Model of fuzzy logic for selection infrastructural investment project of wind farm locations.

Aljicevic, Zajim ; Kostic, Aleksandra ; Dautbasic, Nedis 等

1. Introduction

Increasing environmental concern during the 21th Century has moved the research focus from conventional electricity sources to renewable and alternative energy solutions. In renewable power generation, wind energy has been noted as the fastest-growing energy technology in the world, due to the fact that world has enormous resources of wind energy. Bosnia and Herzegovina also has a big wind potential which need to be used.

At Global Wind Energy Council (GWEC, in April 2016, it is noted that wind power had another record-breaking year. After passing the 50 GW mark for the first time in a single year in 2014, it reached another milestone in 2015 as annual installations topped 63 GW, a 22% increase. By the end of last year, there were about 433 GW of wind power spinning around the globe, a cumulative 17% increase; and wind power supplied more new power generation than any other technology in 2015, according to the IEA[1] (Source: GWEC--Global Wind Energy Council)

Many studies have been proposed to optimize wind power generation. Shata and Hanitsch [2] have evaluated different wind sites in Egypt using classical statistical analysis to estimate the best wind site. Kogak [3] focused entirely on wind speed persistence during weather forecast, site selection for wind turbines and synthetic generation of the wind speed data. Herbert et al [4] developed models for wind resources assessment, site selection and aerodynamic including wake effect to improve the wind turbine performance and to increase its productivity. Lackner et al [5] utilized ground-based measurement devices instead of meteorological towers for wind resource assessment. They investigated the use of a monitoring strategy in more than one site to determine the best wind sites.

This paper utilizes fuzzy logic to evaluate various sites in the bases of their benefits and costs to adopt the higher priority sites in terms of different criteria. The paper is organized so that in Section 2 is given a general overview of the installed capacity of wind power. In Section 3 are given wind characteristics in Bosnia and Herzegovina. In Section 4 is shown fuzzy logic methodology. In Section 5 are presented fuzzy logic rules. In the sixth Section is explained defuzzification. The conclusion was made verification using data on existing wind farm in Podvelezje, and plan future research.

2. Renewable resources

Today, new renewable resources provide only a small share of global energy production (Figure 1). However, the global market for wind power has been expanding faster than any other sources of renewable energy. The world wind power capacity has been duplicated around twenty five times from just 17.4 GW in 2000 to reach over than 432.8 GW at the end of 2015 (Figure 2).

In 2015, increase in wind generation was equal to almost half of global electricity growth. This was surprising but welcome news. It became apparent from new IEA analysis revealing that, for the second successive year, global CO2 emissions remained stable despite growth in the world economy.

It is generally known that we need to invest in wind power. The main problem is how to choose the best location for investment in wind power plants.

This paper utilizes fuzzy logic to evaluate various sites in the bases of their benefits and costs to adopt the higher priority sites in terms of different criteria, i.e., same benefits factors are: wind resources, prevailing wind direction, above ground level, site capacity, site accessibility, soil conditions, site elevation. Cost factors are: land cost, land roughness, temperature, cultural and environmental concerns, aviation/telecommunications conflicts, nearby resident's concerns, site environmental issues (corrosion, humidity), and distance to transmission line, as is used in [6].

3. Wind characteristics in Bosnia and Herzegovina

In the previous period in Bosnia and Herzegovina wind characteristics have been measured at weather stations. Study of Energetic sector [7] shows average annual wind speed at height 50 m above ground for the period 1997-2006. (Figure 3)

Therefore, the first analysis, research and examinations of wind power potential are linked to the area of south Bosnia and Herzegovina. In April 2002 at the location Podvelezje Mostar the first measuring station with appropriate equipment is installed, and then the other ones followed. Inclusive to 31.12.2007, at 13 areas the total number of locations where measuring equipment has been installed is 33. [8-10].

4. Fuzzy logic methodology for decision making

Fuzzy logic has a wide range of utilizations in decision making since it condenses a large amount of parameters into smaller fuzzy sets [11].

In this paper, we have selected Mamdami method because of its closeness to human understanding, since we had a large number of rules in decision-making.

The Fuzzy input/ output combination is shown in the Figure 4.

The fuzzy logic decision selection of the sites options was applied according to benefits and costs:

Benefits:

B1 = Wind resource,

B2 = Prevailing wind direction,

B3 = Above ground level (AGL) (m),

B4 = Site capacity,

B5 = Site accessibility,

B6 = Soil conditions,

B7 = Site elevation.

Costs:

C1 = Land cost,

C2 = Land roughness,

C3 = Temperature ([degrees]C),

C4 = Cultural and environmental concerns,

C5 = Aviation/Telecommunications conflicts,

C6 = Nearby resident's concerns,

C7 = Site environmental issues,

C8 = Distance to transmission line (m).

The fuzzy logic methodology is applied taking into account each site parameters for location Podvelezje [8-10]. The inputs in tables 1 and 2 are considered to be the fuzzy variables, each variable can vary over a fixed range.

The linguistic variables used in the fuzzy methodology are: Very low (VL), Low (L), Normal (N), High (H), Very high (VH), Poor, Marginal, Satisfactory, Good, Excellent, Rock, Mostly rock, Rock/soil, Mostly soil, Soil, Moderate, None, Extensive, Negligible, Minor, Average, Very close, Close, Not far, Far, Very far. Each fuzzy set is addressed as listed in Table 3.

5. Constructing fuzzy rules

Sixty five rules were used in the current fuzzy method implementation to predict the most preferable options or option out of the Podvelezje site.

Fuzzy logic enabled us to dense large amount of data, collected to compare between different sites, into a smaller set of variable rules, to make a decision in the basis of their merits and barriers to produce higher power output at low cost as well as to capture as maximum wind power as possible.

The benefit to cost ratio is shown in Table 4.

6. Performing fuzzy interface

Finally, the mapping process takes place to provide the final decision as shown in Figure 5.

Defuzzification process assesses the outcome in the decision making process. The parameters are analyzed individually, and in this process it is possible to observe two parameters and decide which of them more influenced our results.

This is especially important and helpful in choosing a location for a wind farms. It helps in determining the optimal location. For example in the case that we have two locations with similar parameters then based on those parameters which are different we can make analysis and bring the correct conclusion that the location is favorable.

Also, on the basis of this method it is possible to perform a number of conclusions. For example analysis parameters, land cost and distance to transmission line can be concluded that the land cost is much more important in the decision-making process (Figure 5). The image is clear that the parameter 8 (distance to transmission line) has very little effect, and its only effect is that if the transmission line is over a distance of 10 km, or more than 10 km while any change in parameter C1 (land cost) leads to changes in the final results.

7. Conclusion

The purpose of this paper is to use fuzzy logic and MATLAB for predict the best location for wind farm.

In deciding the used 15 input parameters, where seven parameters presents benefits and eight parameters represents costs. Used a total of 65 rules for decision making.

This model includes all relevant parameters to determine the optimal location for the construction of a wind farm. The advantage of the model is the possibility of determining the advantages of a parameter to the second parameter to a certain value. It is notable that in some parameters higher changes their values do not lead to large differences in profitability of a wind farm until some parameters and their values minor changes lead to great changes in profitability.

For verification of our solutions, we used data on existing wind farm in Podvelezje. Based on our model, we concluded that the choice of location in this case was quite good. Normalized benefit relative weight 0.905, normalized cost relative weight 0.816), or by defined outputs it is between high and very high.

In further research is necessary to collect all the relevant information for potential locations where it is possible to build wind parks. These data would be used in the current model presented in this paper to determine the viability of investments in the wind park.

DOI: 10.2507/27th.daaam.proceedings.107

8. References

[1] http://www.gwec.net, (2016). GWEC-Global Wind 2015 Report, Accessed on: 2016-09-27

[2] Shata, A.S. &Hanitsch, R. (2006). Evaluation of wind energy potential and electricity generation on the coast of Mediterranean Sea in Egypt, Renewable Energy, Vol.31, No.8, page numbers (1183-1202), 0960-1481

[3] Kogak, K. (2008). Practical ways of evaluating wind speed persistence, Energy, Vol.33, No.1, page numbers (6570), 0360-5442

[4] Herbert, G.M.; Iniyan, S.; Sreevalsan, E. &Rajapandian, S.A. (2007). Review of wind energy technologies, Renewable and Sustainable Energy Reviews, Vol.11, No.6, page numbers (1117-1145), 1364-0321

[5] Lackner, M.A.; Rogers, A.L. &Manwell, J.F. (2008). The round robin site assessment method: A new approach to wind energy site assessment, Renewable Energy, Vol.33, No.9, page numbers (2019-2026), 0960-1481

[6] Badran, O.; Abdulhad, E. &El-Tous, Y. (2011). Fuzzy Logic Controller for Predicting Wind Turbine Power Generation, International Journal of Mechanical and Materials Engineering, Vol.6, No.1, page numbers (51-66), 2198-2791

[7] Study of Energetic sector in B&H, Energetski institut Hrvoje Pozar--Croatia, Soluziona--Spain, Ekonomski institut Banjaluka,--BH, Rudarski institut Tuzla--BH, 2008

[8] Elvir, Z.; Mehmed, B.; Fuad, C. &Enes, S. (2008). Energy resources from wind energy in Bosnia & Herzegovina, current state and prospects, 12 International Research/Expert Conference "Trends in the Development of Machinery and Associated Technology" TMT 2008, Istanbul, Turkey

[9] Behmen, M.; Zlomusica, E. & Catovic, F. (2004). The Influence of Electro-Supply Capacities on the State of the Environment in B&H, Status and Perspectives, 8th International Research/Expert Conference Trend in the Development of Machinery and Associated Technology-TMT 2004, Neum, Bosnia and Herzegovina

[10] Catovic, F.; Behmen, M. & Zlomusica, E. (2004) Trends in the development of the electric power systems based on wind energy in world and in Bosnia and Herezegovina, Journal of Enviromental Protection and Ecology-Official Journal of the Balkan Enviromental Association (B.EN.A), Vol. 5, No.4, page numbers (836-840), 1311-5065

[11] Nyrkov, A; Chernyi, S.; Zhilenkov, A. & Sokolov, S. (2016). The use of Fuzzy Neural Structures to Increase the Reliability of Drilling Platforms, Proceedings of the 26th DAAAM International Symposium, pp.0672-0677, B. Katalinic (Ed.), Published by DAAAM International, ISBN 978- 3-902734-07-5, ISSN 1726-9679, Vienna, Austria

Caption: Fig. 3. Average annual wind speed at height 50 m above ground for the period 1997-2006

Caption: Fig. 4. Fuzzy input/output combination

Caption: Fig. 5. Fuzzy implementation sequence Table 1. Overall fuzzy weights for the selected sites based on benefits B1 B2 B3 B4 B5 B6 B7 Relative weight Podvelezje 1 0.7 50 0.8 0.9 0.3 825 0.737 (BiH) Normalized relative weight Podvelezje 0.905 (BiH) Table 2. Overall fuzzy weights for the selected sites based on costs C1 C2 C3 C4 C4 C6 C7 C8 Relative weight Podvelezje 0.2 0.1 15 0.65 0.1 0.4 0.5 10 0.394 (BiH) Normalized relative weight Podvelezje 0.816 (BiH) Table 3. Fuzzy sets Parametres Symbol Variable Linguistic variables type 1 2 Wind resource B1 INPUT VL L Prevailing B2 INPUT Poor Marginal wind direction Above ground B3 INPUT VL L level Site capacity B4 INPUT Poor Marginal Site B5 INPUT Poor Marginal accessibility Soil B6 INPUT Rock Mostly conditions rock Site B7 INPUT VL L elevation Land cost C1 INPUT VL L Land C2 INPUT VL L roughness Temperature C3 INPUT None Moderate Cultural and C4 INPUT None Moderate environmental concerns Aviation/ C5 INPUT None Moderate Telecommunications conflicts Nearby C6 INPUT Negligible Moderate resident's concerns Site C7 INPUT None Minor environmental issues Distance to C8 INPUT Very close Close transmission line B OUTPUT VL L C OUTPUT VL L Parametres Linguistic variables 3 4 5 Wind resource N H VH Prevailing Satisfactory Good Excellent wind direction Above ground N H VH level Site capacity Satisfactory Good Excellent Site Satisfactory Good Excellent accessibility Soil Rock/Soil Mostly Siol conditions soil Site N H VH elevation Land cost N H VH Land N roughness Temperature Extensive Cultural and Extensive environmental concerns Aviation/ Extensive Telecommunications conflicts Nearby Extensive resident's concerns Site Average Moderate Extensive environmental issues Distance to Notfar Far Veryfar transmission line N H VH N H VH Parametres Range Wind resource 0-1 Prevailing 0-1 wind direction Above ground 10-80 m level Site capacity 0-1 Site 0-1 accessibility Soil 0-1 conditions Site 0-2100 m elevation Land cost 0-1 Land 0-1 roughness Temperature -40-40[degrees]C Cultural and 0-1 environmental concerns Aviation/ 0-1 Telecommunications conflicts Nearby 0-1 resident's concerns Site 0-1 environmental issues Distance to 0-30 km transmission line 0-1 0-1 Table 4. Benefit to cost ratio Site Normalized Normalized B/C Normalized B/C benefit cost relative relative weight weight Podvelezje 0.905 0.816 1.109 0.671 (BiH) Fig. 1. Shares of world electricity generation (EIA, 2015). Coal/Peat (41.3%) Natural Gas (21.7%) Hydro (16.3%) Nuclear (10.6%) Oil (4.4%) Others (Renew.) (5.7%) Note: Table made from pie chart. Fig. 2. Global cumulative wind power capacity (Source: GWEC 2015) GLOBAL CUMULATIVE INSTALLED WIND CAPACITY 2000-2015 2000 17,400 2001 23,900 2002 31,100 2003 39,431 2004 47,620 2005 59,091 2006 73,957 2007 93,9254 2008 120,690 2009 159,016 2010 197,946 2011 238,089 2012 282,842 2013 318,463 2014 369,705 2015 432,883 Source: GWCC Note: Table made from bar graph.