3.1.2. Environmental Factors (*C*2)

This refers to those factors which will influence the immediate surroundings of the charging station. A clean and green environment helps in resonating the theme of electric vehicles and thereby makes a charging station built in such a location a success.


### 3.1.3. Traffic Factors (*C*3)

This refers to those sets of factors which are only noticeable when there is a huge population in the area. A charging station which takes into consideration the traffic factors is one which will be able to have huge implications on the state of travel in the area.


#### 3.1.4. Societal Factors (*C*4)

With the advancement in science and technology, the factors promoting wellness of human life have taken a back seat. The societal factors point out the major problems which may be looked at to improve the quality of the society at large.

• Adverse impact of noise and electromagnetic fields (AI)(*C*41): (Due to the construction of the electric vehicle charging station). An electric vehicle charging station has a constant aura of noise and an electromagnetic field surrounding it during the construction

phase which might cause a certain category of people to develop problems. If proper measures can be taken in the initial stage, then this impact may be minimized since public health is of utmost importance.

• Population density (*C*42): This stands for the number of people living in each unit of area. When the population density in a locality is high, it shows that the area is overcrowded. There will be more consumption in such an area and the construction cost will also be high, but the quality of life will usually be low. The need of transportation in such an area is usually very high and an electric vehicle charging station constructed in such an area may just be what the people need.

Tables 1 and 2 describes the factors, sub-factors and alternatives considered for this research.


**Table 1.** Factors and sub-Factors considered in the present study.

**Table 2.** Alternatives selected in the study and their corresponding nearby location, latitude and longitude.


Figure 3 represents the hierarchical framework of the problem in the study.

The rural, urban and total population for the census year starting from 2001 and 2011 has been represented in a graph (Figure 4). In this figure, the projected rural, urban and total population also been forecasted for the years 2021, 2031, 2041 and 2051 in a chronological way. The graph clearly indicates a consistent increase in the population from 4.23 million in 2001 to 6.60 million in 2051 (projected) with a sharp increase in the urban population, and the urban population growth is much higher than the rural, which

indicates higher increasing infrastructural demand. The projected population made us think about the future transport services in the area.

**Figure 3.** Hierarchical structure of the problem.

**Figure 4.** Projected growth of the population of rural and urban areas of the Howrah district.

Urbanization is considered one of the most noteworthy anthropogenic inputs of the environmental framework, and thus the present study considered the spatio-temporal characteristics of urban growth and its inference in the transport of Howrah. The built-up land (Figure 5) has been generated using the NDBI (normalized difference built-up index) with the following equation NDBI = MIR − (NIR/MIR) + NIR (Zha, 2003) [86]. Here, NIR is anear-infrared band such as ETM+ and TM, and LISSIII is a band no.4; MIR is a middle infrared band such as ETM+ and TM and LISSIII is a band no.5.The index is based on the unique spectral response of built-up lands that have a higher reflectance in the MIR wavelength range as compared to that in the NIR wave length range. Thereafter, the NDBI mapping has been prepared to understand the level of urbanization from 2000 (Figure 5a) to 2010 (Figure 5b) in the study area. It helps to correlate the changes in land use patterns and its consequences to the water storage of the study area. The NDBI values range from −1 to +1. Very low values of the NDBI (0.1 and below) correspond to non-urban features, while higher values indicate the covering of areas of impervious surfaces such as asphalt and concrete. To understand the levels of urbanization, the NDBI values have categorized into five zones, which range from −0.96 to 0.45 in 2000 and from −0.95 to 0.71 in 2010. The result with a negative value of NDBI represents the water bodies and vegetation covers, whereas the higher and positive values represent the build-up areas. The map of the year 2000 indicates a huge range of water bodies spread all over the district along with the two major rivers; the Rupnarayan River in west and south west and the Bhagirathi-Hooghly River in the east and south-east side. The major built-up areas are mostly concentrated over the north-east and north-west corners of the district covering the Bally-Jagacha and Udaynarayanpur blocks, respectively.

**Figure 5.** Normalized difference built-up index mapping for level of urbanization during (**a**) 2000 (**b**) 2010.

The transport network of the Howrah district is mapped (Figure 6) and the closest census towns (Figure 7) to the Howrah Municipal Corporation (HMC) were also plotted to understand the importance of daily communication with the Howrah station or surrounding areas. The railway and the road transport connectivity with the Howrah station, which is located in the HMC, is very good. A large number of daily commuters are traveling to the area, mostly for economic and educational purposes. It indicates the concentration of traffic in the area, and in turn increases the public transport connectivity as well as the requirement for improving the local transport system.

**Figure 6.** Transport Network in and around Howrah.

**Figure 7.** Transport Network of Howrah district. Along with the Census Town.

The ten selected points/locations (Figure 8) were mapped in the HMC to understand the spatial coverage and important transport nodes in the area of the current study. The population distribution of each of the 66 wards was also mapped (Figure 9) to observe the population pressure, which is also able to justify the present selection of the ten locations for the study. Most of the selected locations are in densely populated areas, where the public and local transport services are becoming very essential. Furthermore, the Howrah Maidan (*S*4) and Salkia (*S*8) are highly densely populated (Figure 9) and thus very important in terms of transport services, whereas, Liluah (*S*5), Belur (*S*10) and Shibpur (*S*7) belong to highly densely populated and Kadamtala (*S*6), Bakultala (*S*9), Belgachia (*S*3) belong to moderately populated and Dasnagar (*S*1) and Santragachi (*S*2) belong to comparatively less populated among the ten points. But all these points are almost equally important and essential in terms of either population or transport or both. Sub criteria scores of each location has been represented in Table 3.

**Figure 8.** Location of points in HMC.

**Figure 9.** Ward wise population distribution of HMC.


#### **Table 3.** Sub-factors taken for the study.

Linguistic variables in HFN required for the comparison of factors and sub-factors are shown in Table 4.

**Table 4.** Linguistic term in HFN 1–9 scale.


Comparison between factors in linguistic variables given by two DMs is presented in Table 5.

The above Table 6 represents preference of factors in defuzzified form using Equation (18).Normalized matrix is obtained using Equation (20); priority weight of factors are calculated using Equation (21). Societal factors obtain the maximum weight of 0.430, followed by environmental factor 0.37, followed by traffic factor 0.139 and economic factor 0.065.



**Note.** Two DMs are considered here in the study, their opinions are combined using Equation (19).


**Table 6.** Representation of preference of factors in defuzzified form using Equation (18).

The weight of factors obtained is represented in Table 7.


**Table 7.** Representation of the priority weight of factors.

Using Equation (22), C.I is calculated to be 0.08. For R.I, as n = 4, the value is 0.09. Thus using Equation (23), *C*.*R* = 0.08 0.9 = 0.09 < 0.1, hence the matrix is consistent.

In the similar way, the comparison analysis of sub-factors with the help of two DMs has been calculated. The fuzzy weight of factors, sub-factors and global fuzzy weight are represented in Table 8.

**Table 8.** Hexagonal fuzzy weights of factors, sub-factors and global weight.


The linguistic terms used in this study for the rating of alternatives with respect to sub-factors. Preference of alternatives with respect to sub-factors are expressed in linguistic terms, scores and crisp values are depicted in Table 9. Note 2. The sub-factors (*C*22) and (*C*23) are assigned score using Table 3; later, it is converted to HFN. A score of 9 implies "very high" and so on.


**Table 9.** Comparison analysis in linguistic variables for preference of alternatives with respect to sub-factors.

Note: For *C*22, *C*<sup>23</sup> and *C*<sup>31</sup> are represented with crisp numbers.

#### *3.2. Ranking of Alternatives Using Fuzzy AHP-TOPSIS Method*

Following the steps represented in Section 2.7, distance measure, relative closeness and the ranking of sites has been computed as depicted in Table 10.



*3.3. Ranking of Alternatives UsingFuzzy AHP-COPRAS*

Following the steps represented in Section 2.8, *R*+*g*, *R*−*g*, *Rg*, *R* and ranking are computed as depicted in Table 11.

**Table 11.** Values of *R*+*g*, *R*−*g*, *Rg*, *RR* and ranking are represented.


#### **4. Comparison Analysis and Sensitivity Analysis**

Two different MCDM techniques, fuzzy AHP-TOPSIS and fuzzy AHP-COPRAS were employed for the selection of the optimal site for aelectric vehicle charging station in and around the city of Howrah, West Bengal, India. Figure 10 denotes the comparative ranking obtained under the two methodologies, fuzzy AHP-TOPSIS and fuzzy AHP-COPRAS. We also tried to compare the alternatives ranking using the two different said proposed methods.

**Figure 10.** Representation of the ranking obtained under the two MCDM techniques.

A sensitivity analysis was conducted to see the ranking obtained under different changing conditions. Figures 11 and 12 represents the clustered column chart to compare the ranking with the interchange of the sub-factors' weight. Two different cases are taken. In the first case, the sub-factors parking facilities '*C*15' and population density '*C*42' weights were interchanged. In the second case, land cost '*C*11' and generation of noise and air pollution '*C*21' weights were interchanged. For both these cases, two different methodologies were used in this study, i.e., fuzzy AHP-TOPSIS and fuzzy AHP-COPRAS.

**Figure 11.** Sensitivity analysis ranking obtained under fuzzy AHP-TOPSIS.

Figure 11, i.e., ranking obtained by fuzzy AHP-TOPSIS under sensitivity analysis shows that the alternative (*S*4), (*S*5), (*S*6), (*S*10), (*S*3) and (*S*2) are consistent with first, second, third, fourth, fifth and sixth position, respectively, under the considered two cases, whereas following Figure 12, i.e., ranking yield by fuzzy AHP-COPRAS under sensitivity analysis shows that the sites (*S*4), (*S*5),(*S*2), (*S*10),(*S*3) and (*S*9) scores the rank of first,

second, fifth, seventh, ninth and tenth position, respectively. The other sites' variation of rank is noticed under the sensitivity analysis and depicted in the mentioned figures.

**Figure 12.** Sensitivity analysis ranking obtained under fuzzy AHP-COPRAS.

#### **5. Results and Discussion**

This section discusses the results obtained by the methodology FAHP-TOPSIS, FAHP-COPRAS and the sensitivity analysis. The ranking obtained under the two MCDM techniques yield the site "Howrah Maidan" (*S*4) as the best alternative for e-vehicle site selection followed by "Liluah" (*S*5) and "Belur" (*S*10). The FAHP-TOPSIS ranked the alternative "Belur" (*S*10) and "Salkia" (*S*8) equally at the third position. Ranking obtained for all the sites are presented in Tables 10 and 11. In the sensitivity analysis, where the sub-factors' weight are interchanged as discussed in Section 5, it is seen that the site selected "Howrah Maidan" (*S*4) consistently remains in the first position. The rankings obtained under the sensitivity analysis by the two methods are depicted in Figures 11 and 12.

#### **6. Conclusionsand Future Scope**

Ease of commutation, a pollution-free mode of transport, as well as employment generation are direct and indirect benefits of the e-vehicle. For developing countries where the pollution level is quite high and proportion of the roads is lower, e-vehicles can be a game changer. According to our study, considering ten locations across the city of Howrah Maidan due to its proximity to India's largest railway station Howrah and various other attributes ranked it as number one, followed by Liluah and Belur. These three locations are consistent with rank one, two and three, respectively, irrespective of the two MCDM methodologies applied in this research. High population density, enhanced level of consumerism, and the presence of various public facilities with higher footfalls led to this higher ranking.

This paper used the GIS and MCDM tools FAHP, F-TOPSIS and F-COPRAS to obtain the optimal site selection for the e-vehicle charging station. HFN has been used to give a preferential rating of factors, sub-factors and alternatives. The ranking obtained using MCDM tools are logical and scientific. The present findings provide important references for future potential work and problem solving.

For the site selection, factors such as environmental, economic, traffic, societal are incorporated with their respective sub-factors. Through this research, it is observed that the societal factor is the most significant and out of the sub-factors, population density is the most important.

HFN is used here as it captures the hesitancy and vagueness in an efficient way. To practice the qualitative criteria evaluation for imprecise information, FAHP, F-TOPSIS and F-COPRAS are used. Comparative analysis which uses F-TOPSIS and F-COPRAS in our example has depicted consistent results. The reliability, robustness and efficiency of this methodology has also been tested through sensitivity analysis. HFN captures a wider range of linguistic terms but usage of HFN makes computation a bit longer.

In Howrah, India, a very old and unplanned city, creating e-vehicle charging infrastructure across the city is important. A site like Santragachi (*S*2) in this research which ranks lower at present could acquire a significant position due to huge infrastructural investment by the government in that area. A big railway terminus is coming up, which has the potential to change the demography of the area. Thus futuristic sub-factors can be incorporated into future research. The absence of futuristic sub-factors is a limitation at present. A larger number of decision makers based on the administrative point of view can be explored in future research. The proposed methodology used in this research can be applied in different fields such as new vendor selection, and treatment selection for new diseases where more ambiguity and uncertainty is prevalent. The other MCDM tools such as PROMETHEE, VIKOR, and WASPAS can be used in the future with intuitionistic, neutrosophic, and hesitant fuzzy numbers to yield improved, robust and practical solutions.

**Author Contributions:** Conceptualization, A.G., S.P.M.; methodology, S.P.M. and N.G.; software, B.K.M.; validation, A.D. and M.S.G.; formal analysis, A.G., N.G.; investigation, S.K. and A.G.; resources, B.K.M.; data curation, B.K.M. and S.P.M.; writing—original draft preparation, S.K. and A.G.; writing—review and editing, S.P.M. and N.G.; visualization, B.K.M.; supervision, A.G., S.P.M.; project administration, A.G. and S.P.M.; funding acquisition. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** We mention the source of used data in the work.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

