Introduction

Sustainable development (SD) is a concept that has emerged due to the negative environmental and social consequences of unilateral economic development. SD attempts to provide for comprehensive and balanced development. In the past few years, studies have mainly focused on different aspects of such development and have analyzed different variables that could play significant roles in SD in different countries. For example, the analysis of the strategic sustainability of electric vehicles in the European Union today and by 2050 is examined in Boren and Ny (2016). The EU aims to achieve sustainable transport, including a major reduction in greenhouse gas emissions from fossil fuel vehicles, particularly buses, trucks, and cars by 2050. Ahn et al. (2015) evaluate the optimization of energy resources for sustainable development in Korea. Policymaking based on sustainable transport development has been explored to increase the penetration of electric vehicles and thus reduce greenhouse gases in California (Javid and Nejat 2017). Jafari et al. (2017) consider the economic and environmental aspects of sustainability in the waste recycling process.

Compared to conventional vehicles (CVs), electric vehicles (EVs) make less noise and pollution using rechargeable batteries as their energy source and electric motors as their driving force. These vehicles are of three main types: battery electric vehicle (BEV), hybrid electric vehicle (HEV), and plug-in hybrid electric vehicle (PHEV). BEVs have an electric motor with batteries for electric power. Batteries can be charged through both the network connection and the car’s brake energy. The main disadvantage of these vehicles is their full dependence on the battery. Therefore, the driving range of BEVs is relatively lower than that of the conventional ones. HEVs have both fuel and electric motors with enough battery power to store fuel from the engine and the car’s brakes. The main disadvantage of these vehicles is the inability to recharge batteries from the network, thus depending on the fossil fuel engine. PHEVs are designed to eliminate the disadvantages of HEVs. PHEVs are rechargeable from the network, and the batteries should be capable of rapid discharge and fast charging. They can work alone with fuel and electric motor.

The use of electric vehicles has many benefits in reducing CO2 emissions and reducing dependence on fossil fuels in the transport sector. Therefore, many countries have determined targets for developing EVs in recent years and have implemented policies to achieve environmental goals and reduce energy consumption. Sustainable options for electric vehicle technologies have been investigated in different studies (Poullikkas 2015). The introduction of incentive policies for the approval of electric vehicles across countries was investigated in Zhang et al. (2014). The charging of electric vehicles in China’s power system was evaluated by Li et al. (2016). In their study, energy, economic, and environmental issues of EV development were studied, and its policy elements were analyzed. Various policy scenarios for the development of electric vehicles were presented in Zhou et al. (2015), Statharas et al. (2019), and Wang et al. (2018).

Various studies (see for example, Axsen and Kurani 2009) have shown that distribution networks are severely affected by the high penetration of electric vehicles that are charged uncoordinatedly. These effects can include increasing the maximum load and power losses and decreasing the system voltage. It has been suggested that the main solution to these inappropriately uncoordinated vehicle charging effects is to create a coherent process for charging. An effective way to manage the charging of electric vehicles is to use domestic consumption tariffs. Several power companies in different countries have offered hourly tariffs for their customers. The demand response is another effective method that can be used to manage the charge of vehicles (Green et al. 2010). In this method, the operator will allow the customer to cut off at peak time. According to Davidov and Panto (2017), the planning of electric vehicle infrastructures has been assessed based on the reliability of the charge and the quality of the services. Pye et al. (2015) suggested an uncertainty analysis for long-term energy planning in the UK. Based on their study, about 55% of air pollution caused by transportation was caused by cars. Thus, as they proposed, a pressing need is felt for sustainable development-based policymaking to reduce air pollution caused by cars.

There are several methods to evaluate policymaking. Using the data from cities around the world, Haghshenas et al. (2015) tried to assess the sustainable policy in the urban transport system by the use of dynamics systems. According to Lee et al. (2016), the hybrid electric vehicle market penetration model is capable of determining the best policy mix based on the consumer ownership cycle approach.

In the same line, the models of prediction of hybrid electric vehicle market have been investigated through Vensim software by the causal loop diagram. In Prebeg et al. (2016), long-term energy planning for the Croatian power system is presented using a multi-objective optimization approach focusing on renewable energies and the integration of electric vehicles. Sellitto et al. (2014) propose a regular environmental approach using the multi-criteria analysis to evaluate bus transit performance.

There are various methods to evaluate the performance of policies most important among which are fuzzy TOPSIS (Chen and Hwang 1992) and VIKOR (El-Santawy 2012). Kougias et al. (2016) also proposed a tool for the sustainability policymaking of renewable energy systems by MATLAB software. Moreover, Lechman et al. (2019) proposed a life cycle cost analysis method to produce two-layer laminates. The disadvantages of these methods are that they only consider one group of decision-makers. The game theory is another way to evaluate policy performance. The game theory approach considers models of interaction between different decision-makers (Attia et al. 2016). A dynamic game theory approach is proposed to develop a robust and secure distribution policy to optimize non-random Markov policy (Yang 2018). Madani and Hooshyar (2014) propose a learning theory-based game theory to determine the optimal performance of policies in multi-operator multi-repository systems. Multi-objective game theory and fuzzy programming approaches have been proposed to equilibrium economic development and environmental impacts (Moradi and Limaei 2018). The dual goals of multi-objective game theory and fuzzy programming approaches to minimize the detrimental environmental impacts and to maximize the economic revenues derived from various land uses have been outlined in Proskuryakova (2018). A three-level Stackelberg game has been proposed for modeling the interaction between the supplier, the charging infrastructure operator, and investors by Zhu et al. (2017). In Guo and Zhao (2015), the multi-criteria decision-making method for choosing the location of EVs’ charging station is provided based on the TOPSIS fuzzy method.

In this paper, sustainable development policymaking for the acceptance and penetration of PHEVs in Tehran has been analyzed using fuzzy TOPSIS and game theory. The purpose of the study is to evaluate policy scenarios for PHEV development in Tehran from both the vehicle buyers and the state’s point of view. The first section of the paper involves the introduction, which is followed in section two, by describing and analyzing fuzzy TOPSIS and game theory methods. Then, we discuss the extended model of sustainable development and examine the sustainability of existing vehicles in Tehran. Also, in this section, the parameters affecting the evaluation of vehicle sustainability in Tehran are determined over a 16-year period from 2002 to 2018. In the third section, sustainable development policymaking is evaluated by introducing some policy indicators to expand the number of PHEVs in Tehran using fuzzy TOPSIS and game theory. In the fourth section, the selected policy indicators are discussed, and finally, the number of PHEVs is predicted in 2032.

Background and problem statement

In this study, both the fuzzy TOPSIS and game theory method are used to increase the consistency of the obtained policymaking results. The advantage of this approach is that it considers both fuzzy TOPSIS and game theory for the development of PHEV from the vehicle buyers and the state’s point of view.

Fuzzy TOPSIS method

People’s thoughts are always associated with uncertainty, and this uncertainty affects decisions. Fuzzy logic is a type of reasoning method that is similar to how one would reason. The fuzzy logic approach follows human decision-making. One of the fuzzy decision-making methods is fuzzy TOPSIS for ranking options. In this method, the elements of the decision matrix, or the weight of the indices, or both are expressed as fuzzy numbers. The steps of the fuzzy TOPSIS method are as follows:

  • Make a fuzzy decision matrix with dimensions \(m \times n\) for individuals’ views.

$$\tilde{D} = \left[ {\begin{array}{*{20}c} {\tilde{d}_{11} } & {\tilde{d}_{12} } & \ldots & {\tilde{d}_{1n} } \\ . & . & . & . \\ . & . & . & . \\ {\tilde{d}_{m1} } & {\tilde{d}_{m2} } & \ldots & {\tilde{d}_{mn} } \\ \end{array} } \right]$$
(1)

where \(d_{ij}\) represents the ith option in the jth sub-criterion and is a fuzzy number represented by a triangular number as \(\tilde{d}_{ij} = (a_{ij} ,b_{ij} ,c_{ij} )\). The matrix of fuzzy weights is \(\tilde{W} = [\tilde{w}_{1} ,\tilde{w}_{2} , \ldots ,\tilde{w}_{n} ]\), and the fuzzy weights are \(\tilde{w}_{j} = (w_{j1} ,w_{j2} ,w_{j3} )\).

  • Normalize the decision matrix.

In this stage, we need to transform the fuzzy decision matrix into a normalized fuzzy matrix. To achieve a normalized matrix, if a criterion is positive, equitation (2) is used, and if they are negative, equation three is used:

$$\tilde{n}_{ij} = \left( {\frac{{l_{ij} }}{{u_{j}^{*} }},\frac{{m_{ij} }}{{u_{j}^{*} }},\frac{{u_{ij} }}{{u_{j}^{*} }}} \right),\quad u_{j}^{*} = \hbox{max} \;u_{ij}$$
(2)
$$\tilde{n}_{ij} = \left( {\frac{{l_{j}^{ - } }}{{u_{ij} }},\frac{{l_{j}^{ - } }}{{m_{ij} }},\frac{{l_{j}^{ - } }}{{l_{ij} }}} \right),\quad l_{j}^{ - } = \hbox{min} \;l_{ij}$$
(3)

where the decision matrix is as follows:

$$\tilde{N} = \left[ {\begin{array}{*{20}c} {\tilde{n}_{11} } & {\tilde{n}_{12} } & \ldots & {\tilde{n}_{1n} } \\ . & . & . & . \\ . & . & . & . \\ {\tilde{n}_{m1} } & {\tilde{n}_{m2} } & \ldots & {\tilde{n}_{mn} } \\ \end{array} } \right]$$
(4)

  • Make a weighted normalized Fuzzy Matrix V.

To make a weighted matrix, we need to multiply the normalized matrix by the weight of the criteria \(v_{ij} = n_{ij} \times w_{j}\). The matrix is as follows:

$$\tilde{V} = \left[ {\begin{array}{*{20}c} {\tilde{v}_{11} } & {\tilde{v}_{12} } & \ldots & {\tilde{v}_{1n} } \\ . & . & . & . \\ . & . & . & . \\ {\tilde{v}_{m1} } & {\tilde{v}_{m2} } & \ldots & {\tilde{v}_{mn} } \\ \end{array} } \right]$$
(5)

  • Determine the positive ideal solution (\(A^{*}\)) and the negative ideal solution (\(A^{ - }\)).

$$\tilde{A}^{*} = (\tilde{v}_{1}^{*} ,\tilde{v}_{2}^{*} , \ldots ,\tilde{v}_{n}^{*} ),\quad {\text{Where}}\quad \tilde{v}_{j}^{*} = {\text{Max}}(V_{j} )$$
(6)
$$\tilde{A}^{ - } = (\tilde{v}_{1}^{ - } ,\tilde{v}_{2}^{ - } , \ldots ,\tilde{v}_{n}^{ - } )\begin{array}{*{20}c} , & {\text{Where}} & {\tilde{v}_{j}^{ - } = {\text{Min}}(V_{j} )} \\ \end{array}$$
(7)

  • Determine the sum of components distance from the positive ideal and negative ideal values.

$$d_{i}^{ + } = \sum\limits_{j = 1}^{n} {d(\tilde{v}_{ij} } ,\tilde{v}_{j}^{*} )\begin{array}{*{20}c} , & {i = 1,2, \ldots ,m} \\ \end{array}$$
(8)
$$d_{i}^{ - } = \sum\limits_{j = 1}^{n} {d(\tilde{v}_{ij} } ,\tilde{v}_{j}^{ - } )\begin{array}{*{20}c} , & {i = 1,2, \ldots ,m} \\ \end{array}$$
(9)

  • Determine the similarity to the ideal option.

The closeness coefficient (\({\text{CC}}_{i}\)) is calculated by Eq. 10. The option with a higher ranking \({\text{CC}}_{i}\) is better than the others.

$${\text{CC}}_{i} = \frac{{d_{i}^{ - } }}{{d_{i}^{ - } + d_{i}^{ + } }}\begin{array}{*{20}c} , & {i = 1,2, \ldots ,m} \\ \end{array}$$
(10)

Game theory for sustainable development-based policymaking

In the game theory, a game is an interaction or competition between several players where one player’s decision or state affects the others.

A game consists of some players, a set of strategies, and winning. Winning every game does not only depend on luck, but has its own rules and principles, and each player tries to increase its chances for winning by applying those principles (Neck 2010). To define each game, it is necessary to specify the following elements:

  • Players the parties to the game with at least two strategies.

  • Strategies of each player actions that each player can take in different stages of the game.

  • Information structure how much each player can know about their opponent’s moves and preferences at any given moment of the game.

  • Equilibrium the strategy that contains the best choice for all players is the equilibrium point of the game (or Nash equilibrium). Every game always reaches an equilibrium, which means players always come up with a solution based on their criteria. In the Nash point, the equilibrium is the condition obtained by a set of strategies and players’ decisions the deviation from which will reduce the profit.

Problem statement

Sustainable transport has many socioeconomic and environmental benefits that can accelerate local, sustainable development (Meyar-Naimi and Vaez-Zadeh 2012). In the present study, the nested sustainable development framework is considered. In this sustainable development model, priority is given to nature, human, and system, respectively (Fig. 1).

Fig. 1
figure 1

Nested sustainable development framework

Full size image

The human depends on nature; however, nature will continue without humans. Also, the system depends on both the human and the nature. The human and system are limited by nature. In the proposed nested sustainable development model, nature, human (social and economic indicators), and system are prioritized, respectively. The equilibrium point (B) of this model is calculated as follows:

$$B = {\text{Max}}\left( {\frac{{M_{\text{average}} }}{3} \times \left( {\frac{{\left| {A_{\text{N}} - A_{\text{H}} } \right|}}{{(A_{\text{N}} - A_{\text{H}} )}} + \frac{{\left| {A_{\text{H}} - A_{\text{S}} } \right|}}{{(A_{\text{H}} - A_{\text{S}} )}} + \frac{{\left| {A_{\text{N}} - A_{\text{S}} } \right|}}{{(A_{\text{N}} - A_{\text{S}} )}}} \right),0} \right)$$
(11)

where \(A_{\text{N}}\), \(A_{\text{H}}\), and \(A_{\text{S}}\) are the values of nature, human, and system aspects, respectively. \(M_{\text{average}}\) indicates the average of the aspects.

Research methodology

In the present study, a methodology based on the SD model is applied including determination of time, human and nature reference perspective, determination of indicators, normalization of indicators, weighting and aggregating the indicators firstly to build up environmental, social, economic, and technical sub-dimensions and then to prioritize the main aspects, i.e., nature, human, and system. There are three types of electric vehicles on the market, including battery electric vehicle (BEV), hybrid electric vehicle (HEV), and plug-in hybrid electric vehicle (PHEV). In this paper, PHEV is selected for the study. The most important reason to consider PHEVs in this study is the lack of adequate charging stations for them in Tehran.

Determination of indicators

This section represents the effective nature (environmental), human (economic and social), and system (technical) indicators in the development of vehicles in Tehran from 2002 to 2018, which are illustrated in Table 1.

Table 1 Sustainable development indicators on the adoption of vehicles
Full size table

To that end, the weight of the normalized indicators was calculated using the analytical network process (ANP) to sort and aggregate them. In this part, the weights of the aspects, sub-aspects, and indicators of each sub-aspect were calculated using ANP. In the ANP method, the system depends on nature and human. For example, the technical development of new vehicles is dependent on environmental, economic, and social aspects, and human depends on nature.

Sustainability evaluation of vehicles development in Tehran from 2002 to 2018

In this section, the historical development of conventional vehicles (gasoline, CNG, diesel engine) and PHEVs are evaluated from the sustainability point of view. Figures 2, 3, and 4 represent the temporal trend of environmental, economic, social, and technical indicators and their aggregated values from 2002 to 2018 in Tehran.

Fig. 2
figure 2

Temporal trend of aggregated nature indicators

Full size image
Fig. 3
figure 3

Temporal trend of aggregated human indicators

Full size image
Fig. 4
figure 4

Temporal trend of aggregated system indicators

Full size image

The temporal trend of aggregated nature indicators from 2002 to 2018 is shown in Fig. 2. As shown, most of the nature indicators had little growth between 2006 and 2014, while since 2014, most of them have been increasing.

Human indicators are shown in Fig. 3. As it is seen, H1, H2, H3, H4, and H8 have had an increasing trend, whereas H5, H6, H7, and H9 indicators have had a decreasing trend.

System indicators are shown in Fig. 4. As indicated, S1, S2, S5, and S6 at first have been incremental, then decreasing, and then incremental again. S3 and S4 are related to V2G service at the charging station, which has been implemented since 2014. The rest of the technical indicators have been increasing from 2002 to 2018.

The temporal trend of aggregated nature, human, and system aspects is shown in Fig. 5. As reflected in the figure, the human and system aspects remained unchanged in 2002, and the nature aspect fell to its lowest value. In the years 2010–2016, nature ranked higher than the others, and then fell human and system aspects. From 2016, the growth of the system aspect has been more than the others. In this figure, B is the equilibrium point of the nested model. As seen in the figure, from 2006 to 2016, development was sustainable, but it was unstable from 2002 to 2006 and 2016 to 2018.

Fig. 5
figure 5

Temporal trend of aggregated aspects and equilibrium point (B)

Full size image

In Fig. 6, the percentage of gasoline, CNG, diesel, and plug-in hybrid electric vehicles in Tehran from 2002 to 2018 is shown. Based on the figure, since 2008, the CNG vehicles have outnumbered the diesel vehicles. Plug-in hybrid electric vehicles have been used since 2014.

Fig. 6
figure 6

Adoption of gasoline, CNG, diesel, and PHEV in Tehran from 2002 to 2018

Full size image

Sustainable development-based environmental policymaking

In this section, a new approach to PHEV policymaking from the perspective of sustainable development in Tehran is presented. The fuzzy TOPSIS method is used to investigate various factors affecting the development of PHEVs in Tehran. Based on Fig. 5, at the equilibrium point of the nested SD model (B), it can be seen that the development of vehicles in Tehran has been unsustainable since 2016, which confirms the current Tehran air pollution. According to the World Health Organization (WHO), Tehran is one of the most polluted cities in the world in terms of particulate matter (Fig. 7).

Fig. 7
figure 7

Annual ambient level of PM10 (μg/m3)

Full size image

Alternative policy scenarios

In this section, some policy scenarios for the development of PHEV in Tehran are proposed based on the sustainable development indicators (Table 2).

Table 2 Policy description for PHEV development in Tehran
Full size table

Policymaking using fuzzy TOPSIS method

In line with the fuzzy TOPSIS steps, four different criteria, including environmental, social, economic, and technical, are considered in this research. There are also eight alternative policy scenarios Pi, i = {1, 2, …, 8} concerning Tehran’s transportation decisions, regulations, and directives. The criteria weights were assigned to the alternatives for the assessment of different policy scenarios (Fig. 8). The fuzzy variables for the criteria and their weights are shown in Fig. 9. The ratings of alternative policy strategies by decision-makers and the weight of criteria using fuzzy variables are illustrated in Table 3. The distance from the positive and negative ideal values (\(d_{i}^{ + } ,d_{i}^{ - }\)) and the similarity to the ideal option (\({\text{CC}}_{i}\)) from the state and vehicle buyer viewpoints are shown in Tables 4, 5, 6, 7, 8, and 9.

Fig. 8
figure 8

Hierarchical structure for assessing the policy strategies in Tehran

Full size image
Fig. 9
figure 9

Fuzzy variables

Full size image
Table 3 Rating of alternative policy strategies by decision-makers
Full size table
Table 4 Distance from the positive ideal value from the state viewpoint
Full size table
Table 5 Distance from the negative ideal value from the state viewpoint
Full size table
Table 6 Similarity to the ideal option (\({\text{CC}}_{i}\))
Full size table
Table 7 Distance from the positive ideal value from the vehicle buyer viewpoint
Full size table
Table 8 Distance from the negative ideal value from the vehicle buyer viewpoint
Full size table
Table 9 Similarity to the ideal option (\({\text{CC}}_{i}\))
Full size table

The state viewpoint

The fuzzy TOPSIS step from the state viewpoint is shown in Tables 4, 5, and 6. After applying the fuzzy TOPSIS method, the resulted ranking of alternative policy scenarios from the state viewpoint using the closeness coefficient is: \(P_{1} > P_{8} > P_{6} > P_{2} > P_{7} > P_{5} > P_{3} > P_{4}\).

Vehicle buyer viewpoint

The fuzzy TOPSIS step from the vehicle buyer viewpoint is illustrated in Tables 78, and 9. Based on the fuzzy TOPSIS method (Table 9), the resulted ranking of alternative policy scenarios from vehicle buyer viewpoint is: \(P_{6} > P_{3} > P_{4} > P_{7} > P_{8} > P_{5} > P_{1} > P_{2}\).

Sensitivity analysis and game equilibrium point for PHEV policymaking

Sensitivity analysis is a tool for analyzing uncertainty in problems with different decision-makers. Sensitivity analysis helps understand how outputs change with input. Table 10 shows the different weights of the input criteria in the sensitivity analysis. Figure 10 shows the results of the sensitivity analysis in buyers and the state points of view (players in the game theory method). In game theory, each player’s profit in different strategies is first calculated. Then, the Nash equilibrium point is calculated. This point represents the highest profit each player has for themselves in facing other players, so ignoring that the point can reduce their profits as well as others. Table 11 shows the closeness coefficients of the policy scenarios obtained using the Nash equilibrium point.

Table 10 Weights of the criteria in a sensitivity analysis
Full size table
Fig. 10
figure 10

Sensitivity analysis of PHEV buyers and the state point of views

Full size image
Table 11 Closeness coefficient of the obtained policy scenario by Nash equilibrium point
Full size table

Results and discussion

There are several reasons why PHEV is not popular in Tehran. One is that the price of PHEVs is relatively higher than that of CVs because of their inadequate technology. The second reason is that the required charging infrastructure does not yet exist. Therefore, incentives should be given to developing charging stations. It seems that the suggestions proposed in this paper can help increase the number of PHEVs in Tehran.

One of the selected policy indicators for PHEV adoption is providing non-financial incentives such as eliminating traffic constraints in Tehran downtown (P6). To increase the use of PHEVs in Tehran, there must be conditions that encourage citizens to use these vehicles. Such incentives can be both financial and non-financial. Financial incentives that can persuade consumers to buy PHEVs include direct loans, removing taxes and annual charges, discounts on commercial profits and PHEV tariffs, and domestic electrical systems in times other than the on-peak hours. Non-financial incentives can also affect PHEV buyers’ interests. Non-financial incentives can include eliminating traffic constraints and allocating dedicated parking spaces to PHEVs.

Another policy indicator is the peak load shaving with load shifting (P7). To describe this policy, we propose a new concept called time–location shifting (TLS) implemented by PHEV. TLS means power consumption by charging PHEV batteries during off-peak hours (during business hours and in the downtown) and by generating power during on-peak hours by battery discharge (during the evening and night and in the suburbs). The average urban journey duration in Tehran is about 55 min (Samaie et al. 2020). Most people in Tehran are in downtown from 8 am to 5 pm, where there are many commercial and administrative buildings. In the afternoon, they return home from 5 pm. Applying the concept of TLS, power losses are effectively reduced, and as a result, emissions from the power plants supplying these losses are lessened.

Using the TLS scenario will expand the use of PHEV due to being profitable for the vehicle owner as well as the distribution network. To use PHEVs in the network, the time tariff policy can be suggested as given in Table 12. Using the TLS method, PHEVs can be recharged at off-peak hours in the downtown and inject battery power into the grid in on-peak hours in the suburbs, thereby reducing the peak.

Table 12 Time tariff policy based on Tehran load profile
Full size table

Predicted number of PHEVs based on policy scenario from game equilibrium

The growth rate of PHEV in Tehran can be calculated as follows:

$${\text{PHEV}}_{\text{Number}} (t) = {\text{PHEV}}_{{{\text{Number}}@2018}} \times (1 + {\text{CC}}_{\text{equilibrium}} )^{(t - 2018)}$$
(12)

where \({\text{CC}}_{\text{equilibrium}}\) is the amount of average closeness coefficient in the game equilibrium point, \({\text{PHEV}}_{\text{Number}} (t)\) is the PHEV number at year t, and \({\text{PHEN}}_{{{\text{Number}}@2018}}\) is the PHEV number in the year 2018.

In 2018, there were approximately 11,000 PHEVs in Tehran. As the closeness coefficient of the game equilibrium point is 0.4582 for the state player and 0.5591 for the buyer of PHEV player, the average of which is 0.50865. It can be predicted that there will be 3,211,221 PHEVs in 2032.

As shown in Fig. 11a, the transportation sector accounts for about 25% of the total energy consumption in Tehran. Gasoline and natural gas vehicles use 86.44% and 13.5% of energy in the transportation system, respectively. However, the contribution of PHEVs is very small.

Fig. 11
figure 11

a Current energy consumption chart of Tehran in 2018. b Energy consumption forecast of Tehran in 2032

Full size image

As predicted above, PHEVs will account for about 20% of vehicles by 2032. Therefore, they can reduce gasoline consumption and increase electricity consumption, where the electricity consumption of the transportation sector will be about 20%. Based on this scenario, the energy consumption chart for 2032 is shown in Fig. 11b.

Conclusion

Systems’ thinking is based on a worldview in which each large system consists of smaller sub-systems. The transportation system is a set of components that pursue specific goals. So, there is a need for a model that can tell whether the transportation sector development is doing well and what technology can improve the technical, environmental, economic, and social aspects of development. Sustainable development, which includes all aspects of development, can be chosen as a framework for this purpose.

In this study, a new approach to environmental policymaking was presented for sustainable development in Tehran. In this approach, a combination of the ANP, fuzzy TOPSIS, and game theory analysis is used to increase the penetration of PHEVs in Tehran. In so doing, first, the ANP approach was applied to the sustainability assessment of different vehicles in Tehran from 2002 to 2018. And then, some policy scenarios for the development of PHEV in Tehran were proposed, and the fuzzy TOPSIS method was used to rank them. The policies were examined from the point of view of the vehicle buyers and the state, and then using game theory, taking into account the vehicle buyer and the state as the players, the Nash equilibrium point was determined, which yielded the best policy in the point of view of both players. This could help to gain greater knowledge and more understanding of the PHEV development problem, which is the ideal basis for well-founded decision-making. This method can provide for the evaluation of the alternatives, giving important insights into the most suitable policy scenarios for achieving Tehran’s PHEV target by 2032.