Multi-Criteria Evaluation and Selection of Renewable Energy Battery Energy Storage System-A Case Study of Tibet, China

Renewable energy sources such as solar energy and wind energy are characterized by intermittency and volatility due to their over-dependence on weather conditions. Therefore, it is especially important for the power system containing renewable energy to equip the energy storage system which can not only guarantee the flexibility of renewable energy utilization, but also improve the reliability of system power supply. Because the Battery Energy Storage System (BESS) is suitable for mass production and large-scale applications, it has become the main energy storage system scheme for the power system. Because different BESS have differences in efficiency of storage, storage capacity, discharge ability and maintenance, it is necessary to make a comprehensive evaluation and selection of the type of BESS for power system. In this paper, a comprehensive evaluation index system of BESS is established by taking the photovoltaic power station in Xizang region as an example. However, the existing methods are difficult to accurately measure the attribute values of each indicator and the correlation between experts. Therefore, this paper proposes the intuitionistic uncertain language Choquet ordered weighted aggregation operator (IULCWA) by combining intuitionistic uncertain language with fuzzy measure and Choquet integral, and establishes a fuzzy multi-criteria decision-making (MCDM) method for BESS selection. Comparison and analysis also prove that the modified model has good scientificity and operability. This study can provide a new theoretical basis for the selection of energy storage schemes for new energy batteries, and expand the application scope of fuzzy MCDM method.


I. INTRODUCTION
Under the global environmental pollution and energy crisis, renewable energy resources such as wind, solar and others play an important role in the future energy pattern because of their advantages of wide distribution, large scale and no pollution. Due to the excessive dependence of renewable energy on weather conditions and it has obvious intermittency and volatility, which makes the power system containing renewable energy resources very difficult in stabilizing power generation output, enhancing adjustability and improving grid consumption [1], [2]. With the gradual increase The associate editor coordinating the review of this manuscript and approving it for publication was Reinaldo Tonkoski . of penetration of renewable energy resources, this situation is more obvious. Therefore, it is particularly important for the power system containing renewable energy to equip a reasonable energy storage system that can guarantee the flexibility of renewable energy utilization and the reliability of system power supply. In power system, energy storage system realizes energy storage by charging and discharging power of specific energy storage medium. In the low period of electricity consumption, excess electric energy is stored in the energy storage medium in some form. In the peak period of electricity demand, the energy stored in the energy storage medium is released in the form of electric energy in order to achieve the effect of grain cutting and peak filling, which enables the production process, supply process and consumption process VOLUME 9, 2021 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ of electricity to be carried out in different time periods. The power system equipped with energy storage system not only meets the user's demand for electricity, but also improves the power generation efficiency of power plants.
Common energy storage systems include pumped hydroelectric storage system, compressed air energy storage system, BESS, etc. Because of the advantages of flexible configuration, fast response speed and no geographical resource constraints, BESS has become an important part of energy storage system [3], [4]. Among BESS, lead-acid battery, lithium-ion battery and sodium-sulfur battery occupy the main types of BESS [5]. Since Gaston Planté has invented the Lead-acid battery, BESS has been widely used in various energy storage projects [6]. In 2017, the installation of the world's largest lithium-ion battery (100MW, 129MWh from Tesla and Neoen) has completed in South Australia [7]. A 100MW lithium-ion battery storage array built by the US government in California is also coming into service [3]. In China, the new energy sector is booming. In order to improve the economy of the BESS, Bai et al. [8] proposed to use the decommissioned reusable batteries of electric vehicles as the energy storage system of distributed solar photovoltaic power generation.
At present, the basic research of energy storage system mostly focuses on the basic principles, energy efficiency indexes and research status of energy storage system. Luo et al. [9] outlined six types of energy storage systems and their operation principles, technical and economic performance features, and key points of current research. And they introduced Lead-acid batteries, Lithium batteries, Sodium-sulfur batteries, Nickel-cadmium batteries and other energy storage batteries in BESS and provided the data of energy density, power density, specific energy, specific power, power rating and rated energy capacity of each energy storage batteries. By combining the cost characteristics, they provided decision-making basis for investors. Guney and Tepe [10] divided energy storage systems into Chemical energy storage systems, Electrochemical energy storage systems, Electric energy storage systems, Mechanical energy storage systems and Thermal energy storage systems. They summarized the basic principles of various energy storage systems, and evaluated all kinds of energy storage batteries from rated power, energy, self-discharge, power output, charge/discharge energy efficiency, cycle life and so on. Guo et al. [11] introduced the current research status of the battery energy storage technology, the grid technology and the large-scale energy storage technology. Based on the preliminary analysis of several key technologies of large-scale BESS, the standard system construction of energy storage system was proposed. And they forecasted the development of energy storage system.
In the technical research of energy storage system, some researchers put forward the optimal allocation schemes and control strategies for energy storage system, in order to give full play to the best performance of energy storage system. Aiming at the optimal size of the battery in the BESS, Yuqing et al. [12] provided the criteria and methods for evaluating battery size in energy storage system and their applications in various renewable energy systems, to clarify the key index of the battery and to achieve the balance between the technical improvements and the additional cost. Jannesar et al. [13] proposed an optimal configuration considering the location, size and daily charge/discharge performance of batteries in the power distribution network of photovoltaic power generation with high penetration. The fund and maintenance cost of BESS were calculated in terms of energy arbitrage, environmental emission, energy losses, transmission access fee and maintenance. The simulation results were carried out in DigSilent simulation software. The results showed that this method can effectively reduce overvoltage and energy losses, prevent reverse power flow, reduce environmental emissions and maximize economic benefits. Xiong et al. [14] proposed a Hybrid Energy Storage System with battery and ultra-capacitor to meet the needs of high specific power and high specific energy simultaneously in single energy storage system. On the basis of studying the structures and energy control management strategies of this system, specific application examples are given to verify its effectiveness. Hosseinimehr et al. [15] proposed a cooperative control method for battery energy storage units (BES units) based on microgrid and provided an accuratecooperative control strategy provides accurate reactive power sharing among the BES units. In the charging operation of the BES units, proportional-integral controller is adopted to limit the absorbing power of BES units and match its available surplus power. Extensive simulations carried out in power systems computer aided design/Electromagnetic Transients including DC verified its performance.
In order to make the energy storage system serve the power system more efficiently, some scholars have made a comparative study of various energy storage systems in order to select the appropriate BESS according to the practical application requirements. Poullikkas [16] outlined different types of batteries for large-scale electrical storage, especially the current large-scale BESS in operation and compared them with other types of large-scale energy storage systems. The analysis shows that Sodium-sulfur batteries are suitable for large-scale energy storage systems, while the flow batteries, especially the vanadium redox flow batteries are suitable for small-scale energy storage systems. Zhang et al. [3] introduced several types of battery systems, compared the material, electrochemistry performance and cost of batteries, and pointed out that the performance and capacity of large-scale energy storage systems depended on battery and power condition. Nikolaidis and Poullikkas introduced the working principle, operating temperature, voltage, cycling times, storage capacity and application of seven types of energy storage batteries. A comprehensive analysis was conducted according to the mechanical technologies and electrochemical technologies capital cost and performance data that were used in each energy storage system to evaluate the feasibility of these systems [17].
Because different BESS have different performance in efficiency of storage, storage capacity, discharge ability and environmental impact, and the degree of flexibility to solve the problem of power grid absorption and to enhance renewable energy generation is also different. Therefore, what kind of BESS should be equipped for power system needs to be comprehensively considered from multiple factors. Some scholars have studied the selection of energy storage system and achieved some results. Dhundhara et al. [18] used the traditional Economic-technical analysis method to evaluate comprehensively the energy storage systems of Lead-acid batteries and Lithium-ion batteries. However, in practical selection, the selection of BESS is a multi-attribute problem. Traditional economic and technological analysis cannot fully reflect the actual situation of BESS because of its one-sidedness. Therefore, the multi-attribute evaluation method should be adopted to evaluate and select BESS. Olabi [19] proposed multi-attribute selection criteria for energy storage systems, including the available energy resource, energy requirement and application, energy storage efficiency, energy storage cost and energy storage infrastructure. Li et al. [20] constructed a decision-support framework to solve the multi-objective optimization decision-making problem of energy storage technology, by considering three objectives of minimizing cost, maximizing technical suitability and minimizing environmental impact to select the energy storage technology. However, in evaluating these factors, due to the complexity of decision-making and the limitations of decision-makers' cognition, it is impossible to provide accurate data to support decision-making, which makes the decision-making process defective. In addition, reinforcement learning approach can also be used to handle the complex multi-objective optimization problem. Yinliang et al. [21] proposed a fully distributed multiagent-based reinforcement learning method by integrating consensus-based information discovery algorithm and distributed Q-learning algorithm for optimal reactive power. Yang et al. [22] proposed an improved deep deterministic policy gradient algorithm using prioritized experience replay mechanism and L2 regularization, so as to improve the policy quality and learning efficiency of the dispatch strategy.
In group decision-making, because the preferences of experts are influenced by their social status, power, knowledge and other factors, the personal preferences of decision makers are not independent, but cross-related with certain redundancy or complementarity, which leads to the additivity was not valid when aggregating comprehensive values given by experts. In view of the relevance relationship between attributes, Choquet, a French mathematician, put forward the Theory of capacities and defined the Choquet Integral based on the Theory of capacities. This is the earliest systematic study of Non-additive measure [23]. Subsequently, Sugeno, a Japanese scholar, proposed the concept of Fuzzy Measure in order to solve the multiple criteria decision problem in which the indexes are related but without additivity [24].
Murofushi and Sugeno [25] gave a clear explanation of the Fuzzy Measure, pointed out that the Choquet Integral of the Fuzzy measure was reasonable and gave a concise expression, which laid the foundation for its application. Grabisch [26] introduced the Fuzzy measure to deal with the relevance between attributes. Based on Choquet Integral aggregating operators, the multi-attribute decision making method was studied and applied to solve the decision problem. In 2000, Marichal [27] gave the axiom that discrete Choquet integral can handle attribute relevance.
In recent years, Choquet integral and Fuzzy Measure have been widely used in the field of multi-attribute decision making and have achieved fruitful results. In the method, Meyer and Roubens [28] applied Choquet integral and fuzzy numbers to multi-criteria decision-making, and proposed a method of combining Triangular Fuzzy Numbers, Trapezoidal Fuzzy Numbers with Choquet integral. Cui and Li [29] proposed a multi-attribute group decision making method based on Linguistic Variable Choquet Integral Operators; Other scholars proposed an Intuitionistic Fuzzy Sets Choquet Integral operators (IFCI) and its decision-making methods [30]- [32]. Qin and Liu [33] proposed Interval-Numbers Intuitionistic Fuzzy Number Choquet Integral Operators. Ferreira et al. [34] defined Iterval-Numbers Intuitionistic Hesitant Fuzzy Number Choquet Integral Operators for solving multi-attribute decision making problems. In application, Ashayeri et al. [35] used IFCI to solve the selection and allocation of supply chain partners. Büyüközkan et al. [36] used IFCI to solve the decision-making problem of sustainable urban transportation alternatives. Demirel et al. [37] used Hesitant Fuzzy Numbers Choquet Integral Operators to evaluate and select four schemes for preventing soil erosion in order to solve the soil erosion in Turkey. Demirel et al. [38] used Interval Fuzzy Number Choquet Integral Operators to evaluate various schemes for building underground natural gas storage facilities alternatives in order to ensure the reliable supply of natural gas in peak period.
In this paper, a group decision-making method based on fuzzy measure and Choquet integral is proposed. The traditional logic weighted method is developed into Integral synthesis method, which is introduced into the comprehensive evaluation of BESS. On the basis of dealing with uncertain and fuzzy information, the attributes and attribute sets are measured by using fuzzy measure, and the monotonicity of Choquet integral is used to replace the additivity. The intuitionistic uncertain language Choquet ordered weighted aggregation operator (IULCWA) is proposed to measure the evaluation value of the alternatives and solve the problem of relevance between index system and experts, which provides an effective tool for making decision and makes the decision-making process and results more scientific.
The main contributions of this paper are as follows: (1) The characteristics of three major BESS are studied, and a qualitative indexes evaluation system with relevance is constructed from the aspects of economy, sociality, technology and maintainability; (2) On the basis of the existing concepts of Intuitionistic Fuzzy Number and Uncertain Linguistic, the concept of Intuitionistic Uncertain Linguistic is defined, and the operational laws and comparison method of Intuitionistic Uncertain Linguistic Numbers are proposed; (3) Fuzzy measure is used to measure attributes and attribute sets, monotonicity of Choquet integral is used to replace additivity. Proposing IULCWA to measure the evaluation value of the alternatives, the problem of correlation between the attributes and experts was solved. And on this basis, a group-decision making method is proposed for the selection of BESS. The effectiveness of this group-decision making method is verified by example application and comparison with other two methods.
The following structure and content of this paper are as follows. Section 2 introduces the solar energy situation in Ali area of Tibet and the current mainstream three kinds of battery energy storage technology, and determines the research route of this paper. In Section 3, the BESS comprehensive evaluation indexes system is constructed. And the measurement elements of each index are explained. And on the basis of the original theory, a comprehensive evaluation of alternative methods is proposed based on the IULCWA. In Section 4, the comprehensive evaluation model proposed above is used to rank the three kinds of BESS, so as to provide support for selecting the appropriate BESS for a solar power plant in the region. After that, IULCWA is compared with the other two conventional methods, and the results are discussed and analyzed. In the last section, the above mentioned content is summarized and the future research is prospected.

II. RELATED WORKS A. STUDY AREA
Tibet is located in the southwest of Qinghai-Tibet Plateau in China, between 26 • 50 N and 36 • 53 N, and 78 • 25 E and 99 • 06 E. The whole region covers an area of 1,202,230 square kilometers, accounting for about 1/8 of the total area of the whole country, which is second only to Xinjiang among all provinces, municipalities and autonomous regions in China. With an average altitude of over 4,000 meters, it is known as the ''roof of the world''. At the end of 2012, the total resident population of the whole region was 3.08 million, which governed 4 prefecture-level cities, 3 regions, 4 municipal districts and 72 counties. Tibet is the region with the richest solar energy resources in China. The solar energy radiation in most parts of the region reaches 6,000 ∼ 8,000 MJ/per year, which is about twice as high as that in plain areas at the same latitude. Among them, Ali has become the most potential area for developing photovoltaic power generation in Tibet due to its largest area and highest average altitude. In recent years, Tibet has vigorously developed photovoltaic industry. By the beginning of 2019, the total installed capacity of power generation in Tibet Autonomous Region was 2,357,500 kW. Among them, the installed capacity of hydropower is 1.56 million kW, the installed capacity of photovoltaic is 790,000 kW, and the installed capacity of wind power is 7,500 kW. There are 14 photovoltaic power plants currently in use. The total installed capacity of photovoltaic power generation in the new energy field has accounted for more than 30% of the total installed capacity of power generation in Tibet Autonomous Region. The distribution of solar energy resources in Tibet is shown in Figure 1.

B. CONSTRUCTION OF COMPREHENSIVE EVALUATION SYSTEM FOR BESS
According to the research of related literature, this paper proposes a battery energy storage mode selection model based on the IULCWA. The decision-making process includes dealing with fuzzy evaluation indicators, and considering the interaction between indicators and evaluation experts. In order to show the decision framework intuitively, it is necessary to express the logical structure of this paper by using the network flow chart after determining the goals and methods. Figure 2 shows the specific steps of selecting alternative battery energy storage technologies for a photovoltaic power station in Ali, Tibet.

C. CONSTRUCTION OF COMPREHENSIVE EVALUATION SYSTEM FOR BESS
BESS is one of the most commonly used energy storage systems in power system. BESS usually connect one group or more groups of battery units in series or in parallel. Each battery consists of two conductor electrodes and an electrolyte, which are combined in a special, sealed container and connected to the outside to obtain the required voltage and capacity [39]. At present, Lead-acid BESS, sodium sulfur BESS and lithium-ion BESS are widely used.

1) LEAD-ACID BATTERIES ENERGY STORAGE SYSTEM, LEAD-ACID BESS
Lead-acid BESS is an energy storage system with lead-acid battery as the main equipment, and it is the earliest commercial BESS [40]. The electrodes of battery are mainly made of lead and its oxides, and the electrolyte is a solution of sulfuric acid. In discharging state of lead-acid battery, lead dioxide is the main component of the positive electrode, and lead is  [41]. Its working principle is expressed by chemical reaction equation as follows [42]: The negative electrode: Pb + SO 2− 4 ⇔ PbSO 4 + 2e − The positive electrode: Lead-acid batteries have advantages of low price, high specific power, safe operation and high regeneration rate [40]. The market share of lead-acid batteries is as high as 30%. But there are also some problems, such as low specific energy, short cycling life, and massive gas, heavy metal pollution caused by self-discharge [3]. In recent years, lead-carbon super battery using carbon as active material carrier of leadacid batteries is an improvement of traditional lead-acid batteries. Its specific energy and specific power have been greatly improved. Its structure is equivalent to parallel connection of a double layer capacitor with traditional lead-acid batteries, which makes the lead-carbon super battery possess both the high energy of the traditional lead-acid battery and the high specific power of the capacitor. Because carbon can act as buffer and share charge/discharge current with lead negative electrode, especially when high-magnification current is charged/discharged, carbon in composite negative plates firstly responds, which can slow down the impact of high current on the lead negative plates and significantly improve the service life of batteries (> 5000 times) [39], [43]. Projects for worldwide application of lead-acid BESS are shown in Table 1 [44].

2) SODIUM-SULFUR BATTERIES ENERGY STORAGE SYSTEM, NAS BESS
NaS batteries energy storage system is an energy storage system with sodium-sulfur battery as the main energy storage equipment. Sodium sulfur battery was invented by the Ford Company in 1967 [45]. It was originally designed for electric vehicles and then developed into energy storage field. Sodium-sulfur batteries adopt tubular design, with the metal sodium as the negative electrode in the center, the inner tube (β-Al2O3 ceramic tube) is the electrolyte membrane, at the same time, it plays the role of storing sodium metal. And the outer tube is the composite material or stainless-steel metal material, is used to hold the non-metallic sulfur of the positive electrode. At a certain working temperature (above 290 • C), sodium ions react can occur a reversible reaction through the electrolyte diaphragm and sulfur to reach the energy release and storage [40]. Its working principle is expressed by chemical reaction equation as follows [46]: The negative electrode: The positive electrode: The overall reaction: 2Na + xS ⇔ Na 2 S x Sodium sulfur batteries have very high specific energy and specific power. They can achieve high current, high power discharging, high energy storage and conversion efficiency, no self-discharge phenomenon, long cycle life, and they are environmentally friendly, small in size, compact in structure and light in weight. Texas built the world's largest sodium-sulfur battery, which can provide 4 MW of power for up to eight hours when the city's lone line to the Texas power grid goes down [47]. Through combination and series connection, Nas batteries can reach the megawatt level and can be directly used in large-scale energy storage. Abu Dhabi boasts the world's largest NaS battery battery energy storage station (108 MW/648 MWh) [48]. However, sodium-sulfur batteries need to work at temperatures above 290 • C, which puts forward higher requirements for the stability of the VOLUME 9, 2021 electrode material, and there are safety problems in high temperature operation [40], [41]. At the same time, the corrosion resistance of ceramic membrane and sulfur electrode needs to be further improved, and the corresponding high cost is also an important factor restricting their development. Projects for worldwide application of sodium sulfur BESS are shown in Table 2 below [44].

3) LITHIUM-ION BATTERIES ENERGY STORAGE SYSTEM, LITHIUM-ION BESS
Lithium-ion batteries energy storage system is an energy storage system with lithium-ion battery as the main energy storage equipment. When lithium ion batteries are charged, lithium ions are removed from the positive electrode and embedded into the negative electrode through electrolyte and membrane. On the contrary, when the battery discharges, lithium ions are removed from the negative electrode and re-embedded into the positive electrode through electrolyte and diaphragm [9]. Its working principle mainly relies on the intercalation and removal of lithium ions between positive materials (metal oxide) and negative materials (graphite) to achieve energy storage and release [42]. The chemical reaction equation is expressed as follows [49]: The negative electrode: The positive electrode: The overall reaction: Lithium-ion batteries have many advantages, such as high output voltage, high specific energy, high specific power, high charging and discharging efficiency, long cycling life, small self-discharge rate and environment-friendly [3], [38].
However, when applied to large capacity energy storage, lithium-ion BESS faces the problems of battery safety and cost [40]. Protective circuit must be installed in the system to prevent overcharging or over discharging of batteries, which also increases the cost [50]. The safety of Lithium-ion BESS will be improved by using various safe electrode materials and reasonably designing internal and external safety protection measures and battery structure. At the same time, with the development of material preparation technology and the improvement of battery preparation technology, the cost of lithium-ion batteries is expected to be further reduced, which will promote the gradual expansion of lithium-ion batteries to high-power systems such as electric vehicles and large-scale energy storage batteries, and lithium-ion batteries may become a leader in the field of energy storage. Projects worldwide for the application of lithium-ion BESS are shown in Table 3 [44].

4) COMPARISON OF ADVANTAGES AND DISADVANTAGES OF BESS
Through the above analysis, the advantages and disadvantages of all kinds of batteries are summarized as shown in Table 4.

A. THE BESS COMPREHENSIVE EVALUATION INDEXES SYSTEM
Based on the principles of comprehensiveness, operability and objectivity in comprehensive evaluation, this paper constructs the BESS comprehensive evaluation indexes system from four aspects: economy, sociality, technology and maintainability according to the actual situation of BESS and relevant literature.

1) ECONOMY
BESS, as a supporting technology for the development of new energy, it has not yet fully entered the stage of commercial development due to its high cost and imperfect technology level. The government, investors, society and research institutes all pay close attention to the economic issues of BESS development at this stage. Economy of BESS refers to the proportional relationship between system cost and benefit in the whole process of early design and construction and later operation and maintenance. The cost of BESS mainly includes the feasibility study in the early stage and the cost in the design process, the cost of raw materials, equipment, human resources and installation in the construction stage, and the operation and maintenance cost in later period. The benefits of BESS are mainly based on the regional electricity consumption, the discount of peak electricity price and the difference of peak and valley electricity price. Because of the different performance, capacity and size of the BESS, obviously, it is impossible to only compare the cost or benefit, so this paper regards the cost-benefit ratio of the system as an index to measure the economy of BESS, and evaluates the economy of BESS according to the ratio between the cost and benefit of all kinds of batteries.
At the same time, the economy of BESS is also affected by technology and maintainability. Some energy storage batteries need more cost to purchase equipment, install and debug because of complicated technologies or scarcity of raw materials and high manufacturing cost, which leads to poor economy. BESS with low maintainability will consume more human resource cost and maintenance cost, and because of its low maintainability, it is difficult to solve failures and to ensure higher benefit.

2) SOCIALITY
The development of science and technology will inevitably lead to the change of social benefits and ecological environment. BESS, as a new technology, its sociality has also been widely concerned.
The sociality of BESS is an achievement of technological development. The impact of BESS on all aspects of society mainly refers to the impact of the ecological environment and living environment. In the design stage of BESS, it is necessary to consider the impact of the material of energy storage battery on the ecological environment, such as lead-acid batteries, nickel cadmium batteries and so on, which may cause serious pollution to the ecological environment in the process of manufacture, operation and recovery. In the construction process, the damage to the surrounding ecological environment should also be considered. For example, construction waste and sewage discharge and so on are one of the factors to evaluate the sociality of BESS. At the same time, the construction and operation of BESS in cities or densely populated areas will also have an impact on the lives of the surrounding residents. On the one hand, the sound pollution, light pollution and electromagnetic radiation caused by the operation of BESS will have a certain impact on people's lives and safety. On the other hand, the BESS brings convenience for people to use electricity, as well as the impact on government revenue and finance.

3) TECHNOLOGY
Technical evaluation always plays a key role in the evaluation of scientific and technological achievements. Technology is related to all aspects of scientific and technological achievements, and has an important impact on economy, sociality and some other aspects.
The technicality of BESS refers to the performance of the battery and the advanced degree of technology. It is a comprehensive evaluation of the technology used in the whole system. Technical evaluation is an important basis for BESS evaluation, and the performance of BESS is an important basis for technical evaluation of BESS. The performance degree of energy storage battery is evaluated in terms of cycling life, energy storage efficiency, specific capacity and energy density. Cycling life refers to the number of charge-discharge cycles that energy storage battery can carry out under a certain capacity. Energy storage efficiency refers to the ratio of energy stored by energy storage battery elements to input energy. Specific capacity refers to the energy released by energy storage battery of unit mass. Energy density refers to the energy contained in the unit volume of energy storage battery. At the same time, according to the stability and reliability of the system, the technical advanced degree of BESS is evaluated. Stability refers to the ability of the system to perform certain stable state under external influence. Reliability refers to the ability of the system to complete specified functions within a specified time and under specified conditions.
The technical level of BESS has a great impact on the maintainability and economy of the system. The better the performance of energy storage battery and the more advanced the technology, the stronger the continuous operation ability and the higher the maintainability of the energy storage system. However, in the early stage of construction, the cost of equipment may higher and the economy is slightly worse. VOLUME 9, 2021 4) MAINTAINABILITY After any system is put into operation, later maintenance is indispensable. Maintainability of BESS refers to the difficulty of repairability and improvability of system. Repairability refers to the possibility of eliminating or restraining malfunction after system failure, repairing it and returning to original state of normal operation. It mainly evaluates the difficulty of BESS in implementing preventive and corrective maintenance function, including fault detection, diagnosis, repair and whether it can be re-initialized. Improvability is the possibility of accepting the improvement of the existing functions and adding new functions to the BESS. It is a measure of the difficulty of accepting the improvement of the BESS and even making functional modifications to further adapt to the outside world or new environment.
Maintainability of BESS is affected by technology. High technology system has high maintainability. In low technology system, there may be some problems, such as lack of spare parts and inability to add new functions, which leads to low maintainability.
µ is called as a fuzzy measure in P (X ).
From the perspective of multi-criteria decision-making, µ can be seen as the importance of one or more attributes in P (X ).
The determination of attributes and attributes set fuzzy measures is the key to solve practical decision-making problems by using fuzzy integral. There are many algorithms of fuzzy measures. In order to reduce the complexity of solving and calculating fuzzy measures and reflect the interaction between evaluation indexes, Sugeno proposed a fuzzy measure λ.
If X is an attribute set of a multiple attribute decision making problem, ∀A, B ∈ P (X ) , µ (A) and µ (B) can be regarded as the weights of attribute set.
Suppose X is a finite set, ∪ n i=1 x i = X , and x i ∩ x j = φ, i, j = 1, 2, · · · , n, i = j, then As can be seen from Formula (2), for any A ∈ P (X ), there are Because µ (X ) = 1, when the value µ x j is known, the following formula can be used to calculate the value λ.
When, n j=1 µ x j = 1, λ = 0, from the above formula, we can see that when the fuzzy measures of n elements in set X are known, then the fuzzy measures of any subset in P (X ) can be obtained.
For a single element x j , µ x j is called fuzzy density function of x j , it represents the importance of the element x j .

C. CHOQUET INTEGRAL
When the attribute set has relevance, because the weighted operators of additive measure are no longer applicable, here, Choquet integral is used to calculate the comprehensive evaluation value of each alternative.
Definition 3 [23]: Suppose f is a positive real-valued function defined on X and µ is a fuzzy measure on P (X ), then the discrete Choquet integral of function f with respect to the fuzzy measure µ is, Among them, (j) represents a permutation of element subscript in X , satisfies 0 ≤ f x (1) ≤ f x (2) ≤ · · · ≤ f x (n) , and

D. INTUITIONISTIC UNCERTAIN LINGUISTIC SET
Definition 4 [51]: Suppose S = (s 0 , s 1 , · · · , s l ) be a linguistic set consisting of odd number (l + 1) elements, s α(x) , s β(x) ∈S, and X is a finite nonempty set, then is intuitionistic uncertain linguistic set, µ A (x) : X → [0, 1] and ν A (x) : X → [0, 1], and satisfies the conditions 0 ≤ and v A (x) represent respectively the membership degree and non-membership degree with the uncertain linguistic value s α(x) , s β(x) , in addition, represents the degree of indeterminacy of x for the uncertain linguistic number s α(x) , s β(x) , it is also called intuitionistic uncertain linguistic fuzzy index. Triple s α(x) , s β(x) , (µ A (x) , µ A (x)) is also called intuitionistic uncertain linguistic number.
, v A (ã 2 )) as any two intuitionistic uncertain linguistic variables, and λ ≥ 0, then there are the following operational laws, The above calculation results are still as intuitionistic uncertain linguistic variables.
It can be proved that the following calculation rules hold for any two uncertain linguistic numbersã 1 andã 2 .

E. INTUITIONISTIC UNCERTAIN LANGUAGE CHOQUET ORDERED WEIGHTED AGGREGATION OPERATOR
Based on the above related knowledge, this paper proposes the IULCWA to comprehensively screen the alternative battery energy storage technologies.
v A ã j , (j = 1, 2, · · · , n) as a group of intuitionistic uncertain linguistic variables, the IULCWA has the following definitions, and IULCW A : n → . (10), as shown at the bottom of the next page.

F. ALGORITHMIC STEPS
For a multi-attribute group decision-making problem, and the relevance problem between attribute index and experts. Suppose there are m alternatives, b i (i = 1, 2, · · · , m), n decision criteria X j (j = 1, 2, · · · , n) , p experts groups e k (k = 1, 2, · · · , p). The expert e k defines the attribute value of alternative b i under attribute x j as the intuitionistic uncertain linguistic number,    Step 1: Determine the fuzzy measure µ X j of attribute X j (j = 1, 2, · · · , n), and calculate the λ 1 value and the fuzzy measure of each attribute set by the formula (4).
Step 2: Use the formula (10) to aggregate criterion value of alternative b i given by the expert e k to obtain the comprehensive intuitionistic uncertain linguistic valuẽ f k i (i = 1, 2, · · · , m) of alternative b i .
Step 3: Determine the fuzzy measure µ (e k ) of expert e k (k = 1, 2, · · · , p), and then use the formula (4) to get the λ 2 value and the fuzzy measure of expert set.
Step 4: Aggregate the comprehensive intuitionistic uncertain linguistic valuef k i of the alternative b i given by p experts to obtain the group decision collective valuẽ F i (i = 1, 2, · · · , m) of scheme b i by using the formula (10).
step 5: Use the formula (8), (9) and (10) to calculate the expected value E F i ofF i (i = 1, 2, · · · , m), suppose the score function value as S F i and the accuracy function value as H F i .

A. EXAMPLE RESULTS
A Solar power plant in Ali, Tibet is going to add BESS to increase its power generation efficiency and flexibility. In order to select a suitable BESS, the actual situation of three different BESS is comprehensively evaluated. The three energy storage systems are represented by b i (i = 1, 2, 3), represent Lead-acid BESS, Sodium-sulfur BESS and Lithium-ion BESS respectively. The evaluation indexes include maintainability, technology, sociality, and economy, which is represented by x j (j = 1, 2, 3, 4). Three expert groups e k (k = 1, 2, 3) were selected to evaluate the energy storage plants. The three expert groups are from the fields of engineering, power technology and economics, and have rich practical experience in their respective fields. On the basis of the original data, three expert groups α(a (j))( µ(A (j)) −µ(A (j+1))) ,S n j=1 β(a (j))( µ(A (j)) −µ(A (j+1)))   ,   used intuitionistic uncertain linguistic numbers to provide the attribute values of each alternative that are given as follows: Table 5-7. The uncertain linguistic sets adopted by expert groups S = (s 0 , s 1 , s 2 , s 3 , s 4 , s 5 , s 6 ), which represent ''extremely poor'', ''very poor'', ''poor'', and ''average'', ''good'', ''very good'', ''excellent''. Due to the information redundance in technology, economy, sociality and maintainability, mutual independence of economy and society, and redundance, complementarity and independence of attributes, IULCWA can used to deal with this kind of problem.
1) The fuzzy measure of each attribute is calculated by the method proposed in reference [26] as follows: From formula (4), we can calculate λ 1 = −0.2715, then there are 2) Using the intuitionistic uncertain linguistic Choquet integral ordered weighted aggregation operator IULCWA to aggregate the criterion value of alternatives b i given by experts groups e k to obtain the comprehensive intuitionistic uncertain linguistic valuesf k i (i = 1, 2, 3) of alternatives b i .
For example, the intuitionistic uncertain linguistic value of the alternative b 1 given by the experts groups e 1 is aggregated, there aref 1 1 , as shown at the bottom of the next page. 3) Determine the fuzzy measure of each experts group and experts groups set. The fuzzy measure is calculated as follows according to the method proposed in literature [26]: Namely, Sodium-sulfur BESS, which is represented by b 2 , has the best comprehensive evaluation.

B. COMPARISON AND ANALYSIS
The IULCWA presented in this paper is a relatively novel method for BESS selection. It would be better to compare this method with the current mature and widely used decision making methods to verify its stability and feasibility. MCDM methods can be divided into two main categories, one of that considers the behavior of decision makers to be completely rational with the representative methods of TOPSIS, VIKOR and ELECTRE, etc.; the other considers that the behavior is of bounded rationality and such methods generally involve in psychological ideas, such as TODIM and Prospect Theory. In order to validate the reliability of IULCWA in the BESS selection, VIKOR and TODIM will be applied to re-rank the alternatives and then a comparative analysis is performed to describe the changes of the results. Table 10 was obtained by treating Table 5-7 according to the expert weight calculation method proposed in literature [52]. The specific results are shown in Table 11.
As can be seen from Table 11, alternative b 3 is considered more reasonable than b 2 in the VIKOR and TODIM methods. The reason is that VIKOR is based on the comprehensive distance of the criteria to an ideal solution, while TODIM considers the bounded rationality of human that decision makers are often difficult to be completely rational. These two methods do not consider the correlation between attributes and evaluation experts in the calculation process, and ignore the impact of extreme cases on the problem, resulting in a bias in the understanding of evaluation subjects.

V. CONCLUSION
Aiming at the practical problem of how to evaluate BESS comprehensively under the fuzzy environment, this paper constructs a comprehensive evaluation index system of BESS considering economy, sociality, technology and maintainability. In view of the current conventional method cannot well solve the problem of correlation between indicators and experts, the fuzzy measure and Choquet integral method are introduced into the comprehensive evaluation of BESS. Finally, the feasibility and validity of this method are verified by empirical analysis. On the basis of intuitionistic fuzzy number and uncertain linguistic evaluation, this paper defines the concept of intuitionistic uncertain linguistic set, puts forward the operation rule and number comparison method of intuitionistic uncertain linguistic number, and puts forward LULCWA, which solves the relevance in evaluation index and experts and can solve various multi-attribute decisionmaking problems more widely. The comprehensive evaluation index system of BESS constructed in this paper also considers the relevance of the indexes. It basically covers all aspects of the comprehensive evaluation of BESS. It can help the management decision of electric power enterprises and promote the sustainable development of the energy storage industry.
Because BESS involves complex factors, when using this method to make decision analysis, the evaluation of fuzzy information highly depends on individuals or groups and varies from person to person. Therefore, the knowledge level and experience of experts are highly demanded. Different experts may get different results when using different evaluation methods. This paper lists only three common BESS, without considering other available BESS. Next, the author will focus on the expansion of uncertain language sets to reduce the loss of information in the process of transformation. In addition, the IULCWA operator proposed in this paper is more troublesome, which is also the direction to be improved by the author.
In today's energy structure with renewable energy as the main body, the importance of electric energy storage system is increasing day by day. It is not only a powerful guarantee for the safe, reliable and efficient operation of the electric power system, but also a supporting technology for the development of the energy of the Internet. It is also an important part of the Smart Grids. Electric energy storage system provides an effective solution for peak-adjustment of power grid, improving the reliability of power grid and improving power quality, and provides an important guarantee for the sustainable development of renewable energy generation.