TRIZ-based green energy project
evaluation using innovation life cycle
and fuzzy modelling
YUE MENG1,*, HASAN DİNÇER2, SERHAT YÜKSEL2,*
1 Department of Tourism, Sichuan Agricultural University, Chengdu 611130, China 2 School of Business, Istanbul Medipol University, Istanbul, 34815 Turkey
*Corresponding author: Yue Meng (mengyue@sicau.edu.cn); Serhat Yüksel (serhatyuksel@medipol.edu.tr)
ABSTRACT This study eva lua tes theory of the solution of inventive problems (TRIZ)-ba sed green energy project network ba sed on innova tion life cycle. In this context, TRIZ-ba sed combina tions a re obta ined by integer pa tterns for green energy projects. In the next pha se, a ctivity priorities of TRIZ-ba sed principles a re mea sured for green energy projects with Pytha gorea n fuzzy (PF) technique for order preference by simila rity to idea l solution (TOPSIS). Fina lly, project eva lua tion review technique (PERT) a nd PF decision ma king tria l a nd eva lua tion la bora tory (DEMATEL) a pproa ches a re ta ken into considera tion to score the green energy projects ba sed on innova tion life cycle using inventive problem-solving principles. TRIZ-ba sed principles a re a lso ra nked by using Vise Kriterijumska Optimiza cija I Kompromisno Resenje (VIKOR). Additiona lly, a ll ca lcula tions ha ve a lso been ma de with Spherica l fuzzy (SF) sets in a ddition to the PF sets. It is concluded tha t the a na lysis results a re quite relia ble. The ma in contribution of this study is tha t innova tive stra tegies a re identified for green energy investment projects with a novel methodology. The findings indica te tha t dyna micity a nd tra nsforma tion of properties pla y a crucia l role in this process. Therefore, green energy investors should closely follow technologica l developments in this a rea a nd the products used in green energy investments should be ea sily a da pted to these developments. Additiona lly, the principa l of prior a ction should be considered to reduce the costs instea d of periodic a ction. It is obvious tha t the necessa ry controls should be ma de before the project sta rts to increa se the efficiency a nd productivity.
INDEX TERMS Green Energy, PERT, TRIZ, Innova tion Life Cycle, Integer Code Series, DEMATEL, TOPSIS
I. INTRODUCTION
Environmenta l pollution ha s become a crucia l problem throughout the world. Countries ha ve begun to look for the wa ys to reduce this problem. In this context, energy consumption is one of the issues tha t ca use this problem. I n the process of meeting the energy need, a significa nt a mount of ca rbon ga s is relea sed into the a tmosphere by using fossil fuels [1]. Due to this problem, ma ny people suffer from the respira tory disea ses. It is possible to ta lk a bout ma ny disa dvantages of this situa tion for the co u n t ry . Beca use of the increa se in the number of sick people, there will be a significa nt loss of workforce. Additiona lly, with the increa sing number of sick people, hea lthca re spending will go up significa ntly tha t ca uses the countries to ha ve budget deficits. The problems mentioned revea l the importa nce of green energy projects. In its most genera l definition, green energy mea ns energy tha t does not ha rm the environment while being supplied [2]. Ca rbon emissions a re reduced to minimum levels tha nks to the implementa tion of green energy projects.
Green energy projects pla y a crucia l role for the susta ina ble economic improvements. Beca use of this condition, necessa ry a ctions should be ta ken to improve these investments [3]. However, there a re some ba rriers for this situa tion. For exa mple, green energy investment projects ha ve high initia l costs [4]. This condition ma kes the compa nies relucta nt to ma ke investments in this a rea . Furthermore, the green energy investments a re complex and long-term investments. This situa tion increa ses the uncerta inty of these projects beca use it ca uses liquidity risks. During this process, the compa nies ma y ha ve liquidity crisis if a n effective pla nning is not implemented. Moreover, since green energy investments a re complex projects, the compa nies need qua lified employees [5]. Otherwise, it will be very difficult to solve the problems quickly. In a ddition, for green energy projects to be successful, compa nies should ma ke significa nt technology investments.
Simila rly, the popula rity of green energy investments in the litera ture ha s increa sed. Some resea rchers focused on the impa cts of these projects on the economic growth of the countries. Most of them rea ched a conclusion tha t green energy investments pla y a key role for the susta ina ble economic development beca use they lea d to lower ca rbon emission. On the other side, some studies highlighted the difficulties in green energy investment projects. A significa nt pa rt of these studies underlined the significa nce of the high-cost problem for this situa tion. It is understood tha t in the litera ture, the issue of green energy investments is dea lt with mostly from a genera l point of view. However, there is a need for a new study which ca n provide specific a nd cost-effective stra tegies to improve these projects. Owing to innova tive a nd cost-effective investment stra tegies, these projects ca n be increa sed.
It is a imed to develop green energy project network ba sed on innova tion life cycle. A new inventive problem-solving model is suggested for green energy project eva lua tion which ha s four different pha ses. The first pha se includes determining the TRIZ-ba sed principles for green energy projects. In the second pha se, the TRIZ-ba sed combina tions a re used. Furthermore, the third pha se is rela ted to the mea suring the a ctivity priorities of TRIZ-ba sed principles for green energy projects by PF TOPSIS a nd PF VIKOR. Moreover, the fina l pha se focuses on scoring the green energy project ba sed on innova tion life cycle using inventive problem-solving principles. An integra ted method of PERT a nd PF DEMATEL a re considered. In a ddition, a ll ca lcula tions ha ve a lso been ma de by using SF sets to check relia bility.
The contribution is tha t stra tegies a re identified for green energy investment projects by a hybrid multi-criteria decision-ma king (MCDM) model ba sed on integer pa tterns a nd PF sets. There a re a lso some superiorities of this proposed model in compa rison with the previous ones. A hybrid model is crea ted which mea ns tha t different MCDM a pproa ches a re considered for both ra nking the items a nd finding the influentia l rela tionships [6]. This situa tion provides a n objective eva lua tion [7]. Furthermore, the principles for green energy project a ctivities a re defined ba sed on TRIZ technique. In this a pproa ch, more tha n 2 million pa tents a re exa mined, a nd importa nt stra tegies a re defined [8]. Hence, by considering this a pproa ch, innova tive solutions ca n be identified without wa sting too much time [9]. Furthermore, using PF sets provides a n opportunity to reflect uncerta inty in a more suita ble wa y [10-12].
Another novelty is using TOPSIS to define a ctivity priorities. The ma in a dva nta ge is considering both the dista nces to positive a nd nega tive idea l solutions [13,14] which contribute more relia ble results [15]. PERT method is considered in this process [16]. In this technique, the completion times of the a ctivities a re not sta ted precisely, this process is exa mined in a proba bilistic structure [17,18]. This situa tion gives the informa tion tha t the a na lyzes ma d e with the PERT technique a re more rea listic [18]. However, the ma in disa dva nta ge of this technique is tha t there is no a na lysis on how the ra nking of the a ctivities will be [19,20]. Despite this issue, in this study, the green energy project network is constructed by DEMATEL a na lysis. Therefore, it is obvious tha t the more objective eva lua tion ca n be made [21,22]. Moreover, ma king compa ra tive eva lua tions with VIKOR a nd SF sets, the relia bility is mea sured. Using integer pa tterns to find the best combina tions of TRIZ-ba sed principles for innova tion life cycle ha s a lso benefits [23,24]. Thus, it ca n be possible to check the relia bility of the eva lua tions for the pa tterns [25,26].
The rest of the pa per is orga nized a s following. Section 2 includes the litera ture review for both green energy investments a nd methods used in the eva lua tion. Section 3 defines the integer pa tterns a nd geometrica l recognition,
Pytha gorea n fuzzy sets, TOPSIS, DEMATEL a nd proposed mode. Moreover, a na lysis results a re indica ted in the section 4. In the fina l two sections, conclusions a nd discussion a re expla ined.
II. LITERATURE REVIEW
Litera ture eva lua tion is provided for different subjects.
A. GREEN ENERGY INVESTMENTS
The positive contributions of green energy investment projects a re mentioned. It wa s empha sized tha t investments increa se the energy dependence [27,28]. Bhowmik et a l. [29] expla ined tha t clea n energy projects reduce the current a ccount deficit problem of countries [30]. Poggi et a l. [31] underlined tha t since these projects do not ha rm the environment, they contribute to the positive development of the ima ge of the countries. Sha ikh et a l. [32] determined tha t it becomes ea sier for compa nies tha t do not ha rm the environment in energy consumption to obta in loa ns from interna tiona l fina ncia l institutions. In a ddition, Da mette a nd Ma rques [33] highlighted tha t tha nks to green energy projects, the number of people sick is decrea sing. The decrea se in the number of sick people is a fa ctor tha t increa ses the qua lity of life in the country [34].
Another issue exa mined in green energy investments is the fina ncing difficulty [35,36]. Ka ldellis a nd Za fira kis [37] underlined tha t high sta rt-up costs a lso lea d to the problem of fina ncing difficulties. It is not possible to obta in a la rge a mount of investment loa ns from every ba nk. This proble m is rega rded a s the biggest ba rrier to the development of green energy projects. Li et a l. [38] a nd Lin et a l. [39] concluded tha t green energy investments a re long-term projects involving complex processes. As ca n be seen, this problem must be solved in order to increa se green energy investments. Additiona lly, Gong et a l. [40] a nd Breetz et a l. [41] a lso cla imed tha t the governments should support green energy investors. Ta x cuts to be a pplied to these projects will provide investors with a significa nt cost a dva nta ge. In a ddition, governments ca n provide low interest loa ns for green energy projects. This will help investors to get the funds they need. In a ddition, Liu et a l. [42] a nd Nguyen et a l. [43] discussed tha t green bonds a re a nother a pplica tion tha t ma y attract investors' a ttention. Risk ma na gement is a nother importa nt issue for green energy investments. In order for these projects to be ca rried out successfully, the risks encountered in this process must be ma na ged effectively [39]. First, the risks encountered in green energy investments should be clea rly defined [44]. One importa nt risk in this process is fina ncia l risks [45]. Koca a rsla n a nd Soyta s [46] a nd Jones [47] sta ted tha t Some ma teria ls used in green energy projects ma y need to be imported. In this ca se, these products must be pa id in foreign currency. Therefore, a possible increa se in the excha nge ra te ma y ca use these products to be more expensive. This situa tion is a lso va lid for loa ns obta ined from fina ncia l institutions [48,49]. Therefore, a n increa se in
the excha nge ra te ca uses the debt of the compa ny to increa se. Dutta et a l. [50] identified tha t efforts should be ma de to minimize these risks by considering fina ncia l deriva tives. Moreover, Ya ng et a l. [51] defined tha t the opera tiona l risks a re a lso of grea t importa nce in this process. In this context, personnel errors a nd ma lfunctions tha t ma y occur in informa tion technology systems lower the performa nce. Cheung et a l. [52] concluded tha t it should b e ensured tha t opera tiona l risks a re minimized by providing the necessa ry tra ining to the personnel. In a ddition, tha nks to the resea rch a nd development studies to be ca rried out, it will be possible to reduce the disruptions in informa tion technologies.
B. LITERATURE REVIEW ON METHODOLOGY
TRIZ is a technique tha t a rgues tha t the effectiveness of th e current situa tion ca n be increa sed more ea sily, considering pa st experiences [53]. In other words, stra tegy suggestions a re presented for the solution of the existing problem, ta king into considera tion the methods a pplied in solving the problems experienced before. In this wa y, it is possible to solve the problems much fa ster [9]. Hence, the TRIZ ha s been considered for ma ny different purposes. Feniser et a l. [54] focused on the wa ys of increa sing eco-innova tive levels in SMEs with TRIZ. Additiona lly, Moussa et a l. [55] a imed to solve green supply cha in problems by using this technique. Furthermore, Čačo et al. [8] made a study to optimize the a utoma ted ma chine for ultra sonic welding. PERT method is a lso considered in eva lua ting the effectiveness of a project. In this context, this method is very helpful in ma tters such a s esta blishing a fa cility, developing computer systems, a nd developing a new product [16]. Huynh a nd Nguyen [17] a nd Sa ckey a nd Kim [18] a imed to ma ke schedule risk a na lysis with this technique. Furthermore, Sima nkina et a l. [19] focused on energy consumption economy by using PERT a na lysis. Lee et a l. [20] tried to construct a n energy pla nt by this a pproa ch. On the other side, the combina tion of DEMATEL a nd TOPSIS were a lso used for va rious purposes, such a s eva lua ting knowledge tra nsfer effectiveness [15], truck selection [6], medica l tourism a doption [14], eva lua tion of the fina ncia l sectors [7] a nd risk a ssessment of hydrogen genera tion [4].
C. THE IMPORTANT POINTS
The litera ture eva lua tion indica tes tha t clea n energy investments pla y a crucia l role for economic improvement. In a ddition to them, some resea rchers a lso eva lua ted the fa ctors influencing these projects. It is determined tha t there is a need for a new study tha t provide significa nt stra tegies for the improvement of the clea n energy projects. This ma nuscript develops a novel inventive problem-solving model for green energy project eva lua tion ba sed on innova tion life cycle using Pytha gorea n fuzzy sets with integer pa tterns. The ma in contribution is tha t innova tive stra tegies a re identified for green energy investment projects with a n origina l methodology.
III. METHODOLOGY
All methods considered in the a na lysis a re deta iled.
A. INTEGER PATTERNS AND GEOMETRICAL RECOGNITION
This methodology considers integer forma tions to solve complex problems. In this context, I represents a n integer a lpha bet. Additiona lly, the terms ≥ 2. 𝛿𝛿 > 0 𝑎𝑎𝑎𝑎𝑎𝑎 𝜀𝜀 > 0 give informa tion a bout the spa cings of a spa cetime la ttice (𝛿𝛿, 𝜀𝜀). The equation 1 indicates the details of this process [23].
𝐼𝐼𝑛𝑛= {𝑠𝑠 = 𝑠𝑠1… 𝑠𝑠𝑛𝑛, 𝑠𝑠𝑖𝑖∈ 𝐼𝐼, 𝑖𝑖 = 1, … , 𝑎𝑎} (1)
Moreover, 𝑊𝑊𝛿𝛿𝛿𝛿([𝑡𝑡𝑚𝑚, 𝑡𝑡𝑚𝑚+𝑛𝑛]) is the set of piecewise constant functions. Also, 𝑓𝑓 is constant and can get value of (𝑡𝑡𝑖𝑖−1, 𝑡𝑡𝑖𝑖]. Equa tions (2)-(6) a re ta ken into considera tion [24]. 𝑓𝑓: [𝑡𝑡𝑚𝑚, 𝑡𝑡𝑚𝑚+𝑛𝑛] → ℜ1 (2)
𝑖𝑖 = 𝑚𝑚 + 1, … , 𝑚𝑚 + 𝑎𝑎 (3)
𝑓𝑓(𝑡𝑡𝑚𝑚) = 𝑠𝑠1𝛿𝛿 (4)
𝑓𝑓(𝑡𝑡) = 𝑠𝑠𝑖𝑖𝛿𝛿 (5)
𝑡𝑡 ∈ (𝑡𝑡𝑖𝑖−1, 𝑡𝑡𝑖𝑖] and 𝑡𝑡𝑖𝑖= 𝑖𝑖𝜀𝜀 (6)
In these equa tions, 𝑚𝑚 shows an integer and 𝑠𝑠𝑖𝑖, 𝑖𝑖 = 1, … , 𝑎𝑎 show rea l numbers. Moreover, 𝑓𝑓[𝑘𝑘] defines the kth integra l. Also, the equa tion (7) defines the integer code series provides the kth integra l of a function [25]. 𝑓𝑓 ∈ 𝑊𝑊𝛿𝛿𝛿𝛿([𝑡𝑡𝑚𝑚, 𝑡𝑡𝑚𝑚+𝑛𝑛]) (7)
On the other side, kth integra l should sa tisfy the equa tion (8). 𝑓𝑓[𝑘𝑘](𝑡𝑡𝑚𝑚) = 0 (8)
Additiona lly, this function ca n a lso be considered with the code 𝑐𝑐(𝑓𝑓) = 𝑠𝑠1… 𝑠𝑠𝑛𝑛 a nd the powers of integers. This process is deta iled in the equa tions (9)-(11) [26]. 𝑓𝑓[𝑘𝑘](𝑡𝑡𝑚𝑚+𝑙𝑙+1) = ∑𝑘𝑘−1𝑖𝑖=0𝛼𝛼𝑘𝑘𝑚𝑚𝑖𝑖((𝑚𝑚 + 𝑙𝑙 + 1)𝑖𝑖𝑠𝑠1+ ⋯ + (𝑚𝑚 + 1)𝑖𝑖𝑠𝑠 𝑙𝑙+1) 𝛿𝛿𝜀𝜀𝑘𝑘+ ∑𝑘𝑘𝑖𝑖=0𝛽𝛽𝑘𝑘 ,𝑙𝑙+1,𝑖𝑖𝑓𝑓[𝑖𝑖](𝑡𝑡𝑚𝑚)𝜀𝜀𝑘𝑘−𝑖𝑖 (9) 𝛼𝛼𝑘𝑘𝑚𝑚𝑖𝑖 = �𝑘𝑘 𝑖𝑖�((−1)𝑘𝑘−𝑖𝑖−1(𝑚𝑚 + 1)𝑘𝑘−𝑖𝑖+ (−1)𝑘𝑘−𝑖𝑖𝑚𝑚𝑘𝑘−𝑖𝑖) 𝑘𝑘! (10) 𝛽𝛽𝑘𝑘,𝑙𝑙+1,𝑖𝑖=(𝑙𝑙 + 1) 𝑘𝑘−𝑖𝑖 (𝑘𝑘 − 𝑖𝑖)! (11) In a ddition, Figure 1 illustra tes pa tterns.
FIGURE 1.Geometrical patterns
Figure 1 gives informa tion a bout the geometrica l pa tterns. They a re genera ted with the integra tion of the function 𝑓𝑓[𝑘𝑘](𝑡𝑡), 𝑡𝑡0≤ 𝑡𝑡 ≤ 𝑡𝑡16 a nd k=1,2,3. Equa tion (12) shows this
situa tion.
𝑓𝑓[0](𝑡𝑡), 𝑡𝑡𝑗𝑗−1≤ 𝑡𝑡 ≤ 𝑡𝑡𝑗𝑗 𝑡𝑡𝑗𝑗= 𝑗𝑗𝜀𝜀, j = 1,2, … ,16 (12) Moreover, 𝑇𝑇𝑗𝑗0(𝑡𝑡𝑗𝑗−1) = 𝑡𝑡𝑗𝑗−1−𝑡𝑡𝑗𝑗−1= 0 with respect to the tra nsition from one sta te into a nother.
B. PYTHAGOREAN FUZZY SETS
PF a ims to define a new non-sta nda rd fuzzy membership gra des. These gra des a re considered pa ir sets over U which is the universe of discourse. This situa tion is expla ined in equa tion (13) [56].
𝑃𝑃 = �〈𝜗𝜗,𝜇𝜇𝑃𝑃(𝜗𝜗), 𝑎𝑎𝜗𝜗𝜗𝜗𝜗𝜗𝑃𝑃(𝜗𝜗)〉� (13) In this context, 𝜇𝜇𝑃𝑃(𝜗𝜗): 𝜗𝜗 → [0,1] is the degree of membership a nd 𝑎𝑎𝑃𝑃(𝜗𝜗): 𝜗𝜗 → [0,1] explains the degree of
non-membership of the element 𝜗𝜗 𝜗𝜗 𝜗𝜗. Equation (14) shows the deta ils of this process [57].
�𝜇𝜇𝑃𝑃(𝜗𝜗)� 2
+ �𝑎𝑎𝑃𝑃(𝜗𝜗)� 2
≤ 1 (14) Also, the degree of indetermina cy is ca lcula ted by considering equa tion (15) [10].
𝜋𝜋𝑃𝑃(𝜗𝜗) = �1 − �𝜇𝜇𝑃𝑃(𝜗𝜗)�2− �𝑎𝑎𝑃𝑃(𝜗𝜗)�2 (15) Furthermore, in equa tions (16)-(20), the deta ils rega rding the opera tions of Pytha gorea n fuzzy sets a re shown [11]. 𝑃𝑃1 = � 〈𝜗𝜗,𝑃𝑃1(𝜇𝜇𝑃𝑃1(𝜗𝜗),𝑛𝑛𝑃𝑃1(𝜗𝜗))〉 𝜗𝜗𝜗𝜗𝜗𝜗 � a nd 𝑃𝑃2 = �〈𝜗𝜗,𝑃𝑃2(𝜇𝜇𝑃𝑃2(𝜗𝜗),𝑛𝑛𝑃𝑃2(𝜗𝜗))〉 𝜗𝜗𝜗𝜗𝜗𝜗 � (16) 𝑃𝑃1⊕ 𝑃𝑃2= 𝑃𝑃 �� 𝜇𝜇1𝑃𝑃1+ 𝜇𝜇2𝑃𝑃2− 𝜇𝜇1𝑃𝑃1𝜇𝜇22 , 𝑎𝑎𝑃𝑃1𝑎𝑎𝑃𝑃1� (17) 𝑃𝑃1 ⊗ 𝑃𝑃2 = 𝑃𝑃 �𝜇𝜇𝑃𝑃1𝜇𝜇𝑃𝑃2,� 𝑎𝑎2𝑃𝑃1+ 𝑎𝑎𝑃𝑃22− 𝑎𝑎𝑃𝑃21𝑎𝑎𝑃𝑃22 � (18)
𝜆𝜆𝑃𝑃 = 𝑃𝑃 ��1 − �1 − 𝜇𝜇2𝑝𝑝 � 𝜆𝜆 , �𝑎𝑎𝑝𝑝� 𝜆𝜆 � , 𝜆𝜆 > 0 (19) 𝑃𝑃𝜆𝜆 = 𝑃𝑃 � �𝜇𝜇 𝑝𝑝�𝜆𝜆,�1 − �1 − 𝑎𝑎𝑝𝑝2 �𝜆𝜆 � , 𝜆𝜆 > 0 (20) For defuzzifica tion, the score function (S) is considered a s in the equa tion (21) [12].
𝑆𝑆(𝜗𝜗) = �𝜇𝜇𝑃𝑃(𝜗𝜗)�2− �𝑎𝑎𝑃𝑃(𝜗𝜗)�2 where 𝑆𝑆(𝜗𝜗) ∈ [−1,1] (21)
C. SPHERICAL FUZZY (SF) SETS
The membership, non-membership, a nd hesita ncy parameters (μ, y and π) are taken into account in SF sets (𝐷𝐷�𝑆𝑆). Equa tions (22) a nd (23) a re used in this process [58].
𝐷𝐷�𝑆𝑆= �⟨𝑢𝑢, (𝜇𝜇𝐷𝐷�𝑆𝑆(𝑢𝑢), 𝑦𝑦𝐷𝐷�𝑆𝑆(𝑢𝑢), 𝜋𝜋𝐷𝐷�𝑆𝑆(𝑢𝑢))|𝑢𝑢 ∈ 𝜗𝜗� (22) 0 ≤ 𝜇𝜇2𝐷𝐷�𝑆𝑆(𝑢𝑢) + 𝑦𝑦 𝐷𝐷�𝑆𝑆 2 (𝑢𝑢) + 𝜋𝜋 𝐷𝐷�𝑆𝑆 2 (𝑢𝑢) ≤ 1 (23) In a ddition, 𝐷𝐷�𝑆𝑆= (𝜇𝜇𝐷𝐷�𝑆𝑆, 𝑦𝑦𝐷𝐷�𝑆𝑆, 𝜋𝜋𝐷𝐷�𝑆𝑆) and 𝐸𝐸�𝑆𝑆 = (𝜇𝜇𝐸𝐸�𝑆𝑆, 𝑦𝑦𝐸𝐸�𝑆𝑆, 𝜋𝜋𝐸𝐸�𝑆𝑆) give informa tion a bout two Spherica l fuzzy sets. Ma thema tica l deta ils a re demonstra ted in equa tions (24)-(27) [59]. 𝐷𝐷�𝑆𝑆⊕ 𝐸𝐸�𝑆𝑆= ⎩ ⎪ ⎨ ⎪ ⎧ �𝜇𝜇2𝐷𝐷�𝑆𝑆+ 𝜇𝜇 𝐸𝐸�𝑆𝑆 2 − 𝜇𝜇 𝐷𝐷�𝑆𝑆 2 𝜇𝜇 𝐸𝐸�𝑆𝑆 2 �12 , 𝑦𝑦𝐴𝐴�𝑆𝑆𝑦𝑦𝐵𝐵�𝑆𝑆,� �1 − 𝜇𝜇2𝐸𝐸�𝑆𝑆�𝜋𝜋 𝐷𝐷�𝑆𝑆2 + �1 − 𝜇𝜇2𝐷𝐷�𝑆𝑆�𝜋𝜋 𝐸𝐸�𝑆𝑆2 − 𝜋𝜋𝐷𝐷�𝑆𝑆2 𝜋𝜋𝐸𝐸𝑆𝑆2 � 1 2 ⎭ ⎪ ⎬ ⎪ ⎫ (24) 𝑫𝑫�𝑺𝑺⊗ 𝑬𝑬�𝑺𝑺= ��𝝁𝝁𝑫𝑫�𝑺𝑺𝝁𝝁𝑬𝑬�𝑺𝑺,(𝒚𝒚𝟐𝟐𝑫𝑫�𝑺𝑺+ 𝒚𝒚 𝑬𝑬�𝑺𝑺 𝟐𝟐 − 𝒚𝒚 𝑫𝑫�𝑺𝑺 𝟐𝟐 𝒚𝒚 𝑬𝑬�𝑺𝑺 𝟐𝟐 �𝟏𝟏𝟐𝟐 ,��𝟏𝟏 − 𝒚𝒚 𝑬𝑬�𝑺𝑺 𝟐𝟐� 𝝅𝝅 𝑫𝑫�𝑺𝑺 𝟐𝟐 + �𝟏𝟏 − 𝒚𝒚𝑫𝑫�𝑺𝑺𝟐𝟐 � 𝝅𝝅 𝑬𝑬𝑺𝑺 𝟐𝟐 − 𝝅𝝅 𝑫𝑫𝑺𝑺 𝟐𝟐 𝝅𝝅 𝑬𝑬�𝑺𝑺 𝟐𝟐 � 𝟏𝟏 𝟐𝟐�𝑫𝑫� 𝑺𝑺⊗ 𝑬𝑬�𝑺𝑺 = ��𝝁𝝁𝑫𝑫�𝑺𝑺𝝁𝝁𝑬𝑬�𝑺𝑺,(𝒚𝒚𝑫𝑫�𝑺𝑺 𝟐𝟐 + 𝒚𝒚 𝑬𝑬�𝑺𝑺 𝟐𝟐 − 𝒚𝒚𝑫𝑫�𝑺𝑺𝟐𝟐 𝒚𝒚 𝑬𝑬�𝑺𝑺 𝟐𝟐�𝟏𝟏𝟐𝟐 ,��𝟏𝟏 − 𝒚𝒚 𝑬𝑬�𝑺𝑺 𝟐𝟐 � 𝝅𝝅 𝑫𝑫�𝑺𝑺 𝟐𝟐 + �𝟏𝟏 − 𝒚𝒚𝑫𝑫�𝑺𝑺𝟐𝟐 � 𝝅𝝅 𝑬𝑬�𝑺𝑺 𝟐𝟐 − 𝝅𝝅𝑫𝑫�𝑺𝑺𝟐𝟐 𝝅𝝅 𝑬𝑬�𝑺𝑺 𝟐𝟐 � 𝟏𝟏 𝟐𝟐� (𝟐𝟐𝟐𝟐) 𝝀𝝀 ∗ 𝑫𝑫�𝑺𝑺= ��𝟏𝟏 − �𝟏𝟏 − 𝝁𝝁𝑫𝑫�𝑺𝑺𝟐𝟐 �𝝀𝝀� 𝟏𝟏 𝟐𝟐 ,𝒚𝒚𝝀𝝀𝑫𝑫�𝑺𝑺,��𝟏𝟏 − 𝝁𝝁 𝑫𝑫�𝑺𝑺 𝟐𝟐 �𝝀𝝀− �𝟏𝟏 − 𝝁𝝁𝑫𝑫�𝑺𝑺𝟐𝟐 − 𝝅𝝅 𝑫𝑫�𝑺𝑺 𝟐𝟐 �𝝀𝝀� 𝟏𝟏 𝟐𝟐 � , 𝝀𝝀 > 𝟎𝟎 (𝟐𝟐𝟐𝟐) 𝑫𝑫�𝑺𝑺𝝀𝝀= �𝝁𝝁𝑫𝑫�𝑺𝑺𝝀𝝀 ,�𝟏𝟏 − �𝟏𝟏 − 𝒚𝒚 𝑫𝑫�𝑺𝑺 𝟐𝟐 �𝝀𝝀� 𝟏𝟏 𝟐𝟐 , ��𝟏𝟏 − 𝒚𝒚𝑫𝑫�𝑺𝑺𝟐𝟐 �𝝀𝝀− �𝟏𝟏 − 𝒚𝒚 𝑫𝑫�𝑺𝑺 𝟐𝟐 − 𝝅𝝅𝑫𝑫�𝑺𝑺𝟐𝟐 �𝝀𝝀� 𝟏𝟏 𝟐𝟐 � , 𝝀𝝀 > 𝟎𝟎 (𝟐𝟐𝟐𝟐) The spherica l weighted a rithmetic mea n (SWAM) is considered a s in equa tion (28) [60].
𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝑺𝒘𝒘�𝑫𝑫�𝑺𝑺𝟏𝟏,. ., 𝑫𝑫�𝑺𝑺𝑺𝑺� = 𝒘𝒘𝟏𝟏𝑫𝑫�𝑺𝑺𝟏𝟏+ ⋯ + 𝒘𝒘𝑺𝑺𝑫𝑫�𝑺𝑺𝑺𝑺= ��𝟏𝟏 − � �𝟏𝟏 − 𝝁𝝁𝑫𝑫�𝑺𝑺𝑺𝑺𝟐𝟐 �𝒘𝒘𝑺𝑺 𝑺𝑺 𝑺𝑺=𝟏𝟏 � 𝟏𝟏 𝟐𝟐 ,� 𝒚𝒚𝑫𝑫�𝑺𝑺𝑺𝑺𝒘𝒘𝑺𝑺 𝑺𝑺 𝑺𝑺=𝟏𝟏 , �� �𝟏𝟏 − 𝝁𝝁𝑫𝑫�𝑺𝑺𝑺𝑺𝟐𝟐 �𝒘𝒘𝑺𝑺 𝑺𝑺 𝑺𝑺=𝟏𝟏 − � �𝟏𝟏 − 𝝁𝝁𝑫𝑫�𝟐𝟐𝑺𝑺𝑺𝑺 𝑺𝑺 𝑺𝑺=𝟏𝟏 − 𝝅𝝅𝑫𝑫�𝟐𝟐𝑺𝑺𝑺𝑺�𝒘𝒘𝑺𝑺� 𝟏𝟏 𝟐𝟐 � (𝟐𝟐𝟐𝟐) D. TOPSIS
It is used to ra nk the a lterna tives. First, the norma lized va lues a re used a s in equa tion (29) [13].
𝒓𝒓𝑺𝑺𝒊𝒊 = 𝑿𝑿𝑺𝑺𝒊𝒊 �∑𝒎𝒎𝑺𝑺=𝟏𝟏𝑿𝑿𝑺𝑺𝒊𝒊𝟐𝟐
𝑺𝑺 = 𝟏𝟏, 𝟐𝟐, 𝟑𝟑, . . 𝒎𝒎 𝒂𝒂𝑺𝑺𝒂𝒂 𝒊𝒊 = 𝟏𝟏, 𝟐𝟐, 𝟑𝟑,. . 𝑺𝑺 (𝟐𝟐𝟐𝟐)
The second step includes weighting the va lues a s in equa tion (30).
𝒔𝒔𝑺𝑺𝒊𝒊= 𝒘𝒘𝑺𝑺𝒊𝒊 × 𝒓𝒓𝑺𝑺𝒊𝒊 (𝟑𝟑𝟎𝟎) La ter, the positive (𝑺𝑺+) a nd nega tive (𝑺𝑺−) idea l solutions a re defined. In this context, equa tion (31) a nd (32) a re u se d [14].
𝐴𝐴+ = �𝑠𝑠
1𝑗𝑗, … , 𝑠𝑠𝑚𝑚𝑗𝑗� = �max 𝑠𝑠1𝑗𝑗 𝑓𝑓𝑓𝑓𝑓𝑓 ∀ 𝑗𝑗 ∈ 𝑎𝑎� (31) 𝑺𝑺− = �𝒔𝒔
𝟏𝟏𝒊𝒊,… , 𝒔𝒔𝒎𝒎𝒊𝒊� = �𝒎𝒎𝑺𝑺𝑺𝑺 𝒔𝒔𝟏𝟏𝒊𝒊 𝒇𝒇𝒇𝒇𝒓𝒓 ∀ 𝒊𝒊 ∈ 𝑺𝑺� (𝟑𝟑𝟐𝟐) Next, the dista nces a re computed a s in equa tions (33) a nd (34). In this process, 𝑫𝑫𝑺𝑺+ a nd 𝑫𝑫𝑺𝑺− show the dista nce to the best a nd the worst items [15].
𝑫𝑫𝑺𝑺+= �∑ �𝒔𝒔𝑺𝑺𝒊𝒊− 𝑺𝑺𝒊𝒊+� 𝟐𝟐 𝑺𝑺 𝒊𝒊=𝟏𝟏 (𝟑𝟑𝟑𝟑) 𝑫𝑫𝑺𝑺−= �∑ �𝒔𝒔𝑺𝑺𝒊𝒊− 𝑺𝑺𝒊𝒊−� 𝟐𝟐 𝑺𝑺 𝒊𝒊=𝟏𝟏 (𝟑𝟑𝟑𝟑) Fina lly, the rela tive closeness (RCi) is ca lcula ted with equa tion (35). It is considered to ra nk the a lterna tives.
𝑹𝑹𝑹𝑹𝑺𝑺 = 𝑫𝑫𝑺𝑺− 𝑫𝑫𝑺𝑺++ 𝑫𝑫 𝑺𝑺 − (𝟑𝟑𝟐𝟐) E. VIKOR
VIKOR a pproa ch a lso ra nks different a lterna tives. The first step of TOPSIS methodology is a lso simila r for VIKOR. Next, fuzzy best a nd worst va lues (𝑓𝑓̃𝑗𝑗∗, 𝑓𝑓̃𝑗𝑗−) are computed by equa tion (36) [61].
Mea n group utility (𝑆𝑆̃𝑖𝑖) and maximal regret (𝑅𝑅�𝑖𝑖) are identified a s in equa tions (37) a nd (38) [62].
𝑺𝑺�𝑺𝑺= � 𝒘𝒘�𝒊𝒊 ��𝒇𝒇�𝒊𝒊∗− 𝒎𝒎�𝑺𝑺𝒊𝒊�� ��𝒇𝒇�𝒊𝒊∗− 𝒇𝒇�𝒊𝒊−�� 𝑺𝑺 𝑺𝑺=𝟏𝟏 (𝟑𝟑𝟐𝟐) 𝑹𝑹�𝑺𝑺= 𝒎𝒎𝒂𝒂𝒎𝒎𝒊𝒊�𝒘𝒘�𝒊𝒊 ��𝒇𝒇�𝒊𝒊∗− 𝒎𝒎�𝑺𝑺𝒊𝒊�� ��𝒇𝒇�𝒊𝒊∗− 𝒇𝒇�𝒊𝒊−��� (𝟑𝟑𝟐𝟐) In these equa tions, 𝑤𝑤�𝑗𝑗 represents fuzzy weights. Next, the va lue of 𝑄𝑄�𝑖𝑖 is ca lcula ted a s in equa tion (39) [63].
𝑸𝑸�𝑺𝑺= 𝒗𝒗 �𝑺𝑺�𝑺𝑺− 𝑺𝑺�∗� �𝑺𝑺�� − − 𝑺𝑺�∗� + (𝟏𝟏
− 𝒗𝒗) �𝑹𝑹�𝑺𝑺− 𝑹𝑹�∗� �𝑹𝑹�� −− 𝑹𝑹�∗� (𝟑𝟑𝟐𝟐) Within this context, the stra tegy weights a re shown a s v. On the other side, 1-v indica tes the weight of the individua l regret. In this process, items a re ra nked by using the va lues of S, Q a nd R. Equa tion (40) identifies the deta ils of the first requirement.
𝑄𝑄�𝐴𝐴(2)� − 𝑄𝑄�𝐴𝐴(1)� ≥ 1
(𝑗𝑗 − 1) (40) Additiona lly, the second requirement is rela ted to the a ccepta ble sta bility.
F. DEMATEL
Firstly, the expert tea m eva lua tes the criteria . After tha t, the direct rela tion ma trix (A) is obta ined. Equa tion (41) expla ins the deta ils of this ma trix [6].
A= ⎣ ⎢ ⎢ ⎢ ⎡ 0 𝑎𝑎12 𝑎𝑎13 ⋯ 𝑎𝑎1𝑛𝑛 𝑎𝑎21 0 𝑎𝑎23 ⋯ 𝑎𝑎2𝑛𝑛 𝑎𝑎31 𝑎𝑎32 0 ⋯ 𝑎𝑎3𝑛𝑛 ⋮ ⋮ ⋮ ⋱ ⋮ 𝑎𝑎𝑛𝑛1 𝑎𝑎𝑛𝑛2 𝑎𝑎𝑛𝑛3 ⋯ 0 ⎦ ⎥ ⎥ ⎥ ⎤ (41)
Within this fra mework, aij indica tes the influence of criterion i on the criterion j. Next, equa tions (42) a nd (43) a re considered to genera te the norma lized ma trix (B) [7].
𝐵𝐵 =𝑚𝑚𝑎𝑎𝑚𝑚 𝐴𝐴
1≤𝑖𝑖≤𝑛𝑛∑𝑛𝑛𝑗𝑗=1𝑎𝑎𝑖𝑖𝑗𝑗 (42) 0 ≤ 𝑏𝑏𝑖𝑖𝑗𝑗≤ 1 (43) La ter, tota l rela tion ma trix (C) is developed by using equa tion (44). The identity ma trix is denoted by I [21].
𝐶𝐶 = 𝐵𝐵(𝐼𝐼 − 𝐵𝐵)−1 (44) The sums of rows a nd columns (D a nd E) a re computed with equa tions (45) a nd (46) [22].
𝐷𝐷 = �∑𝑛𝑛𝑗𝑗=1𝑒𝑒𝑖𝑖𝑗𝑗�𝑛𝑛𝑛𝑛1 (45) 𝐸𝐸 = �∑𝑛𝑛𝑖𝑖=1𝑒𝑒𝑖𝑖𝑗𝑗�1𝑛𝑛𝑛𝑛 (46) Within this scope, the va lue of D+E is used for weighting the fa ctors. For this purpose, threshold va lue (𝛼𝛼) is taken into considera tion in equa tion (47).
𝛼𝛼 = ∑ ∑ �𝑒𝑒𝑖𝑖𝑗𝑗� 𝑛𝑛 𝑗𝑗=1 𝑛𝑛 𝑖𝑖=1 𝑁𝑁 (47) G. PROPOSED MODEL
A new 4-sta ge model ha s been suggested by integra ting the methods expla ined a bove. Figure 2 demonstra tes a ll steps of this proposed model.
FIGURE 2.The details of proposed model
Pha se 1: Determining the TRIZ-ba sed principles for green energy projects.
Step 1: Select the cha ra cteristics of green energy projects for TRIZ technique.
Step 2: Construct the contra diction ma trix for green energy projects.
Step 3: Define the TRIZ-ba sed principles for green energy project a ctivities.
Pha se 2: Collecting the TRIZ-ba sed decision combina tions. Step 4: Determine the inventive problem-solving model using innova tion life cycle with integer pa tterns.
Step 5: Collect the linguistic eva lua tions for green energy project processes in terms of innova tion life cycle.
Step 6: Select the combina tions for TRIZ-ba sed principles. Step 7: Identify the best combina tions with the integer cod e series.
Pha se 3: Mea suring the a ctivity priorities of TRIZ-ba sed principles for green energy projects with PF TOPSIS a nd PF VIKOR. Additiona lly, the a na lysis ha s a lso been ma de by considering SF sets.
Step 8: Construct the fuzzy preferences of TRIZ-ba sed principles for decision ma trix.
Step 9: Ca lcula te Pytha gorea n fuzzy rela tion ma trix. Step 10: Compute the defuzzified va lues of decision ma trix with score function.
Step 11: Norma lize the decision ma trix.
Step 12: Ca lcula te the weighted decision ma trix.
Step 13: Ra nk the project a ctivity priorities of TRIZ-ba sed principles with the va lues of CCi.
Pha se 4: Scoring the green energy project ba sed on innova tion life cycle using inventive problem-solving principles by PF a nd SF DEMATEL.
Step 14: Collect the linguistic eva lua tions of TRIZ-ba sed principles for rela tion ma trix.
Step 15: Compute the a vera ge fuzzy preferences for direct rela tion ma trix.
Step 16: Compute Pytha gorea n fuzzy rela tion ma trix. Step 17: Construct the defuzzified rela tion ma trix. Step 18: Norma lize the rela tion ma trix.
Step 19: Determine the immedia te Predecessors of a ctivities with tota l rela tion ma trix.
Step 20: Construct the green energy project network ba sed on inventive problem-solving model.
Step 21: Collect the linguistic eva lua tions for a ctivity dura tion a nd costs.
Step 22: Ca lcula te the a vera ge fuzzy va lues for a ctivity dura tion a nd costs.
Step 23: Mea sure the pa th performa nces of green energy projects.
The ma in novelty of this proposed model is considering a hybrid methodology. This mea ns tha t different MCDM models (TOPSIS a nd DEMATEL) a re used in the a na lysis process [6]. In other words, both ra nking the fa ctors a nd ma king the ca usa l rela tionships a re performed objectively [7]. On the other side, in some models in the litera ture, the a lterna tives a re ra nked with a MCDM model, but the a uthors selected the weights [64,65]. Hence, by compa ring with these models, the proposed model provides better results [66,67]. Additiona lly, defining the principles for green energy project a ctivities ba sed on TRIZ technique is a nother novelty of this model beca use it provides innova tive solutions [8,9]. In other words, specific stra tegies a re developed for the investment decisions. This situa tion provides a competitive a dva nta ge for the green energy investors.
Also, considering PF sets ha s some benefits. With respect to the intuitionistic fuzzy sets, the sum of membership a nd non-membership degrees should be ma ximum 1 [68-71]. However, rega rding the Pytha gorea n fuzzy sets, there is not such a necessity so tha t owing to these fuzzy sets, uncerta inties ca n be reflected in a more suita ble wa y [10-12]. Additiona lly, the ma in a dva nta ge of TOPSIS is tha t more a ppropria te results ca n be rea ched due to considering a lso nega tive optima l solutions [13-15]. However, in some simila r methods, the shortest dista nces to positive optima l solutions a re only ta ken into considera tion [72-74]. Thus, it is clea r tha t TOPSIS helps to rea ch more effective ra nking results.
Moreover, the green energy project network is genera ted by using impa ct-rela tion ma p genera ted by DEMATEL a na lysis so tha t more objective results ca n be rea ched [21,22]. However, in most of MCDM techniques, only the weights ca n be ca lcula ted, but the ca usa lity a na lysis ca nn o t be identified. Therefore, this proposed model ca n genera te more effective a nd relia ble stra tegies in compa rison with previous models [75-80]. Fina lly, considering integer pa tterns to define the best combina tions of TRIZ-ba sed principles for innova tion life cycle ha s a contribution to increa se the relia bility of the eva lua tions for the pa tterns [23-26]. Both VIKOR a nd TOPSIS a pproa ches a re ta ken into considera tion. Simila rly, the a na lyses a re a lso performed with SF sets in a ddition to the PF sets. With the help of these compa ra tive eva lua tions, it ca n be possible to mea sure the relia bility. Therefore, this proposed model ha s significa nt superiorities by compa ring with the previous models tha t include only one MCDM a pproa ch.
Furthermore, this proposed model is a lso quite a ppropria te the purpose of the ma nuscript. A novel inventive problem-solving model is crea ted for green energy project eva lua tion. In this context, green energy project a ctivities a re genera ted by considering TRIZ-ba sed principles. Green energy investment projects ha ve ma ny benefits. However, beca use of high-cost problems, it becomes quite difficult to increa se these projects. Hence, in the a na lysis process, cost effectiveness should be ta ken into considera tion. TRIZ is a
technique which provides innova tive a nd cost-effective solutions. Thus, preferring TRIZ in this fra mework contributes to more a ppropria te results. In a ddition to this condition, in this study, these a ctivities a re ra nked to find the best process in green energy project development. Beca use it is a crucia l situa tion, both TOPSIS a nd VIKOR methods a re considered in the ra nking process. With the help of this compa ra tive eva lua tion, it ca n be possible to ha ve more relia ble results. Moreover, the green energy project network is genera ted in this study, a s well. Due to this situa tion, DEMATEL method is considered beca use of the impa ct rela tion ma p.
On the other ha nd, there a re a lso some limita tions of this proposed model. In this model, the criteria a re weighted by considering only DEMATEL methodology. In order to ma ke a compa ra tive eva lua tion, both SF a nd PF sets a re ta ken into considera tion. However, in the proposed model, there is not a compa ra tive a na lysis for DEMATEL method. Another limita tion of the proposed model is tha t the a ctivities a re defined by considering TRIZ-ba sed principles. In this process, the experts a re requested to genera te items by considering only 39 different TRIZ principles. Fina lly, immedia te predecessors a re determined with DEMATEL. Beca use of this situa tion, the pa ths a re limited to only 4 different a ctivity sets.
IV. ANALYSIS RESULTS
The proposed model consists of four different pha ses. Ana lysis results a re given for ea ch pha se.
PHASE 1: DETERMINING THE TRIZ-BASED
PRINCIPLES FOR GREEN ENERGY PROJECTS
In this pha se, firstly, the cha ra cteristics of green energy projects a re defined for TRIZ technique. Ba sed on this eva lua tion, 6 different items a re selected out of TRIZ-ba sed engineering pa ra meters. Ta ble 1 indica tes the deta ils of the selected cha ra cteristics.
TABLE 1.The characteristics of green energy projects for TRIZ
Characteristics Literature
Loss of energy (F1) [19],[20] Relia bility (F2) [27],[28] Convenience of use (F3) [29],[30] Repa ira bility (F4) [31],[32] Ada pta bility (F5) [19],[30]
Ca pa city (F6) [27],[32]
Firstly, necessa ry ca re should be ta ken to a void energy loss in green energy projects. Otherwise, the efficiency of these projects will decrea se significa ntly. On the other ha nd, the most importa nt point in these projects is tha t they do not ha rm the environment. In this context, it should be ensured tha t people's confidence in these projects is increa sed by ta king necessa ry mea sures. In a ddition, ca re should be ta ken to ensure tha t these projects a re user-friendly. In this wa y, it will be possible for the products to be preferred more by the customers. In a ddition, in ca se of a potentia l problem in the project, this ma lfunction should be repa ired quickly. In this wa y, uninterrupted electricity will be produced. Very ra pid developments a re occurring in green energy technologies. Therefore, it is importa nt to design projects so tha t they ca n be a da pted very ea sily to these developments. Fina lly, the electrica l energy to be obta ined must ha ve a high ca pa city.
After tha t, the contra diction ma trix is constructed for green energy projects. In this fra mework, 3 different experts made eva lua tions. These people consist of a ca demicia ns a nd ma na gers who work a t the level of directors a t lea st.The expert tea m a lso ha s a minimum of 20 yea rs of experience in the subject. These people ha ve the knowledge to ma ke effective eva lua tions for green energy projects. With respect to the crea ting of the contra diction ma trix (CM), the expert tea m ma de a compa ra tive eva lua tion for these 6 different fa ctors. The left side of this ma trix gives informa tion a bout the improvement of these items wh e re a s the right side sta tes the worsening conditions. As a result, experts identified significa nt TRIZ-ba sed principles a s in Ta ble 2.
TABLE 2.CM for green energy projects
Worsening Characteristics Im p r o v in g C h a ra c ter is ti c s Characteristics F1 F1 - F2 15 15,19 F3 10,15 F4 15,19,35 F5 10,15,19 F6 F2 10,15,19 - 10,15,19 10,15,19 19,35 15,19 F3 10,15 10,15,19 - 15,19 10,15 10,35 F4 10,15,19 10,15,19 10,15,19,35 - 10 15,35 F5 10,15 10,15,35 15,19,35 10,19,35 - 15,19,35 F6 15,19,35 10,15,19 15,19,35 10,15 15,19,35 -
It is obvious tha t by considering the eva lua tions empha sized in Ta ble 2, 4 different issues a re identified out of 40 different TRIZ principles. Ta ble 3 gives informa tion a bout these fa ctors.
TABLE 3.TRIZ-based principles for green energy project activities
Principle
Numbers Definition
Supported Literature 10 Prior Action (Principle 1) [19],[20] 15 Dyna micity (Principle 2) [31],[32] 19 Periodic Action (Principle 3) [29],[30] 35 Tra nsforma tion of Properties (Principle 4) [27],[32]
Prior a ction refers to the pre-pla nning of a project. In this context, it is a imed to clea rly identify the issues a ffecting the customers' preferences. On the other ha nd, dyna micity includes the designing the object for the outside environment for the best solution. In this context, improvements a re ma de in the processes of the project by mea suring the customers' rea ctions. In other words, necessa ry a da pta tions should be implemented in green
energy projects, especia lly considering technologica l developments. Moreover, rega rding the periodic a ction, it should be a imed to identify possible ma lfunctions in green energy projects in a timely ma nner by conducting a udits a t irregula r interva ls. Fina lly, the tra nsforma tion of properties refers to cha nges to be ma de in the physica l properties of the object when necessa ry.
PHASE 2: COLLECTING THE TRIZ-BASED COMBINATIONS
The inventive problem-solving model is identified using innova tion life cycle with integer pa tterns in the first step of this pha se. Figure 3 shows the process of inventive problem-solving model using innova tion life cycle with integer pa tterns.
FIGURE 3.Process of inventive problem-solving model using innovation life cycle with integer patterns
Innova tion life cycle is defined in 16 periods of time. Accordingly, selected processes of inventive problem-solving model entitled TRIZ a pproa ch a re a na lyzed in the time of innova tion life cycle, respectively. Figure 4 represents tha t k number of process is given in n periods of innova tion life cycle. Ea ch process is divided into 16 periods of time a nd k number of process is subject to the innova tive resea rches. At the end of project processes, the most a ppropria te inventive problem-solving results could be eva lua ted for the efficient decision-ma king results of green energy projects. In this study, selected principles from the TRIZ method a re considered a s a process set a nd they a re a pplied using innova tion life cycle in 64 periods of time for the inventive problem-solving model of green energy projects. In the next step, the linguistic eva lua tions
a re obta ined for green energy project processes in terms of innova tion life cycle. Ta ble 4 defines the preference numbers a nd integer a lpha bet for decision ma trix.
TABLE 4.Preference numbers and integer alphabet for decision matrix
Linguistic Scales Preference Numbers Integer Alphabet Wea kest .10 -2 Poor .25 -1 Fa ir .50 0 Good .75 +1 Best 1 +2
The eva lua tions rega rding different a ctivities a re obta ined. La ter, the best combina tions a re determined. With respect
to the principa l 1, the ca lcula tions of the 4 different combina tions a re indica ted below.
Combina tion 1: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (1)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (1)0− (1)0 = 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0 = 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (1)1− (2)1+(2)1− (2)1≠ 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (2)1− (2)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (1)1− (1)1+(1)1− (1)1= 0
Beca use 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) ≠ 0, it is seen that the combination 1 does not sa tisfy the innova tion life cycle pa ttern.
Combina tion 2: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (1)0− (1)0 = 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0 = 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (2)1− (2)1+(2)1− (2)1= 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (2)1− (2)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (1)1− (1)1+(1)1− (1)1= 0 At the level 3, 𝑓𝑓[2](𝑡𝑡1, 𝑡𝑡8) = (2)2− (2)2+(2)2− (2)2+ (2)2− (2)2+(1)2− (1)2= 0, 𝑓𝑓[2](𝑡𝑡 9, 𝑡𝑡16) = (1)2− (1)2+(1)2− (1)2+ (1)2− (1)2+(1)2− (1)2= 0 At the level 4, 𝑓𝑓[3](𝑡𝑡1, 𝑡𝑡16) = (2)3− (2)3+(2)3− (2)3+ (2)3− (2)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3= 0
It is concluded tha t the combina tion 2 is consistent a t the level 4 for the geometrica l pa tterns of innova tion life cycle in terms of TRIZ-ba sed principle 1. Accordingly, combina tion 2 is selected a s a best decision combina tion of principle 1. Combina tion 3: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (1)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (1)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (1)0− (1)0 = 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0 = 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (1)1− (2)1+(2)1− (2)1≠ 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (2)1− (2)1+(1)1− (2)1≠ 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1 = 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (1)1− (1)1+(1)1− (1)1= 0
Since 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) ≠ 0 and 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) ≠ 0, it is determined tha t the combina tion 3 does not provide the hiera rchica l form of innova tion life cycle pa ttern a t the level 2.
Combina tion 4: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (1)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0= 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (2)1− (2)1+(2)1− (2)1 = 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (2)1− (2)1+(1)1− (2)1≠ 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1 = 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (1)1− (1)1+(1)1− (1)1= 0
The computa tion process of integer code series ca nnot continue beca use 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) ≠ 0.
Additiona lly, rega rding the principa l 2, the hiera rchica l forms a re a lso computed for other TRIZ-ba sed principles a t the different levels of integer codes. The best decision combina tion of principle 2 is selected a s combina tion 1. The results a re given a s follows.
Combina tion 1: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (2)0− (2)0= 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0= 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (2)1− (2)1+(1)1− (1)1 = 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (2)1− (2)1+(2)1− (2)1= 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1 = 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (1)1− (1)1+(1)1− (1)1= 0 At the level 3, 𝑓𝑓[2](𝑡𝑡1, 𝑡𝑡8) = (2)2− (2)2+(1)2− (1)2+ (2)2− (2)2+(2)2− (2)2 = 0, 𝑓𝑓[2](𝑡𝑡 9, 𝑡𝑡16) = (1)2− (1)2+(1)2− (1)2+ (1)2− (1)2+(1)2− (1)2= 0 At the level 4, 𝑓𝑓[3](𝑡𝑡1, 𝑡𝑡16) = (2)3− (2)3+(1)3− (1)3+ (2)3− (2)3+(2)3− (2)3+ (1)3− (1)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3 = 0
For the principle 3, the best decision set of innova tion life is combina tion 7. The consistency results of this combina tion a re given below. Combina tion 7: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (1)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (2)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (1)0− (1)0 = 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) =
(1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (2)0− (2)0 = 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0 = 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (1)1− (2)1+(2)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (1)1− (1)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (2)1− (2)1+(1)1− (1)1= 0 At the level 3, 𝑓𝑓[2](𝑡𝑡1, 𝑡𝑡8) = (1)2− (2)2+(2)2− (1)2+ (1)2− (1)2+(1)2− (1)2= 0, 𝑓𝑓[2](𝑡𝑡 9, 𝑡𝑡16) = (1)2− (1)2+(1)2− (1)2+ (2)2− (2)2+(1)2− (1)2= 0 At the level 4, 𝑓𝑓[3](𝑡𝑡1, 𝑡𝑡16) = (1)3− (2)3+(2)3− (1)3+ (1)3− (1)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3+ (2)3− (2)3+(1)3− (1)3= 0
Moreover, with respect to the principle 4, the combina tion 7 presents the best decision set of innova tion life cycle with integer codes. Four level computa tion results of this combina tion a re defined a s following.
Combina tion 7: At the level 1, 𝑓𝑓[0](𝑡𝑡1, 𝑡𝑡2) = (1)0− (1)0 = 0, 𝑓𝑓[0](𝑡𝑡3, 𝑡𝑡4) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡5, 𝑡𝑡6) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 7, 𝑡𝑡8) = (1)0− (1)0 = 0, 𝑓𝑓[0](𝑡𝑡9, 𝑡𝑡10) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡 11, 𝑡𝑡12) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡13, 𝑡𝑡14) = (1)0− (1)0= 0, 𝑓𝑓[0](𝑡𝑡15, 𝑡𝑡16) = (1)0− (1)0= 0, At the level 2, 𝑓𝑓[1](𝑡𝑡1, 𝑡𝑡4) = (1)1− (1)1+(1)1− (1)1 = 0, 𝑓𝑓[1](𝑡𝑡5, 𝑡𝑡8) = (1)1− (1)1+(1)1− (1)1= 0, 𝑓𝑓[1](𝑡𝑡9, 𝑡𝑡12) = (1)1− (1)1+(1)1− (1)1 = 0, 𝑓𝑓[1](𝑡𝑡13, 𝑡𝑡16) = (1)1− (1)1+(1)1− (1)1= 0 At the level 3, 𝑓𝑓[2](𝑡𝑡1, 𝑡𝑡8) = (1)2− (1)2+(1)2− (1)2+ (1)2− (1)2+(1)2− (1)2 = 0, 𝑓𝑓[2](𝑡𝑡 9, 𝑡𝑡16) = (1)2− (1)2+(1)2− (1)2+ (1)2− (1)2+(1)2− (1)2= 0 At the level 4, 𝑓𝑓[3](𝑡𝑡1, 𝑡𝑡16) = (1)3− (1)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3+ (1)3− (1)3+(1)3− (1)3 = 0
PHASE 3: MEASURING THE ACTIVITY PRIORITIES OF TRIZ-BASED PRINCIPLES FOR GREEN ENERGY PROJECTS WITH PF TOPSIS
Firstly, the fuzzy preferences of TRIZ-ba sed principles a re constructed for decision ma trix a s in Ta ble 5.
TABLE 5.Fuzzy preferences of TRIZ-based principles for decision matrix
Time Process Phase Principle I Principle II Principle III Principle IV
T1 Emerging Pha se 1 1 1 .75 .75 T2 Pha se 2 .10 .10 .10 .25 T3 Growth Pha se 1 1 .75 1 .75 T4 Pha se 2 .10 .25 .25 .25 T5 Ma turity Pha se 1 1 1 .75 .75 T6 Pha se 2 .10 .10 .25 .25 T7 Aging Pha se 1 .75 1 .75 .75 T8 Pha se 2 .25 .10 .25 .25 T9 Emerging Pha se 1 .75 .75 .75 .75 T10 Pha se 2 .25 .25 .25 .25 T11 Growth Pha se 1 .75 .75 .75 .75 T12 Pha se 2 .25 .25 .25 .25 T13 Ma turity Pha se 1 .75 .75 1 .75 T14 Pha se 2 .25 .25 .10 .25 T15 Aging Pha se 1 .75 .75 .75 .75 T16 Pha se 2 .25 .25 .25 .25
The Pytha gorea n fuzzy rela tion ma trix is genera ted. The norma lized ma trix is constructed by considering the bounda ries of 𝜇𝜇2𝑝𝑝+ 𝑎𝑎𝑝𝑝2 = 1 The matrix is shown in Table 6.
TABLE 6.Pythagorean fuzzy decision matrix
Time Principle I Principle II Principle III Principle IV T1 [.90,.05] [.90,.05] [.68,.16] [.68,.16] T2 [.09,.46] [.09,.46] [.09,.46] [.23,.39] T3 [.90,.05] [.68,.16] [.90,.05] [.68,.16] T4 [.09,.46] [.23,.39] [.23,.39] [.23,.39] T5 [.90,.05] [.90,.05] [.68,.16] [.68,.16] T6 [.09,.46] [.09,.46] [.23,.39] [.23,.39] T7 [.68,.16] [.90,.05] [.68,.16] [.68,.16] T8 [.23,.39] [.09,.46] [.23,.39] [.23,.39] T9 [.68,.16] [.68,.16] [.68,.16] [.68,.16] T10 [.23,.39] [.23,.39] [.23,.39] [.23,.39] T11 [.68,.16] [.68,.16] [.68,.16] [.68,.16] T12 [.23,.39] [.23,.39] [.23,.39] [.23,.39] T13 [.68,.16] [.68,.16] [.90,.05] [.68,.16] T14 [.28,.39] [.23,.39] [.09,.46] [.23,.39] T15 [.68,.16] [.68,.16] [.68,.16] [.68,.16] T16 [.23,.39] [.23,.39] [.23,.39] [.23,.39] After defuzzifica tion a nd norma liza tion,the weights of time a re considered equa lly a s 6.25% a nd weighted decision ma trix is given in Ta ble 7.
TABLE 7.Weighted decision matrix
Time Principle I Principle II Principle III Principle IV T1 .039 .039 .021 .021 T2 .035 .035 .035 .017 T3 .039 .021 .039 .021 T4 .047 .024 .024 .024 T5 .039 .039 .021 .021 T6 .040 .040 .020 .020 T7 .024 .046 .024 .024 T8 .024 .047 .024 .024 T9 .031 .031 .031 .031 T10 .031 .031 .031 .031 T11 .031 .031 .031 .031 T12 .031 .031 .031 .031 T13 .024 .024 .046 .024 T14 .024 .024 .047 .024 T15 .031 .031 .031 .031 T16 .031 .031 .031 .031
In the fina l step, the project a ctivity priorities of TRIZ-ba sed principles a re ra nked with the va lues of CCi. The ra nking results of TRIZ-ba sed principles a re indica ted in Ta ble 8.
TABLE 8.Ranking the priorities of TRIZ-based principles
Principles D+ D- CCi Ranking of Project activity Priorities (Principle 1) .045 .047 .512 (A2) (Principle 2) .044 .049 .527 (A1) (Principle 3) .051 .041 .442 (A3) (Principle 4) .066 .000 .000 (A4) Ta ble 8 sta tes tha t dyna micity is the first a ctivity. After tha t, the prior a ction is considered. Also, the third a ctivity is the periodic a ction. Furthermore, the tra nsforma tion of properties is on the la st ra nk. In a ddition to this situa tion, these items a re a lso ra nked by considering VIKOR. Moreover, the a na lysis is a lso performed by using SF sets in a ddition to PF sets. The compa ra tive a na lysis results a e summa rized in Ta ble 9.
TABLE 9.Comparative ranking results
Principles SF VIKOR SF TOPSIS PF VIKOR PF TOPSIS P1 2 2 2 2 P2 1 1 1 1 P3 3 3 3 3 P4 4 4 4 4
It is defined tha t the findings of both techniques a re a lmost the sa me for both SF a nd PF sets. This situa tion expla ins tha t the findings of PF TOPSIS a re relia ble.
PHASE 4: SCORING THE GREEN ENERGY PROJECT BASED ON INNOVATION LIFE CYCLE USING INVENTIVE PROBLEM-SOLVING PRINCIPLES
Green energy projects a re eva lua ted ba sed on innova tion life cycle a nd TRIZ-ba sed principles using a n integra ted method of PERT a nd PF DEMATEL. PF DEMATEL is used for defining the immedia te predecessors of a ctivity in the project eva lua tion. Critica l pa ths a s well a s the cost a nd dura tion scores a re a ssessed by using PERT. On the other side, the linguistic rela tion eva lua tions for TRIZ-ba sed principles a re sta ted on Ta ble 10.
TABLE 10.Linguistic relation evaluations for TRIZ-based principles
A1 A2 A3 A4 E1 E2 E3 E1 E2 E3 E1 E2 E3 E1 E2 E3 A1 M H H M M M S S M A2 M H H S S M A3 H H M A4
Avera ge fuzzy preferences a re computed a s in Ta ble 11.
TABLE 11.Average fuzzy preferences for direct relation matrix
A1 A2 A3 A4
A1 0.67 0.50 0.33
A2 0.67 0.33
A3 0.67
A4
The norma lized ma trix is constructed by considering the bounda ries of 𝜇𝜇2𝑝𝑝+ 𝑎𝑎𝑝𝑝2= 1 The matrix is shown in Table 12.
TABLE 12.Pythagorean fuzzy relation matrix
A1 A2 A3 A4
A1 [0.45,0.28] [0.45,0.28] [0.23,0.39]
A2 [0.45,0.28] [0.23,0.39]
A3 [0.68,0.16]
A4
The a vera ge va lues of membership a nd non-membership degrees a re computed, a nd the score function va lues a re ca lcula ted for obta ining the defuzzified va lues of Pytha gorea n fuzzy rela tion ma trix. Ta ble 13 shows the deta ils.
TABLE 13.Defuzzified relation matrix
A1 A2 A3 A4
A1 .000 .320 .127 .033
A2 .000 .000 .320 .033
A3 .000 .000 .000 .320
A4 .000 .000 .000 .000
After tha t, the norma lized ma trix is crea ted a s in Ta ble 14.
TABLE 14.Normalized relation matrix
A1 A2 A3 A4
A1
.000
.668
.265
.068
A2
.000
.000
.668
.068
A3
.000
.000
.000
.668
A4
.000
.000
.000
.000
Next, the immedia te predecessors of a ctivities a re determined with tota l rela tion ma trix. The a ctivity impa cts for immedia te predecessors a re defined a s in Ta ble 15.
TABLE 15.Activity impacts for immediate Predecessors
A1 A2 A3 A4
A1
.000
.668
.710
.587
A2
.000
.000
.668
.513
A3
.000
.000
.000
.668
A4
.000
.000
.000
.000
In Ta ble 15, the bold va lues a re the higher va lues tha n threshold. Thus, it is possible to illustra te the immedia te predecessors of green energy projects in Ta ble 16.
TABLE 16.Immediate predecessors of the TRIZ-based principles
TRIZ-based Principles Activity Immediate Predecessors Dyna micity A1 - Prior Action A2 A1 Periodic Action A3 A1, A2
Tra nsforma tion
of Properties A4 A2, A3
In the following step, the green energy project network is genera ted ba sed on inventive problem-solving model. Figure 4 indica tes the flowcha rt of green energy project network.
FIGURE 4.The flowchart of green energy project network
Eva lua tions for a ctivity dura tion a nd costs or provided. Ta ble 17 demonstra tes eva lua tions for a ctivity dura tion a nd costs.
TABLE 17.Linguistic evaluations for activity duration and costs
Duration Costs Activity E1 E2 E3 E1 E2 E3 A1 M H H M VH H A2 H VH M L VH M A3 VH H H VH H H A4 H L M H H M
Additiona lly, the optimistic a nd pessimistic va lues for a ctivity dura tion a nd costs a re demonstra ted in Ta ble 18.
TABLE 18.Optimistic and pessimistic values for activity duration and costs
Duration Costs
Activity Optimistic Pessimistic Optimistic Pessimistic
A1 M H M VH
A2 M VH L VH
A3 H VH H VH
A4 L H M H
After tha t, the a vera ge fuzzy va lues for a ctivity dura tion a nd costs a re ca lcula ted a s in Ta ble 19.
TABLE 19.Average fuzzy values for activity duration and costs
Activity Average fuzzy preferences
Duration Costs
A1 0.67 0.75
A2 0.75 0.58
A3 0.83 0.83
A4 0.50 0.67
La ter, the pa th performa nces of green energy projects a re mea sured a s in Ta ble 20.
TABLE 20.The weights (Ws) of TRIZ-based activities
Activity D E D+E D-E Ws
A1 1.965 .000 1.965 1.965 .258
A2 1.181 .668 1.848 .513 .242
A3 .668 1.378 2.045 -.710 .268
A4 .000 1.768 1.768 -1.768 .232
Ta ble 20 identifies tha t a ctivity 3 (periodic a ction) ha s the grea test weight. Also, the a ctivity 1 (dyna micity) is a nother significa nt principle. Activity 2 (prior a ction) a nd a ctivity 4 (tra nsforma tion of properties) ha ve low weights. Ta ble 21 indica tes the pa th performa nces of green energy projects with respect to the dura tion a nd costs.
TABLE 21.Path performances of green energy projects by duration and costs
Paths Activity Set Weighted preference
of duration Ranking by duration Weighted preference of costs Ranking by costs Pa th 1 A1, A2, A4 .641 1 .668 1 Pa th 2 A1, A3, A4 .675 2 .754 3
Pa th 3 A1, A2, A3, A4 .693 3 .713 2
Activity fuzzy preferences a re multiplied with their weights for the a ctivity set of pa ths. Thus, the weighted preference results a re obta ined to mea sure the pa th performa nces by the dura tion a nd costs. It is concluded tha t pa th 1 is the
shortest pa th by dura tion. On the other side, pa th 3 is the longest pa th by dura tion/CPM. Additiona lly, pa th 1 ha s the lowest cost wherea s pa th 2 ha s the highest cost. These ca lcula tions a re a lso ma de by using SF sets. Ta ble 22 expla ins the deta ils of these ca lcula tions.
TABLE 22.Comparative ranking results by duration and costs
Paths Activity Set Ranking by
duration (SF) Ranking by costs (SF) Ranking by duration (PF) Ranking by costs (PF) Pa th 1 A1, A2, A4 1 1 1 1 Pa th 2 A1, A3, A4 2 3 2 3
Pa th 3 A1, A2, A3, A4 3 2 3 2
Ta ble 22 demonstra tes tha t the ra nking results a re the sa m e for both SF a nd PF sets. Hence, the findings a re consistent.
V. CONCLUSION
It is a imed to identify green energy project network ba sed on innova tion life cycle by using PF sets with integer pa tterns. For this purpose, four pha se-hybrid decision ma king a pproa ch is a pplied by considering the TRIZ, integer code series, PERT, Pytha gorea n fuzzy TOPSIS a nd DEMATEL, respectively. The novelty of this study is to propose a hybrid decision support system for green energy projects a nd integra te the inventive problem-solving model into the PERT using PF sets properly. TRIZ technique is used for designing the inventive problem-solving model, PERT is a pplied for illustra ting the critica l pa ths of green energy projects, PF TOPSIS is employed for defining the a ctivity priorities of green energy projects, a nd PF DEMATEL is computed for figuring out the immedia te predecessors of a ctivities. Additiona lly, TRIZ-ba sed principles a re a lso ra nked by using VIKOR. Moreover, a ll ca lcula tions ha ve a lso been ma de with Spherica l fuzzy (SF) sets in a ddition to the PF sets. The results expla in tha t dyna micity is the initia l a ctivity of green energy projects while tra nsforma tion of properties is the fina l a ctivity. However, dyna micity, prior a ction, a nd tra nsforma tion of properties a re the set of a ctivities with the shortest pa th by dura tion a s well a s the lowest cost.
VI. LIMITATIONS AND IMPLICATIONS
It is identified tha t dyna micity a nd tra nsforma tion of properties ta ke pla ce in a ll pa ths. This situa tion gives informa tion tha t these a ctivities a re crucia l in this condition. By considering these issues, it is understood tha t for the
best solution, products should be designed in a ccorda nce with the externa l environment. Green energy investments ha ve high initia l costs. This situa tion crea tes a ba rrier for investors to focus on this a rea a s it nega tively a ffects profita bility. In this context, technologica l developments rega rding green energy investments crea te a n opportunity to reduce these costs. Therefore, green energy investors should closely follow technologica l developments in this a rea . The importa nt point here is tha t the products used in green energy investments should be ea sily a da pted to these developments. Otherwise, these technologica l developments will not be ea sily a pplica ble to projects. Thus, compa nies tha t ca nnot ga in a cost a dva ntage will a lso lose a significa nt competitive a dva nta ge. In this context, green energy compa nies must be rea dy for technologica l developments tha t will emerge in every sense. Hence, depa rtments within the compa ny should be designed to be a ble to rea ct quickly. In this wa y, it will be possible for the compa ny to a da pt to innova tions very quickly. Simila rly, Må nberger et a l. [81], Ga lla gher et a l. [82] a nd Egli et a l. [83] ma de a n eva lua tion to understa nd the significa nt indica tors to a ffect the performa nce of the green energy investment projects. They underlined tha t these compa nies should follow technologica l improvements to ha ve susta ina ble success in these investments.
Another importa nt result of this study is tha t the principa l of prior a ction is importa nt to reduce the costs instea d of periodic a ction. It is essentia l to ma ke the necessa ry controls before the project sta rts in order to increa se the efficiency a nd productivity. The inspections to be ma de a fter the project is implemented will contribute to the identifica tion of problems. However, correcting these problems will cost the compa ny. In this context, these problems should be reduced a s much a s possible in order to
reduce these costs. The most helpful a spect to this situa tion is tha t the necessa ry controls a re ma de before the project is implemented. It is vita l tha t costs ca n be reduced. In the litera ture, lots of resea rchers highlighted the significa nce of this situa tion. For insta nce, Kha n et a l. [84], Adefa ra ti a nd Ba nsa l [85] a nd Bellotti et a l. [86] defined tha t prelimina ry pla nning pla ys a crucia l role to reduce the costs of these projects so tha t the efficiency a nd the effectiveness of the se projects ca n be increa sed.
The ma in contribution is tha t innova tive stra tegies a re identified for green energy investment projects with a novel hybrid MCDM model ba sed on integer pa tterns a nd Pytha gorea n fuzzy sets. Nonetheless, there a re a lso some limita tions of the proposed model in this study. The ma in limita tion is tha t innova tive fa ctors a re defined by only TRIZ principles. Hence, in the future studies, different
methodologies ca n be considered to determine the criteria . For insta nce, the criteria ca n be genera ted by considering SWOT a na lysis. Hence, different fa ctors ca n be considered a t the sa me time. Also, the results of DEMATEL a re not compa red with other MCDM models. In the future studies, a na lytic hiera rchy process (AHP) ca n a lso be considered so tha t it ca n be possible to ma ke robustness check. Additiona lly, immedia te predecessors a re determined with DEMATEL. Due to this issue, the pa ths a re limited to only 4 different a ctivity sets. On the other ha nd, a more specific a na lysis of green energy types ca n be ma de. In a ddition to this condition, different a pproa ches ca n a lso be used in the a na lysis process. The results could be widened for the future studies by using the different MCDM a pproa ches such a s Entropy a nd considering the cross-industria l a na lysis.
