \end{matrix}$ (10)

In which,

${\tilde{v}}_{I}^{+} = \max ({\tilde{v}}_{IJ}), {\tilde{v}}_{I}^{-} = \min ({\tilde{v}}_{IJ})$ (11)

Step 6- Calculating the relative distances between each solution and the fuzzy ideal solution, and rank the solutions.

The distances from the solution i to A⁺ and A^– are respectively

${d_{i}}^{+} = \sum_{i = 1}^{n} d ({\tilde{v}}_{IJ}, {\tilde{v}}_{I}^{+}), i = 1, 2, \dots, m$ (12)

${d_{i}}^{-} = \sum_{i = 1}^{n} d ({\tilde{v}}_{IJ}, {\tilde{v}}_{I}^{-}), i = 1, 2, \dots, m$ (13)

In which, d(·, ·) is the distance between the two triangular fuzzy numbers

$C_{i} = \frac{{d_{i}}^{-}}{{d_{i}}^{+} + {d_{i}}^{-}}$ (14)

According to the size of C_i, the alternatives can be ranked.

4 Result analysis and discussion

4.1 Resilience enhancement strategies

Emergency logistics SC resilience enhancement is to establish the necessary response mechanism and help the emergency logistics SC to recover quickly to the normal state or a better state, which can minimize the loss caused by the emergency when the links suffer from interruption caused by the external impact. This paper collects, classifies and arranges a series of strategies and develops a Fuzzy-TOPSIS approach to enhance emergency logistics SC resilience by a large number of literature analysis and investigation visit to experts in emergency management. Based on the all-round and multi-level strategic concept of resilience enhancement ^[17], 8 emergency logistics SC resilience enhancement strategies are proposed based on the principles of redundancy, resilience pre-embedding, SC flexibility and agility [1], as shown in Table 3.

Table 3
Emergency logistics SC resilience enhancement strategies and description

Number Strategies Description Enhancement principles Source

1 S1: Increasing the number of transport links Increasing the number of transportation links from the emergency material storage point to the disaster area, so as to disperse the risk and reduce the road network vulnerability Redundancy, resilience pre-embedding Visit and investigation

2 S2: Increasing the capacity of critical road Reducing the possibility of road congestion and increasing road capacity to improve transportation efficiency Redundancy, Resilience pre-embedding Visit and investigation

3 S3: Establishing green channel Establishing emergency support channels between regions or countries to simplify the delivery cycle of emergency materials Agility Visit and investigation

4 S4: Improving the accuracy of the plan Carrying out real-time monitoring and early warning for disasters and their diffusion, and generating emergency plans in time Flexibility, agility Bottani, Rizzi, [27] Streka, et al. [28]

5 S5: Improving the ability of information monitoring and early warning Advanced sensing, detection equipment and comprehensive database. High visibility can help disaster assessment and emergency decision-making Agility Huang, et al. [29]

6 S6: Strengthening the layout rationality of emergency material reserve centers Including reasonable quantity and reasonable location. Reasonable layout of the reserve centers can shorten the transportation route and improve the transportation efficiency Resilience pre-embedding Visit and investigation

7 S7: Improving the ability to raise emergency material Flexible application of emergency reserve, compulsory requisition, government procurement, social donation and international assistance Agility Visit and investigation

8 S8: Strengthening the standardization of emergency facilities, equipment and processes Standardization can make emergency equipment and process be substitutable, so as to improve the flexibility of emergency logistics SC Flexibilty Visit and investigation

Number	Strategies	Description	Enhancement principles	Source
1	S1: Increasing the number of transport links	Increasing the number of transportation links from the emergency material storage point to the disaster area, so as to disperse the risk and reduce the road network vulnerability	Redundancy, resilience pre-embedding	Visit and investigation
2	S2: Increasing the capacity of critical road	Reducing the possibility of road congestion and increasing road capacity to improve transportation efficiency	Redundancy, Resilience pre-embedding	Visit and investigation
3	S3: Establishing green channel	Establishing emergency support channels between regions or countries to simplify the delivery cycle of emergency materials	Agility	Visit and investigation
4	S4: Improving the accuracy of the plan	Carrying out real-time monitoring and early warning for disasters and their diffusion, and generating emergency plans in time	Flexibility, agility	Bottani, Rizzi, [27] Streka, et al. [28]
5	S5: Improving the ability of information monitoring and early warning	Advanced sensing, detection equipment and comprehensive database. High visibility can help disaster assessment and emergency decision-making	Agility	Huang, et al. [29]
6	S6: Strengthening the layout rationality of emergency material reserve centers	Including reasonable quantity and reasonable location. Reasonable layout of the reserve centers can shorten the transportation route and improve the transportation efficiency	Resilience pre-embedding	Visit and investigation
7	S7: Improving the ability to raise emergency material	Flexible application of emergency reserve, compulsory requisition, government procurement, social donation and international assistance	Agility	Visit and investigation
8	S8: Strengthening the standardization of emergency facilities, equipment and processes	Standardization can make emergency equipment and process be substitutable, so as to improve the flexibility of emergency logistics SC	Flexibilty	Visit and investigation

4.2 Establishment of the evaluation index system

Combined with the previous analysis for emergency logistics SC risk and resilience, this paper establishes a set of index system of emergency logistics SC resilience enhancement from 4 aspects, which are supply network robustness, node members’ resource support ability, decision-making response ability and cooperation ability. Table 4 shows the comprehensive evaluation index system of emergency logistics SC resilience enhancement.

Table 4
The weight of the evaluation index system

Target Primary index Secondary indexes The weight

The performance evaluation for emergency logistics SCs’ Supply network robustness (1) the dispersion degree of emergency resource reserve centers (0.5,0.7,0.85)

resilience enhancement (2) the dispersion degree of distribution point (0.45,0.65,0.85)

(3) the redundancy of transportation capacity (0.25,0.65,0.8)

Node members’ resource support capability (4) manufacturers capacity redundancy (0.1,0.15,0.35)

(5) inventory redundancy of emergency resource reserve centers (0.15,0.3,0.15)

Decision making ability (6) monitoring and evaluation ability (0.2,0.35,0.55)

(7) richness of emergency case resource base (0.1,0.25,0.45)

(8) resource allocation ability (0.1,0.2,0.4)

Cooperation ability (9) command and coordination efficiency (0.2,0.4,0.6)

(10) laws and regulations guarantee mechanism (0,45,0.65,0.8)

Target	Primary index	Secondary indexes	The weight
The performance evaluation for emergency logistics SCs’	Supply network robustness	(1) the dispersion degree of emergency resource reserve centers	(0.5,0.7,0.85)
	resilience enhancement	(2) the dispersion degree of distribution point	(0.45,0.65,0.85)
		(3) the redundancy of transportation capacity	(0.25,0.65,0.8)
	Node members’ resource support capability	(4) manufacturers capacity redundancy	(0.1,0.15,0.35)
		(5) inventory redundancy of emergency resource reserve centers	(0.15,0.3,0.15)
	Decision making ability	(6) monitoring and evaluation ability	(0.2,0.35,0.55)
		(7) richness of emergency case resource base	(0.1,0.25,0.45)
		(8) resource allocation ability	(0.1,0.2,0.4)
	Cooperation ability	(9) command and coordination efficiency	(0.2,0.4,0.6)
		(10) laws and regulations guarantee mechanism	(0,45,0.65,0.8)

(1) supply network robustness

Robustness refers to the ability of the system to maintain other performances when it receives certain disturbances or disturbances. It emphasizes the stability of the system. When the main nodes of the network are scattered and the transportation capacity is redundant, the supply network robustness is considered to be effectively enhanced. The emergency supply network usually takes the form of three-level network: emergency resource reserve centers ⟶ emergency distribution point ⟶ disaster area. Thus the supply network robustness can be divided into three indexes: the dispersion degree of emergency resource reserve centers, the dispersion degree of distribution point and the redundancy of transportation capacity.

(2) Node members’ resource support capability

Emergency logistics SC resilience enhancement requires that the members of the supply chain alliance at all levels, such as the material production enterprises and the emergency material reserve center, should maintain certain resilient. Node members’ resource support capability can be divided into two indexes: Manufacturers’ capacity redundancy and inventory redundancy of emergency resource reserve centers.

(3) decision making ability

In emergency logistics, the decision-making ability of the departments is the most important aspect reflecting its agility. Setting up an expert group with rich experience from the field of emergency decision-making, enhancing the accuracy of emergency plans and improving the advanced nature of emergency rescue database, can all effectively improve the ability of emergency decision-making. The decision-making ability can be divided into three secondary indicators: monitoring and evaluation ability, richness of emergency case resource base and resource allocation ability.

(4) cooperation ability

A good cooperation among supply chain members can optimize the process. Establishing a green rescue mechanism, directly delivering for important materials, and strengthening the sharing level of data and information such as emergency relief demand, material reserves, production plans, rescue action plans, etc., are all help to improve the cooperation. Cooperation is often achieved through the formulation of various mechanisms in non-emergency state. It can be divided into command and coordination efficiency, laws and regulations guarantee mechanism.

4.3 Performance evaluation

We find many experts from the emergency command center and relevant research institutes to score the weight of evaluation index. The weights of the evaluation index system of the alternatives are calculated, as shown in Table 4.

The experts are organized to discuss the resilience enhancement effect of alternatives. The discussion result are transferred and the fuzzy multi-attribute decision matrix can be obtained, as shown in Table 5.

Table 5
The fuzzy multi-attribute decision matrix of resilience enhancement alternatives

Number Alternatives Index of evaluation

Index 1 Index 2 Index 3 Index 4 Index 5 Index 6 Index 7 Index 8 Index 9 Index 10

1 S1 (7,9,9) (7,9,9) (7,9,9) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3)

2 S2 (7,9,9) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3)

3 S3 (1,1,3) (1,1,3) (1,3,5) (1,1,3) (1,1,3) (1,1,3) (5,7,9) (1,3,5) (5,7,9) (7,9,9)

4 S4 (1,1,3) (1,1,3) (1,1,3) (1,3,5) (1,3,5) (7,9,9) (7,9,9) (7,9,9) (7,9,9) (5,7,9)

5 S5 (1,1,3) (1,1,3) (1,3,5) (1,1,3) (1,3,5) (7,9,9) (5,7,9) (5,7,9) (5,7,9) (5,7,9)

6 S6 (3,5,7) (3,5,7) (7,9,9) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3)

7 S7 (1,1,3) (1,1,3) (1,1,3) (7,9,9) (7,9,9) (1,1,3) (1,1,3) (1,1,3) (7,9,9) (5,7,9)

8 S8 (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (1,1,3) (5,7,9) (1,3,5) (5,7,9) (1,3,5)

Number	Alternatives	Index of evaluation
1	S1	(7,9,9)	(7,9,9)	(7,9,9)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)
2	S2	(7,9,9)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)
3	S3	(1,1,3)	(1,1,3)	(1,3,5)	(1,1,3)	(1,1,3)	(1,1,3)	(5,7,9)	(1,3,5)	(5,7,9)	(7,9,9)
4	S4	(1,1,3)	(1,1,3)	(1,1,3)	(1,3,5)	(1,3,5)	(7,9,9)	(7,9,9)	(7,9,9)	(7,9,9)	(5,7,9)
5	S5	(1,1,3)	(1,1,3)	(1,3,5)	(1,1,3)	(1,3,5)	(7,9,9)	(5,7,9)	(5,7,9)	(5,7,9)	(5,7,9)
6	S6	(3,5,7)	(3,5,7)	(7,9,9)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)
7	S7	(1,1,3)	(1,1,3)	(1,1,3)	(7,9,9)	(7,9,9)	(1,1,3)	(1,1,3)	(1,1,3)	(7,9,9)	(5,7,9)
8	S8	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(1,1,3)	(5,7,9)	(1,3,5)	(5,7,9)	(1,3,5)

The alternatives of resilience enhancement are ranked by the calculation steps mentioned above, as shown in Table 6.

Table 6

The relative distances between each alternative and the fuzzy ideal solution

Number	Alternatives	d+	d–	c	Rank
1	S1 Increasing the number of transport links	3.983	2.389	0.375	1
2	S2: Increasing the capacity of critical road	4.665	1.586	0.254	8
3	S3: Establishing green channel	4.326	2.147	0.332	4
4	S4: Improving the accuracy of the plan	4.199	2.34	0.358	3
5	S5: Improving the ability of information monitoring and early warning	4.213	2.361	0.359	2
6	S6: Strengthening the layout rationality of emergency material reserve centers	4.385	1.969	0.31	6
7	S7: Improving the ability to raise emergency material	4.4	1.989	0.311	5
8	S8: Strengthening the standardization of emergency facilities, equipment and processes	4.543	1.909	0.296	7

Through the evaluation of Fuzzy-TOPSIS approach, it is concluded that “increasing the number of transport links” is the best solution of emergency logistics SCs’ resilience enhancement, followed by “improving the ability of information monitoring and early warning” and “improving the accuracy of the plan”. In practice, the priority of the above three strategies is conducive to enhance the resilience of emergency logistics SC.

4.4 Sensitivity analysis

Sensitivity analysis is a quantitative method to describe the importance of input variables to output variables, which is used to study and predict the impact of changes in some indicators on the output value of the model. Sensitivity analysis is applied in many fields such as economy, ecology and engineering. In the evaluation above, the resilience enhancement effect of the 8 alternatives is ranked. In order to verify the resilience enhancement effectiveness of each alternative, we try to disturb the evaluation index. If the ranking results of alternatives change with the disturbance of the index, it shows that the eight alternatives proposed in this paper have certain resilience enhancement effect.

We conduct 15 experiments, as shown in Table 7. In experiments 1–10, the weight of one certain index is set as high/important, and thus the fuzzy number is (0.7,0.9,0.9); the other indexes are low/unimportant, and thus the fuzzy number is (0.1,0.1,0.3). From experiment 11 to experiment 15, the weight ratings of all evaluation indexes are respectively set as high / important, high/important, medium, low/unimportant, low/unimportant. According to the calculation steps of fuzzy TOPSIS, the 8 alternatives have been ranked for 15 times.

Table 7
Sensitivity analysis

Number Definitions Rank

1 $\tilde{W_{1}} = (0.7.0.9, 0.9), \tilde{W_{2 \sim 10}} = (0.1, 0.1, 0.3)$ S1>S2>S4>S5>S6>S3>S7>S8

2 $\tilde{W_{2}} = (0.7.0.9, 0.9), \tilde{W_{1, 3 \sim 10}} = (0.1, 0.1, 0.3)$ S1>S4>S5>S6>S3>S7>S8>S2

3 $\tilde{W_{3}} = (0.7.0.9, 0.9), \tilde{W_{1, 2, 4 \sim 10}} = (0.1, 0.1, 0.3)$ S1>S6>S5>S4>S3>S7>S8>S2

4 $\tilde{W_{4}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 3, 5 \sim 10}} = (0.1, 0.1, 0.3)$ S7>S4>S5>S3>S1>S8>S6>S2

5 $\tilde{W_{5}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 4, 6 \sim 10}} = (0.1, 0.1, 0.3)$ S4>S5>S8>S3>S1>S7>S7>S2

6 $\tilde{W_{6}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 5, 7 \sim 10}} = (0.1, 0.1, 0.3)$ S4>S5>S3>S1>S7>S8>S6>S2

7 $\tilde{W_{7}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 6, 8 \sim 10}} = (0.1, 0.1, 0.3)$ S4>S5>S3>S8>S1>S7>S6>S2

8 $\tilde{W_{8}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 7, 9 \sim 10}} = (0.1, 0.1, 0.3)$ S4>S5>S3>S1>S7>S8>S6>S2

9 $\tilde{W_{9}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 8, 10}} = (0.1, 0.1, 0.3)$ S3>S7>S4>S5>S8>S1>S6>S2

10 $\tilde{W_{10}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 10}} = (0.1, 0.1, 0.3)$ S4>S5>S7>S3>S8>S1>S6>S2

11 $\tilde{W_{1 \sim 10}} = (0.1, 0.1, 0.3)$ S4>S5>S3>S1>S7>S8>S6>S2

12 $\tilde{W_{1 \sim 10}} = (0.1, 0.3, 0.5)$ S4>S5>S3>S1>S7>S8>S6>S2

13 $\tilde{W_{1 \sim 10}} = (0.3, 0.5, 0.7)$ S4>S5>S3>S1>S7>S8>S6>S2

14 $\tilde{W_{1 \sim 10}} = (0.5, 0.7, 0.9)$ S4>S5>S3>S1>S7>S8>S6>S2

Number	Definitions	Rank
1	$\tilde{W_{1}} = (0.7.0.9, 0.9), \tilde{W_{2 \sim 10}} = (0.1, 0.1, 0.3)$	S1>S2>S4>S5>S6>S3>S7>S8
2	$\tilde{W_{2}} = (0.7.0.9, 0.9), \tilde{W_{1, 3 \sim 10}} = (0.1, 0.1, 0.3)$	S1>S4>S5>S6>S3>S7>S8>S2
3	$\tilde{W_{3}} = (0.7.0.9, 0.9), \tilde{W_{1, 2, 4 \sim 10}} = (0.1, 0.1, 0.3)$	S1>S6>S5>S4>S3>S7>S8>S2
4	$\tilde{W_{4}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 3, 5 \sim 10}} = (0.1, 0.1, 0.3)$	S7>S4>S5>S3>S1>S8>S6>S2
5	$\tilde{W_{5}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 4, 6 \sim 10}} = (0.1, 0.1, 0.3)$	S4>S5>S8>S3>S1>S7>S7>S2
6	$\tilde{W_{6}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 5, 7 \sim 10}} = (0.1, 0.1, 0.3)$	S4>S5>S3>S1>S7>S8>S6>S2
7	$\tilde{W_{7}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 6, 8 \sim 10}} = (0.1, 0.1, 0.3)$	S4>S5>S3>S8>S1>S7>S6>S2
8	$\tilde{W_{8}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 7, 9 \sim 10}} = (0.1, 0.1, 0.3)$	S4>S5>S3>S1>S7>S8>S6>S2
9	$\tilde{W_{9}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 8, 10}} = (0.1, 0.1, 0.3)$	S3>S7>S4>S5>S8>S1>S6>S2
10	$\tilde{W_{10}} = (0.7.0.9, 0.9), \tilde{W_{1 \sim 10}} = (0.1, 0.1, 0.3)$	S4>S5>S7>S3>S8>S1>S6>S2
11	$\tilde{W_{1 \sim 10}} = (0.1, 0.1, 0.3)$	S4>S5>S3>S1>S7>S8>S6>S2
12	$\tilde{W_{1 \sim 10}} = (0.1, 0.3, 0.5)$	S4>S5>S3>S1>S7>S8>S6>S2
13	$\tilde{W_{1 \sim 10}} = (0.3, 0.5, 0.7)$	S4>S5>S3>S1>S7>S8>S6>S2
14	$\tilde{W_{1 \sim 10}} = (0.5, 0.7, 0.9)$	S4>S5>S3>S1>S7>S8>S6>S2

Experiments 1 to 10 highlight individual index is one by one, reflecting different pertinence. Among them, Experiments 1 to 3 focus on the robustness of the supply network, and resulted by which, “S1: increasing the number of transport links” become the best solution for 3 times. Experiment 4 to 5 highlight the resource support ability of node members, and resulted by which, “S4: improving the accuracy of the plan” and “S7: improving the emergency material raising ability” become the best solution respectively. Experiment 6∼8 highlight the decision-making ability, and resulted by which, and “S4: improving the accuracy of the plan” become the best solution for 3 times. Experiment 9∼ experiment 10 highlight the cooperation ability, and resulted by which, “S3: Establishing green channel” and “S4: improving the accuracy of the plan” become the best solution respectively. The solution with the most times of ranking the first is S4, which reflects its good resilience enhancement ability; S1 also has a good performance; and then are S3 and S7. In addition, among 10 experiments, the ranking of each alternative has changed at least 1 time (such as S2), and some has changed 9 times (such as S7). It shows that the resilience enhancement effect of each alternative is different, and they will play different roles in different evaluation standards.

In experiment 11–15, the weight of each index is made equal, but the weight values are set to different sizes. The results show that the alternatives rank the same when all index values are equal. It shows that the contribution of each alternative to the resilience enhancement of the emergency logistics SC system is different.

4.5 Implications and recommendations

It has been discussed in detail about how to enhance the resilience of emergency logistics SCs and which alternatives can be given priority through the evaluation above. We are going to further explain how to operate and implement these prioritized alternatives.

(1) Increasing the number of transport links

Increasing the number of transport links is essentially to create redundancy of transport capacity. When some certain links of the emergency supply chain are interrupted, if there are alternative lines, the emergency materials supply can continue to maintain. Transport links can be increased in the following ways: first, increasing the number of emergency material reserve centers; second, increasing the number of emergency distribution points. These two approaches are aimed at the three-level network of “emergency resource reserve centers ⟶ emergency distribution points ⟶ disaster area” in the emergency logistics SC, and can disperse risks effectively. Third, applying diversified transportation modals to emergency distribution. For example, in case of serious damage to some key roads, aircraft transportation can be mobilized; in addition, various transportation modals such as railway, road, aircraft, Inland River, etc. can be combined to form a multimodal transportation system, which can increase transportation links effectively.

(2) Improving the ability of information monitoring and early warning

Emergency logistics SCs includes a series of links such as the supply, manufacturing, purchasing, distribution, distribution and demand feedback of emergency materials. The information exchange and processing between each link is a crucial issue in emergency management. The following methods can be used to improve the ability of information monitoring and early warning: first, applying advanced data collection technology and GIS information technology to grasp risk sources in time; second, applying visual display technology and video intelligent analysis technology to provide real-time image preview, video recording, early warning record query and other functions; third, applying big data technology to analyze the risk degree of major risk sources and real-time feedback of external environment changes, to realize dynamic risk warning.

(3)Improving the accuracy of the plan

The emergency department needs to formulate the emergency plan in real time according to the feedback of the disaster. An accurate plan can effectively speed up the rescue progress and improve the rescue efficiency and effect. The accuracy of the plan can be improved in the following ways: first, collecting and analyzing the relevant data and expert information of emergency operation at home and abroad, to establish and update the database in time; second, applying big data analysis and mining algorithm to analyze the risk factors with multi-dimension, and summarizing the topography, population, transportation, resources and other historical data to predict the risk trend.

5 Conclusions

Emergency logistics SC is a complex and risky system. The effective way to overcome the vulnerability of emergency logistics SC is to enhance the resilience of it. The resilience enhancement process of emergency logistics SC is a full-depth and multi-level defense system, involving a wide range of factors. In this paper, 8 resilience enhancement strategies of emergency logistics SC are put forward. Through adopting fuzzy-TOPSIS approach and sensitivity analysis, the influence mechanism of each evaluation index is deduced. “Increasing the number of transport links”, “Improving the ability of information monitoring and early warning” and “Improving the accuracy of the plan” are more effective resilience strategies in this case.

Although the paper studies the resilience of emergency logistics SC from system perspective, the mechanism of each enhancement strategy needs further analysis; meanwhile, the risk management of emergency logistics is a time-consuming and expensive project, but the cost of resilience enhancement of emergency logistics SC has not been considered in the evaluation. This will be the focus of the next step in research.

Footnotes

Acknowledgment

The work is supported by the Natural Science Foundation of Zhejiang Province (Grant: Y17G010025).

References

Sun

C.Y.

and Qin

Q.W.

, A discussion on several fundamental theoretical issues of paroxysmal events, Journal of Southwest China Normal University (Humanities and Social Sciences Edition)31 (2005), 50–53.

Chen

Z.Y.

, On synergy management in emergency supply chain dealing with unconventional emergencies, Journal of Beijing Institute of Technology (Social Sciences Edition)15 (2013), 95–99.

Kumar

, Managing risks in a relief supply chain in the wake of an adverse event, OR Insight24 (2011), 131–157.

Kumar

and Havey

, Before and after disaster strikes: A relief supply chain decision support framework, International Journal of Production Economics145 (2013), 613–629.

Chen

Z.J.

, Xu

and Cheng

, Reliability of emergency logistics supply chain, Railway Transport and Economy31 (2009), 69–71.

Gong

, Chen

and Hu

, Fuzzy entropy clustering approach to evaluate the reliability of emergency logistics system, Energy Procedia16 (2012), 0–283.

, Zhou

and Liu

, et al.Disaster risk assessment based on variable fuzzy sets and improved information diffusion method, Human and Ecological Risk Assessment: An International Journal19 (2013), 857–872.

Peng

, Li

J.L.

and Wang

S.S.

, et al.Robust location routing optimization for multi class emergency resource allocation, Chinese Journal of Management Science6 (2017), 143–150.

Tofighi

, Torabi

S.A.

and Mansouri

S.A.

, Humanitarian logistics network design under mixed uncertainty, European Journal of Operational Research250 (2015), 239–250.

10.

W.Y.

, Robust model of emergency material allocation under uncertain network structure, Control and Decision28 (2013), 1898–1902.

11.

Jeong

K.Y.

, Hong

J.D.

and Xie

, Design of emergency logistics networks, taking efficiency, risk and robustness into consideration, International Journal of Logistics Research and Applications17 (2014), 1–22.

12.

Chen

C.X.

, Study on invulnerability of emergency logistics network based on complex network, Application Research of Computers29 (2012), 1260–1262.

13.

, Zhou

and Liu

, The evaluation and optimization of post-earthquake emergency rescue road network in shanghai historic area, Transportation Systems Engineering and Information17 (2017), 227–233.

14.

Martin

and Helen

, Building the resilient supply chain, International Journal of Logistic Management1 (2004), 56–61.

15.

Peck

, Drivers of supply chain vulnerability: an integrated framework, International Journal of Physical Distribution & Logistics Management34 (2005), 210–232.

16.

Rice

and Caniato

, Building a secure and resilient supply network, Supply Chain Management Review9 (2009), 435–443.

17.

Rice

J.B.

and Sheffi

, A supply chain view of the resilient enterprise, Mit Sloan Management Review47 (2005), 41–48.

18.

Cimellaro

G.P.

, Reinhorn

A.M.

and Bruneau

, Framework for analytical quantification of disaster resilience, Engineering Structures32 (2010), 3639–3649.

19.

Paton

, Smith

and Violanti

, Disaster response: Risk, vulnerability and resilience, Disaster Prevention and Management9 (2000), 173–180.

20.

Calgaro

, Lloyd

and Dominey-Howes

, From vulnerability to transformation: a framework for assessing the vulnerability and resilience of tourism destinations, Journal of Sustainable Tourism22 (2014), 341–360.

21.

Carvalho

, Barroso

A.P.

and Virgínia

, et al., Supply chain redesign for resilience using simulation, Computers & Industrial Engineering62 (2012), 329–341.

22.

Lam

J.S.L.

and Bai

, A quality function deployment approach to improve maritime supply chain resilience, Transportation Research Part E: Logistics & Transportation Review92 (2016), 16–27.

23.

Y.W.

and Zhao

L.D.

, To discuss on the resilience of emergency logistics network, Value Engineering27 (2008), 1–3.

24.

, Yang

J.Q.

and Wang

H.Y.

, et al., Research on resilience design model of emergency material distribution network, Journal of Wuhan University of Technology (Transportation Science & Engineering)42 (2018), 727–731.

25.

Nujoom

, Wang

and Mohammed , et al., Optimization of a sustainable manufacturing system design using the multi-objective approach, International Journal of Advanced Manufacturing Technology96 (2018), 2539–2558.

26.

, Yang

J.Q.

and Wang

H.Y.

, et al., A Quality Function Deployment to Emergency Logistics Supply Chain Resilience, Journal of Wuhan University of Technology (Transportation Science & Engineering)42 (2018), 942–946.

27.

Bottani

and Rizzi

, Strategic management of logistics service: A fuzzy QFD approach, International Journal of Production Economics103 (2006), 585–599.

28.

Straka

, Malindzakova

and Trebuna

, et al., Application of extendsim for improvement of production logistics efficiency, International Journal of Simulation Modelling16 (2017), 422–434.

29.

Huang

S.T.

, Bulut

and Duru

, Service quality assessment in liner shipping industry: An empirical study on Asian shipping case, International Journal of Shipping & Transport Logistics72 (2015), 221–242.