A new soft computing approach for analyzing the influential relationships of critical infrastructures
Introduction
Critical infrastructure (CI) includes both the physical facilities and virtual assets upon which a nation's strength is built. Every nation relies on its CI to maintain the normal functions of society. Any damage to or failure of CI can seriously affect a nation's safety, public health, and economy [1]. Infrastructure can be defined as a network of independent, mostly privately owned, man-made systems and processes that function collaboratively and synergistically to produce and distribute a continuous flow of essential goods and services [2]. Thus, CI is equivalent to the central nervous system of the country, damage to a part of which can affect the security of the whole. Furthermore, a nation cannot achieve its goals, be it energy sustainability and economic or social development, if its CI network is at risk or undeveloped [3]. During normal operation, the interdependence among the components of CI is often unnoticeable. However, when operational anomalies occur, such failures due to aging, poor maintenance, deliberate attacks, or natural disasters, the strong degree of correlation between components becomes evident. For example, the 2011 Fukushima Daiichi nuclear disaster was one of the largest CI accidents in Asia. It caused the power supply to fail, which in turn affected communications and transportation throughout the region [4]. It should also be noted that the interdependencies of CI components are usually bidirectional. For example, dams not only store water, but they can also generate electricity through kinetic energy, which in turn supplies the energy required for the operation of the dam. As the number incidents of CI failure continues to increase worldwide, researchers are increasingly motivated to study what leads to vulnerability and resilience, building influential network relationships for CI systems to assess the effectiveness of their components and determine priorities. Two issues always considered by decision makers when CI protection plans are formulated, namely the CI network itself (the interdependencies of the components) and how to prioritize individual parts of the network according to their complex interactions [5], [6], [7], [8], [9], [10], [11], [12].
Various models and methods have been developed for the analysis of interdependent infrastructure systems. Utilizing qualitative and semiquantitative research methodologies, Utne et al. [5] introduced a two-stage assessment method for the city facilities of Oslo, Norway. In the first stage, historical hazard events were categorized and CI interdependencies identified from the results of hazard and operability analysis. In the second stage, fault tree analysis and event tree analysis methods were employed to determine the degree of CI dependence. Fang and Sansavini proposed a multimode restoration model [13]. They used the restoration framework to investigate the resilience and interinfluence of CI failures considering the uncertainties of repair time and availability of resources. Ouyang [10] developed a mathematical framework which they called resilience optimization of interdependent CI systems against spatially localized attacks (ROICISs-SLAs). This can be categorized as a nonlinear mixed-integer programming problem. They also proposed an algorithm to determine the optimal solution for CI resilience. Foytik and Robinson [11] developed a real-time evacuation planning and simulation model (RtePM) to decide upon measures for the evacuation of the components of CI taking into consideration the importance of connections between them. The results indicate that disruption of communication influences traffic congestion and the efficiency of crowd evacuation.
In recent years, several studies have employed multi-criteria decision-making (MCDM) methods to define the influential relationships within the CI system. MCDM is often used by industry professionals to prioritize alternatives derived from expert opinions relevant to specific decision-related problems. These methods can be further categorized as multi-attribute decision-making or multi-objective decision-making methdods. State-of-the-art reviews of MCDM and its applications can be found in several articles [14], [15], [16], [17], [18], [19], [20], [21]. However, very few researchers have employed MCDM models, either directly or indirectly to CI related problems [7,8]. Although past studies have made significant contributions, it has been assumed in the modeling that the elements of CI are independent or crisp numbers have been used to reflect the experts’ subjective opinions. Unlike prior studies that relied heavily on computer simulations, Singh et al. [7] developed a hybrid model that combines the interpretive structural modeling (ISM) technique and cross-impact matrix multiplication applied to classification (MICMAC) method, which they used to explore the dependencies among CI sectors. The model serves as the basis by which the Government of India develops CI management policies. However, ISM can only determine whether CIs influence each other, not the level of influence. To overcome the shortcomings of ISM, Huang et al. [8] combined the graph-theory-based decision-making trial and evaluation laboratory (DEMATEL) technique with the analytic network process (ANP) approach (called DANP) for construction of the interdependent relationships among CIs. Unfortunately, they used an arithmetic average method for integration of questionnaire responses from multiple experts, which led to information loss and did not allow information uncertainty to be considered. Therefore, we propose a new hybrid MCDM model that integrates the rough number, fuzzy theory and DANP approaches (called rough-fuzzy DANP or RF-DANP) for analyzing influential relationships among CIs. The selection and definition of CIs are based on the 2019 report of the Homeland Security Office of Taiwan. The contributions and innovations of our study are outlined below:
- (I)
The incorporation of fuzzy theory into the DANP method ensures that the evaluation procedure is comprehensive in uncertain environments.
- (II)
The rough number-based approach is combined with DANP, which can effectively integrate the judgments of multiple experts to enhance the practicability of DANP. In our review of the relevant literature, we found no studies combining the rough-fuzzy method with DANP when discussing CI.
- (III)
Our study compares a variety of MCDM methods, including the analytic hierarchy process (AHP) [14], the best worst method (BWM) [22], and ANP [17]. However, it is assumed in both AHP and BWM that the criteria are independent, which is not the case in actual situations. In addition, ANP cannot describe the influential relationships among the criteria. The proposed method is more effective and practical in real-world cases.
- (IV)
The proposed approach serves as a better evaluation framework for determination of the influential relationships of CIs. The quality and efficiency of the solution is not affected by the number of CIs.
In summary, the RF-DANP method is a new concept for evaluating influence relationships among alternatives capable of comprehensively considering various qualitative factors derived from expert knowledge. The remainder of this paper is organized as follows. Section 2 presents the DEMATEL, ANP, and rough number methods as well as the fuzzy theory. Section 3 introduces the proposed RF-DANP approach. Section 4 describes a real-world application in Taiwan. Section 5 compares a few models and discusses the managerial implications. In Section 6 some conclusions are offered and directions for future research discussed.
Section snippets
Literature review
A review of the relevant literature and methods is presented in this section starting with an introduction to the rough number method and description of a few basic concepts of fuzzy set theory and the DEMATEL technique. Finally, we introduce the DANP method.
Proposed approach: rough-fuzzy DANP method
This section details the proposed rough-fuzzy DANP approach and its calculation process, which can be divided into three subsections: methodological assumptions and scope, construction of a complex evaluation system by using rough-fuzzy DEMATEL, and finally, obtaining criteria weights by using rough-fuzzy DANP.
Empirical example
To protect national CIs, Taiwan's government launched their NIPP, in which seven major CIs (sectors) are identified, each comprising 3–10 subsectors. The latest major CI classification was released in 2019. Because Taiwan is an island with few natural resources, its economy is completely dependent upon industrial production. The related CIs are vital to Taiwan's economic development. The main objectives of the NIPP include assuring operation of various CIs, protecting critical facilities,
Discussion and managerial implications
As disasters continue to occur, researchers and practitioners are increasingly focusing on the interdependence of critical infrastructures when making plans to protect them. Many frameworks for quantitative and qualitative analysis of CIs have been proposed [4], [5], [6],[8], [9], [10],[55], [56], [57], [58], [59]. The traditional methods for the analysis of interdependencies between the CI components are computer simulations, such as agent-based, dynamic or Monte Carlo simulations. However,
Conclusions and remarks
The safety of the CI system is one of the most significant issues facing nations around the world. Achieving the goals of energy sustainability, economic development, or social development is difficult if the infrastructure network is at risk. Taiwan is endangered by many natural calamities (e.g., typhoons and earthquakes) every year because of its geographical location. Therefore, determining the influence relationships and importance of CIs is important for assisting the government in making
Declaration of Competing Interest
All authors declare that they have no conflict of interests.
Ethical approval
This article does not contain any studies with human or animals performed by any of the authors.
References (60)
- et al.
A contribution of axiomatic design principles to the analysis and impact of attacks on critical infrastructures
Int. J. Crit. Infrastruct. Prot.
(2018) Review on modeling and simulation of interdependent critical infrastructure systems
Reliab. Eng. Syst. Saf.
(2014)- et al.
Methodologies and applications for critical infrastructure protection: state-of-the-art
Energy Policy
(2011) - et al.
Hybrid systems modeling for critical infrastructures interdependency analysis
Reliab. Eng. Syst. Saf.
(2017) - et al.
A method for risk modeling of interdependencies in critical infrastructures
Reliab. Eng. Syst. Saf.
(2011) - et al.
Risk analysis of critical infrastructures emphasizing electricity supply and interdependencies
Reliab. Eng. Syst. Saf.
(2012) - et al.
Identifying critical infrastructure sectors and their dependencies: an indian scenario
Int. J. Crit. Infrastruct. Prot.
(2014) - et al.
A method for exploring the interdependencies and importance of critical infrastructures
Knowl Based Syst
(2014) - et al.
Critical infrastructure dependencies: a holistic, dynamic and quantitative approach
Int. J. Crit. Infrastruct. Prot.
(2015) A mathematical framework to optimize resilience of interdependent critical infrastructure systems under spatially localized attacks
Eur. J. Oper. Res.
(2017)
Weighting critical infrastructure dependencies to facilitate evacuations
Int. J. Disaster Risk Reduct.
Critical infrastructure vulnerability–a method for identifying the infrastructure service failure interdependencies
Int. J. Crit. Infrastruct. Prot.
Power planning in ict infrastructure: a multi-criteria operational performance evaluation approach
Omega (Westport)
Interrelationships of risks faced by third party logistics service providers: a dematel based approach
Transp. Res. Part E: Logist. Transp. Rev.
An integrated dematel-anp approach for renewable energy resources selection in turkey
Int. J. Prod. Econ.
Critical success factors for reverse logistics in indian industries: a structural model
J. Clean Prod.
A multi-criteria port suitability assessment for developments in the offshore wind industry
Renew. Energy
A review of multi criteria decision making (MCDM) towards sustainable renewable energy development
Renew. Sustain. Energy Rev.
A novel failure mode and effect analysis model for machine tool risk analysis
Reliab. Eng. Syst. Saf.
An integrated ahp and vikor for design concept evaluation based on rough number
Adv. Eng. Inf.
Developing sustainable supplier selection criteria for solar air-conditioner manufacturer: an integrated approach
Renew. Sustain. Energy Rev.
Fuzzy sets
Inf. Control
An integrated model for solving problems in green supplier selection and order allocation
J. Clean Prod.
Why triangular membership functions?
Fuzzy Sets Syst.
The concept of a linguistic variable and its application to approximate reasoning–I
Inf. Sci. (Ny)
A DEMATEL method in identifying key success factors of hospital service quality
Knowl. Based Syst.
Using a hybrid method for evaluating and improving the service quality of public bike-sharing systems
J. Clean Prod.
Building an effective safety management system for airlines
J. Air Transp. Manag.
Integrated model of hot spring service quality perceptions under uncertainty
Appl. Soft Comput.
Sustainable recycling partner selection using fuzzy DEMATEL-AEW-FVIKOR: a case study in small-and-medium enterprises (SMEs)
J. Clean Prod.
Cited by (25)
Evolution of cyber-physical-human water systems: Challenges and gaps
2023, Technological Forecasting and Social ChangeBuilding an assessment model for measuring airport resilience
2021, Journal of Air Transport ManagementCitation Excerpt :Taiwan's special geographical position, an island surrounded by the sea, means that most people rely upon the air transport system to enter and leave the country. In addition, earthquakes occur all year round and there are frequent typhoons in the summer (Lo et al., 2020). In Section 2, we discussed the establishment of a framework for airport resilience assessment, which includes the dimensions of detection capability (D1), resistance capability (D2), rescue capability (D3), and recovery capability (D4), as well as the 27 criteria that are classified under them.
A group decision-making approach for exploring trends in the development of the healthcare industry in Taiwan
2021, Decision Support SystemsCitation Excerpt :Only the influential relationship is obtained without the degree of influence. DEMATEL optimizes the evaluation performance of ISM for construction of the network influential relationship and influential degree among the factors as an influential network relationship map (INRM) [23]. This study proposes a modified Z-DEMATEL technique, which not only considers the uncertainty of the decision environment, but also measures the reliability (confidence) of experts in the evaluation.
Exploring competitive sports technology development: using a MCDM model
2023, Journal of Physical Education and Sport