International Journal of Energetica (IJECA) https://www.ijeca.info ISSN: 2543-3717 Volume 4. Issue 1. 2019 Page 01-10 IJECA-ISSN: 2543-3717. June 2019 Page 1 Selection of parameters to predict dew point temperature in arid lands using Grey Theory: a case study of Iran Mojtaba Qolipour 1 , Ali Mostafaeipour 1 , Mostafa Rezaei 1* , Elham Behnam 1 , Hossein Goudarzi 2 , Ali Razmjou 3 1Industrial Engineering Department, Yazd University, Yazd, IRAN. 2School of Architecture and Planning, University of New Mexico, NM, USA 3Escolla Técnica Superior d’Enginyeria Industrial de Barcelona, Universitat Politècnica de Catalunya, Barcelona, Catalunya, SPAIN Email * : mm.sr6870@yahoo.com AbstractDew point is the temperature at which water vapor in the air condenses into liquid with the same rate it evaporates. Dew point study is important in arid lands with low rainfall, also in other regions with various hydrological and climatological conditions. In this study, the Grey theory is applied for the first time to propose a framework approach to identify the important parameters affecting the prediction of dew point temperature. The ability of Grey theory to estimate and rank the parameters of a problem with missing data and uncertain conditions means that it has a good potential for mentioned application. For this research, 8 parameters are selected using literature review including: global solar radiation on a horizontal surface (H), water vapor pressure (VP), atmospheric pressure (P), sunshine duration (n), minimum air temperature (Tmin), maximum air temperature (Tmax), average air temperature (Tavg), and Relative Humidity (RH). The study is conducted for the city of Abadeh in Iran by using the data pertaining to a 10 year period between 2005 and 2015. The findings show that RH, Tavg, P, Tmax, Tmin, H, n and Vp with the grey possibility degrees of, respectively, 0.534, 0.551, 0.608, 0.622, 0.635, 0.695, 0.697 and 0.712, are the most important and effective parameters in prediction of dew point temperature. The proposed method also prioritizes the studied parameters in the order of their effectiveness on predicted dew point temperature. Keywords:Dew Point Temperature; Grey Theory; Daily Dew Point Temperature; Fuzzy; Abadeh city. Received: 01/01/2019 – Accepted: 27/02/2019 Nomenclature DPT Dew Point Temperature G Fuzzy Number ANN Artificial Neural Network )(X G  Fuzzy Membership Function ANFIS Adaptive Neuro Fuzzy Inference System )( GL  Length of grey number ],[ GG SVR Support Vector Regression iD i-th Grey's dimension SVM Support Vector Machine Parameters ELM Extreme Learning Machine H Horizontal global solar radiation MLP Multi-Layer Perceptron n Sunshine Hour RBNN Radial Basis Neural Networks Vp Water Vapor Pressure GRNN Generalized Regression Neural Networks P Atmospheric Pressure CANFIS Coactive Neuro Fuzzy Inference system Tmin Minimum Air Temperatures Abder Image placée Mojtaba Qolipour et al IJECA-ISSN: 2543-3717. June 2019 Page 2 MLR Multiple Linear Regression Tmax Maximum Air Temperatures GEP Gene Expression Programming Tavg Average Air Temperatures Greece Letters RH Relative Humidity w Grey's Attribute weight Td Daily Dew PointTemperature r Grey's Attribute range T Temperature xi The i-th value of variable ''x'' xmin The minimum value of variable ''x'' Xo Normalized value of variable ''x'' xmax The maximum value of variable ''x'' I. Introduction Dew point is the temperature at which water vapor in the air (at constant barometric pressure) condenses into liquid with the same rate it evaporates. When air temperature goes below the dew point temperature, some portion of the water vapor in the air condenses until reaching a new equilibrium. The water vapor condenses on a solid surface is called dew [1, 2]. In fact, dew point temperature (Td) is the temperature below which the air can no longer hold its water vapor. In this process, pressure (P) and mixing ratio (r) of moist air remain constant [3, 4]. When air pressure is constant and no water vapor gets added or removed from the air, the mixing ratio of air remains constant. At the dew point temperature, water content of the air is at saturation point; therefore throughout the process the mixing ratio remains equal to its value at the point of saturation (at dew point temperature). In other words, at this temperature, the relative humidity is always 100% [5]. The presence of a relatively cool layer of air near the ground at night can push the temperature down to dew point. A basic method of measuring dew point is to acquire a container with a fully smooth outer wall, fill it with a mixture of water and ice, and then measure the temperate of water at the moment when water droplets (dews) gradually appear on the outer wall [6]. As regards the high importance of dew, many researches have been done into it. Several are summarized in the following. Meng and Wen [7] investigated the amount, frequency, and also duration of dew events in two croplands, arid and sub-humid, in order to quantify the characteristics of dew events and analyze the underlying mechanism of dew formation in different ecosystems. Their finding show that dew events play a more important role in crop growth in arid areas in comparison to sub-humid areas considering the dew occurrence frequency and the amount per night. In another research, Zhuang and Zhao [8] scrutinized the effects of dew events on non-mucilaginous seed germination of annual desert plant species during dry seasons. Finally the results showed that the presence of dew significantly increased seed germination percentages and decreased the nonviable seed percentages. Wang et al. [9] studied dew formation, amount, and duration of rain-fed jujube trees in a semi- arid loess hilly region of China. In this regard they used dew intensity and duration, relative humidity, temperature, and wind speed. The results indicated that atmospheric conditions of relative humidity of >78% and dew point temperature of 1°C–3°C are significantly favorable to dew formation. Pan and Wang [10] performed measurement of accumulated dew formation amount and duration in order to determine the dew formation on different kinds of soils associated with various shrub species and microhabitats. The results showed that the accumulated dew formation amount was four times larger at open spaces as compared to under the canopy, and it was nearly twice as much under living Artemisia ordosica plants as compared to under living Caragana korshinskii plants. These mentioned studies show that dew is of significant importance and needs more investigation into its different aspects, as a result, it is attempted to predict dew point temperature in this study. The accurate and reliable data for dew point is very important in agronomical and climatological related research works. Usually, dew point is used with relative humidity for identifying of the moisture level in the air. Dew point can also be used for providing the estimation of the humidity for near-surface of plants that influences the stomatal closure. Dew point study is important in arid lands with low rainfall, also in other regions with various hydrological and climatological conditions [11]. The majority of studies on dew point temperature prediction and evaluating the factors affecting this process have used a variety of artificial neural networks or in other words the soft computing methods. Some researchers have employed the soft computing techniques to evaluate and determine the most important parameters affecting the dew point prediction for a variety of case studies. The following researches were also found in the literature. Amirmojahedi et al. [11] utilized a hybrid computational intelligence method for predicting dew. Mojtaba Qolipour et al IJECA-ISSN: 2543-3717. June 2019 Page 3 They used a hybrid approach including the extreme learning machine (ELM) with wavelet transform (WT) algorithm for prediction of daily dew point temperature for city of Bandar Abbas in Iran. In a study by Baghban et al. [2], the dew point of atmospheric moist air was estimated by LSSVM and ANFIS methods. The model developed in this study was tested with 100 data instances, and the MSE and R- Square (R 2 ) of results were found to be 0.000016, 1.0000 for LSSVM model and 0.382402, 0.9987 for ANFIS model respectively. In a study by Sarkar [12], a new theoretical formulation for dew point temperatures was developed with the use of modified Clausius–Clapeyron equation. The developed formulation was aimed for comfort air- cooling systems. This study reported that the error of proposed method is far less than that of standard psychrometric measures. These authors also used the DPT equation to plot the wet sensible and latent loads of a cooling dehumidifying coil based on discretized temperature and humidity. Evaluation and determination of the most important and most effective parameter in prediction of dew point temperature with due consideration given to availability and reliability of data is of utmost importance. The majority of research conducted to determine the most effective parameter in dew point temperature prediction has neglected an issue, and that is the uncertainty in the accuracy of data pertaining to parameters and the number of samples available for each parameter. Although most studies have used an acceptable number of samples, the accuracy of the data used in earlier researches and the effect of human and instrument error in data recorded and collected by meteorological offices have been overlooked. Therefore, when evaluating and determining the most effective parameter in dew point temperature prediction, it is crucial to use a method that can deal with inaccuracy and the possibility of errors in data and provide a satisfactory output with the lowest possible error. In this regard, Grey theory, which can use consecutive formulation, standardize the data, and take the environmental uncertainty for each parameter into account (through fuzzy calculation), can be employed to prioritize the parameters and determine the most important and effective ones. This paper seeks to use the Grey theory to cover the mentioned gap in the research on this subject and provide a decision making framework for ranking and determination of the most important parameters in dew point temperature prediction. The rest of paper is structured as follows: In section 2, Geographic characteristics are presented. The methodology is brought in Section 3. In Section 4, analysis of the research is done. Finally, conclusion remarks are drawn in Section 5. II. Area of study Iran as a developing country is located in Western Asia with a population of about 81 million people and an area of approximately 1,648,195 km 2 [13-15]. The country has different types of climates including mild and wet, Mediterranean, cold, cold semi-desert, dry and hot dry [16,17], which Fars province lays in areas with semi-arid, cold and Mediterranean climates [18]. Abadehis capital of Abadeh County in Fars Provinceof Iran that is located in 31°09′39"N52°39′02"E[18]. As of 2014, the population was estimated to be 94042 [19]. It features a continental semi-arid climate with extreme heat and dryness over summer, and cold (extreme at times) and wet winter, with huge variations between daytime and nighttime throughout the year [20]. Figure 1 illustrates map of Iran including location of Abadeh. Figure 1.Map of Iran including Abadeh area. III. Methodology In this research, a new methodology by using Grey Theory for selection of parameters to predict dew point temperature is proposed. In this section, the description of the proposed theorem is presented. Grey theory was introduced in 1982 by Deng. This theory is a modern mathematical theory developed from the concept of Grey set [21]. Its general procedure includes obtaining a set of initial data and repetitively processing them to weaken the effect of random information on these data with the intention of establishing a robust order (rule) in the data arrangement. The theory consists of five main parts: grey prediction, grey relational analysis, grey decision, grey programming and grey control [22, 23]. Based on the degree of https://en.wikipedia.org/wiki/Fars_Province https://en.wikipedia.org/wiki/Iran https://tools.wmflabs.org/geohack/geohack.php?pagename=Abadeh_County¶ms=31_15_N_52_30_E_type:city_region:IR IJECA-ISSN: 2543-3717. June 2019 information, the grey theory calls the system “white” when its information is fully known, calls it “black” when its information is fully unknown, and calls it “Grey” when it information is partly known.This fuzzy approach is a very effective method for solving problems involved with uncertainty, discrete data and incomplete information [24]. Other reasons behind the widespread use of this method include its simple calculati small size of its required sample, lack of need for distribution of the initial sample, its ability to deal with discrete data, and the fact that qualitative output obtained from Grey relational grade cannot be seen in the results of quantitative analyses [25]. Application of Grey theory is classified into 5 categories of evaluation, modeling, predicting, decision-making and control [ In an empirical formula provided by the researchers, dew point is a function of temperature and relative humidity [27]: )1121.0()9.0112( 125.0  TRHTTd Researchers have introduced the following eight parameters for predicting the dew point temperature and evaluating the factors affecting the prediction of dew point: global solar radiation on a horizontal surface (H), water vapor pressure (VP), atmospheric pressure (P), sunshine duration (n), minimum air temperature (T maximum air temperature (Tmax), average air temperature (Tavg), relative humidity (RH) [12, 28 these eight parameters are used to achieve the research objective. Grey set is defined as a set of uncertain data that can be described by a series of Grey numbers, Grey equations, Grey matrices, etc. the Grey set G is defined by reference set X with )(XG and G lower bounds of the membership function [ Where  , and when these two variables are equal Grey set turn into a fuzzy set. This means that Grey set is a fuzzy set that is flexible in dealing with different problems. Figure 2 shows a representation of Grey system procedures. ]1,0[:)( ]1,0[:)(   xX xX G G   Mojtaba Qolipour et al information, the grey theory calls the system “white” when its information is fully known, calls it “black” when its information is fully unknown, and calls it it information is partly known.This fuzzy approach is a very effective method for solving problems involved with uncertainty, discrete data and incomplete ]. Other reasons behind the widespread use of this method include its simple calculations, the small size of its required sample, lack of need for distribution of the initial sample, its ability to deal with discrete data, and the fact that qualitative output obtained from Grey relational grade cannot be seen in the results ]. Application of Grey theory is classified into 5 categories of evaluation, modeling, making and control [26]. In an empirical formula provided by the researchers, dew point is a function of temperature and relative (1) Researchers have introduced the following eight parameters for predicting the dew point temperature and evaluating the factors affecting the prediction of dew horizontal surface (H), water vapor pressure (VP), atmospheric pressure (P), sunshine duration (n), minimum air temperature (Tmin), ), average air temperature 28]. In this study, parameters are used to achieve the research Grey set is defined as a set of uncertain data that can be described by a series of Grey numbers, Grey equations, Grey matrices, etc. the Grey set G is defined )(X G as upper and lower bounds of the membership function [29]: (2) )()( XX GG   , and when these two variables are equal Grey set turns into a fuzzy set. This means that Grey set is a fuzzy set that is flexible in dealing with different problems. Figure 2 shows a representation of Grey system procedures. Figure 2.Grey system [ Assuming each fuzzy number as the fuzzy relationships between two Grey numbers ],[ 111 G and [2  G follows [31]: ],[ 212121   GG ,[ 12121   GG )],,,max( ,,[min(* 21212121 212121   GG ] 1 , 1 [*],[ 22 1121   GG ],[],[*. 11111  kkkGk  Length of grey number ],[ GG GGGL  )( In a Grey set, an option can be chosen over another only when we can establish that those two options are actually comparable and one can be considered more preferable. In deterministic sets, this fact can be established one through simple judgment, but in fuzzy sets this can be done via multi objective decision making techniques [ In the following, the instruments of Grey evaluation of dew point temperature parameters are described. 5 evaluated dimensions are represented by  521 ,...,, DDDDi  . The scale of Attribute ratings (shown in table 1) are used to assess the qualitative judgments of respondents for each criterion and, the scale of Attribute weights (shown in Table 2) a evaluate the weight of parameters [ amounts of table 1, we first normalized Abadeh's meteorology data (including the parameters) using following formula and then multiply them in 10 to changed Abadeh's meteorology data in G have: minmax min 0 xx xx x i      xxi Page 4 Grey system [30] Assuming each fuzzy number as ],[  G , the fuzzy relationships between two Grey numbers ], 22  can be defined as (3) ]2 (4) ),, 2121  (5) (6) ] (7) ] is defined as: (8) In a Grey set, an option can be chosen over another only when we can establish that those two options are actually comparable and one can be considered more preferable. In deterministic sets, this fact can be established one through simple judgment, but in fuzzy sets this can be done via multi-parameters or multi- objective decision making techniques [32]. In the following, the instruments of Grey evaluation of dew point temperature parameters are described. Here, 5 evaluated dimensions are represented by . The scale of Attribute ratings (shown in table 1) are used to assess the qualitative judgments of respondents for each criterion and, the scale of Attribute weights (shown in Table 2) are used to evaluate the weight of parameters [31]. For calculate the amounts of table 1, we first normalized Abadeh's meteorology data (including the parameters) using following formula and then multiply them in 10 to changed Abadeh's meteorology data in Grey data. So we 100 x (9) Mojtaba Qolipour et al IJECA-ISSN: 2543-3717. June 2019 Page 5 Table 1 illustrates the scale of attribute ratings, and Table 2 shows the scale of attribute weights. Table 1. The scale of Attribute ratings Scale Attribute ratings ( r ) Very poor(VP) [0, 1] Poor(P) [1, 4] Moderate(M) [4, 6] Good(H) [6, 9] Very good (VG) [9, 10] Table 2. The scale of Attribute weights Scale Attribute weights ( w ) Very Low(VL) [0.0, 0.1] Low(L) [0.1, 0.4] Medium(M) [0.4, 0.6] High(H) [0.6, 0.9] Very high(VH) [0.9, 1.0] Equation 10 is used to assess the opinions of received data [29]: ]...[ 1 ],[ 21 t ijijij t ij t ij t ij GGG t GGG  (10) In the above equation, t ijG isvalue for the t-th data for i-thobservation with respect to j-th parameter, and can be represented by the Grey number ],[ tij t ij t ij GGG  . The value of j-th parameter can be calculated via equation 11 [33]: 3,2,1; 7 7 1   j G G t tj j (11) After calculating the weights of parameters, options need to be evaluated. This evaluation is conducted via equation 12. The obtained Grey values t ijG is then used to form Grey decision matrix D:              mnm n GG GG D    1 111 (12) The evaluation numbers are all positive, so the decision matrix can be normalized. Equation 13 shown below is used to obtain normalized matrix *D [31]. ),( maxmax * j ij j ij ij GG G   (14) s.t }1;{maxmax miG ijj   (13)              ** ** * 1 111 mnm n GG GG D    Multiplying the normalized decision matrix by the obtained weight matrix gives the normalized weighted decision matrix: ijijij WGN  *               mnm n n NN NN D    1 111 (15) Hypothetical and important parameter of dew point temperature in case study, which is denoted by maxB , is calculated via the following equation [32]: ]}max,[max],...,max,{[max },...,,{ 111111 maxmax 2 max 1 max inmiinmiimiimi nGGGB    (16) The concept of preference degree is used to compare two Grey numbers with each other. Preference degree of Grey number 1G with respect to Grey number 2G is calculated via equation 17 [34]: )()( ),0max(),0max( )( 121 2121 21 GGGG GGGG GGp     (17)And when 1G equals 2G , equation 18 holds true: )( 21 GGp   = 5.0)( 12  GGp   )( 21 GGp   + 1)( 12  GGp  (18) This means that sum of the preference degrees of two Grey numbers always equal with 1. The Grey possibility degree between each decision parameter (dew point temperature) and the important parameter is then calculated via equation 19. Parameters with a lower Grey possibility are more suitable [29]:   n j jiji GNP n BBp 1 maxmax }{ 1 )( (19) Next section will describe the research findings. IV. Case study and data collection First, the data concerning the eight study parameter was collected from the meteorology office of Abadeh. This data pertained to a 10-year period from 2005 to 2015. Figure 3 shows the average monthly minimum temperature, maximum temperature and mean temperature of the study area during the period of 2005 to 2015. According to Figure 3, during this period, the maximum temperature of Abadeh has been 35°C s and its minimum temperature has been -5°C. International Journal of Energetica (IJECA) https://www.ijeca.info ISSN: 2543-3717 Volume 4. Issue 1. 2019 Page 01-10 IJECA-ISSN: 2543-3717. June 2019 Page 6 Figure 3.Monthly average of Tmax, Tmin and Tavg for case study over a period of 10 years from 2005to2015 Figure 4 illustrates the graph of average monthly precipitation and monthly mean temperature of Abadeh during the period of 2005 to 2015. According to Figure 4, during this period, the minimum precipitation has been zero and maximum precipitation has been 16 mm. Figure 4. Monthly average of Temperature and Precipitation for case study over a period of 10 years from 2005to2015 Figure 5 displays the graph of minimum humidity, maximum humidity, and monthly mean humidity of Abade during the period of 2005 to 2015 (the va Figure 5. Monthly average of RHmax, RHmin and RHavg for case study over a period of 10 years from 2005to2015 -5 0 5 10 15 20 25 30 35 40 T o C T Max T Avg. T Min 0 10 20 30 40 50 60 0 5 10 15 20 25 30 P re c ip it a ti o n m m T o C Temprature Percipitation 0 10 20 30 40 50 60 70 80 90 H u m id it y % RH min RH max RH avg. IJECA-ISSN: 2543-3717. June 2019 Figure 6 presents the graph of monthly mean hours of sunshine and monthly means water vapor in study area (in millimeter). According to Figure 6, in the period between 2005 and 2015, the maximum monthly mean water vapor has been 390 mm in July, and the minimum Figure 6. Monthly average of n and V V. Analysis Equations 10 and 11 were used to calculate the attribute weight of each parameter for the given data. Table 3 illustrates attribute weights of dew point's parameters as follows: Table 3. Attribute weights of dew point's parameters Parameters Attribute weights H [0.00,0.08] n [0.08,0.14] Vp [0.14,0.18] P [0.18,0.24] Tmin [0.24,0.30] Tmax [0.30,0.32] Tavg [0.32,0.36] RH [0.36,0.40] Attribute weights of each parameter were evaluated via equation 12 to assess the priority and selection dew point temperature's parameters. Grey decision table of the obtained data is illustrated in table 4: Table 4. Grey decision table D1 D2 D3 D4 H [0.5, 1.5] [2.4, 5.0] [3.4, 6.6] [7.5, 8.5] N [0.4, 2.5] [2.3, 4.0] [4.0, 7.8] [3.0, 6.2] Vp [0.5, 1.9] [2.5, 3.6] [4.2, 5.7] [5.5, 8.0] 0 100 200 300 400 500 W a te r v a p o r m m , H o u rs Mojtaba Qolipour et al re 6 presents the graph of monthly mean hours of sunshine and monthly means water vapor in study area (in millimeter). According to Figure 6, in the period between 2005 and 2015, the maximum monthly mean water vapor has been 390 mm in July, and the minimum monthly mean water vapor has been 5 mm in January and February. Meanwhile, the maximum monthly mean hours of sunshine has been 340 hours in June, and the minimum monthly mean hours of sunshine has been 200 hours in December Monthly average of n and Vp for case study over a period of 10 years from 2005to2015 Equations 10 and 11 were used to calculate the attribute weight of each parameter for the given data. Table 3 illustrates attribute weights of dew point's Attribute weights of dew point's parameters ights Attribute weights of each parameter were evaluated via equation 12 to assess the priority and selection dew point temperature's parameters. Grey decision table of the Grey decision table 4 D5 [7.5, 8.5] [6.5, 9.0] [3.0, 6.2] [7.5, 9.8] [5.5, 8.0] [8.5, 9.5] P [0.8, 2.7] [2.0, 3.2] [3.3, 5.1] Tmin [0.1, 3.5] [0.9, 2.7] [3.0, 4.4] Tmax [0.6, 2.3] [1.8, 2.5] [3.7, 6.3] Tavg [0.2, 2.8] [1.5, 4.4] [2.9, 6.1] Rh [0.3, 3.1] [0.6, 2.3] [3.5, 5.6] Using equation 13, the matrix of Table 4 is normalized which is shown in Table 5: Table 5. Normalized Grey decision table D1 D2 D3 H [0.2238, 0.8339] [0.1386, 0.7694] [0.3501, 0.4879] n [0.2588, 0.7513] [0.1493, 0.3499] [0.1966, 0.4090] Vp [0.2551, 0.8622] [0.0349, 0.2577] [0.2511, 0.4172] P [0.3188, 0.5060] [0.1407, 0.7254] [0.3160, 0.6715] Tmin [0.3077, 0.6991] [0.0631, 0.2543] [0.2075, 0.4733] Tmax [0.4336, 0.8909] [0.2050, 0.7147] [0.3517, 0.7172] Tavg [0.3426, 0.5993] [0.2435, 0.8143] [0.6302, 0.8308] Rh [0.1093, 0.6591] [0.1241, 0.9293] [0.4889, 0.5853] Finally, the normalized decision weighted matrix of Table 6 was obtained by multiplying the normalized matrix by the weight matrix. Vp n Page 7 monthly mean water vapor has been 5 mm in January and February. Meanwhile, the maximum monthly mean hours of sunshine has been 340 hours in June, and the minimum monthly mean hours of sunshine has been 200 over a period of 10 years from 2005to2015 [3.3, 5.1] [5.9, 7.2] [7.7, 9.0] [3.0, 4.4] [5.3, 8.2] [7.0, 9.4] [3.7, 6.3] [6.2, 7.4] [8.5, 10.0] [2.9, 6.1] [5.4, 8.0] [8.2, 9.9] [3.5, 5.6] [4.5, 7.0] [7.3, 9.7] Using equation 13, the matrix of Table 4 is normalized Normalized Grey decision table D4 D5 [0.3501, 0.4879] [0.0347, 0.5497] [0.0540, 0.8095] [0.1966, 0.4090] [0.3034, 0.9172] [0.5308, 0.9443] [0.2511, 0.4172] [0.2858, 0.7269] [0.4384, 0.7792] [0.3160, 0.6715] [0.2934. 0.7528] [0.3256, 0.9340] [0.2075, 0.4733] [0.2174, 0.6832] [0.1299, 0.7549] [0.3517, 0.7172] [0.3804, 0.8884] [0.3705, 0.5688] [0.6302, 0.8308] [0.1471, 0.5678] [0.4694, 0.7115] [0.4889, 0.5853] [0.0689, 0.7590] [0.0119, 0.8637] Finally, the normalized decision weighted matrix of 6 was obtained by multiplying the normalized Mojtaba Qolipour et al IJECA-ISSN: 2543-3717. June 2019 Page 8 Table 6. Standard Grey weighted decision matrix D1 D2 D3 D4 D5 H [0.0000, 0.1649] [0.0532, 0.1558] [0.1209, 0.1749] [0.1256, 0.2106] [0.1063, 0.2328] n [0.0192, 0.1277] [0.1381, 0.1812] [0.1470, 0.2173] [0.1818, 0.2802] [0.2012, 0.3509] Vp [0.0301, 0.1326] [0.0154, 0.1961] [0.1963, 0.2687] [0.1811, 0.2638] [0.1111, 0.2085] P [0.0135, 0.1742] [0.0872, 0.1427] [0.1844, 0.2414] [0.1455, 0.2294] [0.1259, 0.2502] Tmin [0.0112, 0.1655] [0.0430, 0.1491] [0.1890, 0.2868] [0.1316, 0.2707] [0.1948, 0.2759] Tmax [0.0632, 0.1943] [0.1067, 0.1535] [0.2835, 0.3685] [0.1693, 0.2620] [0.2305, 0.3176] Tavg [0.0630, 0.1133] [0.1690, 0.2614] [0.2998, 0.3001] [0.2350, 0.2893] [0.1832, 0.3234] Rh [0.0524, 0.1022] [0.0046, 0.1477] [0.2599, 0.3314] [0.2468, 0.3797] [0.1707, 0.3722] Equation 16 was then used to calculate the value of important parameter: maxB = ([0.00, 0.1934], [0.0046, 0.2614], [0.1209, 0.3685], [0.1256, 0.3797], [0.1111, 0.3722]) Equation 19 was then used to compare the values of weighted decision matrix with the ideal values. Table 7 shows the Grey possibility degrees of parameters of dew point temperature in Abadeh with respect to important parameters. Table 7. Grey possibility degree Possibility Degree P{H ≤ Bmax} 0.695 P{Vp ≤ Bmax} 0.712 P{P ≤ Bmax} 0.608 P{n ≤ Bmax} 0.697 P{Tmin ≤ B max} 0.635 P{Tmax ≤ B max} 0.622 P{Tavg ≤ B max} 0.551 P{Rh ≤ B max} 0.534 Options with lower Grey possibility degrees with respect to ideal option are more suitable [27]. According to the results of Table 7, among the assessed parameter's Grey possible, parameters including: RH, Tavg, P, Tmax, Tmin, H, n and Vp These parameters were determined –in that order- as the most effective parameters in dew point temperature prediction. RH and Tavg were determined as the most important parameters in dew point temperature prediction, and P was obtained as the third most important factor in this process. Given that at a fixed P, formula (1) holds true, these findings seem reasonable and acceptable. Therefore, the proposed method can properly evaluate the importance of parameters in dew point temperature prediction and prioritization of the parameters in the order of their effective and general importance. The findings of this study were found to be consistent comparing with previous studies. Options with lower Grey possibility degrees with respect to ideal option are more suitable [27]. According to the results of Table 7, among the assessed parameter's Grey possible, parameters including: RH, Tavg, P, Tmax, Tmin, H, n and Vp are important and effective respectively. Then, in spite of RH and Tavg, it received that P (air pressure) has noticeable effect on dew point temperature. To sum up, the Grey approach would be efficient for precise predictions of daily dew point temperature and also higher accuracy. Also, the findings are compatible to pervious research. VI. Conclusions The aim of this paper was to use Grey theory to evaluate and determine the most important parameters in prediction of dew point temperature for the city of Abadeh located in southern Iran. The used method is able to properly assess the problem and its dimensions in presence of any number of data, and has been used extensively by many researchers for different applications. In this study, 8 parameters affecting the dew point temperature including Vp, n, H, P, Tmin, Tmax, Tavg and RH were obtained from literature and then the data pertaining to these parameters was gathered from meteorological organization of the study area. Then, the proposed method was used to devise the table of parameter weights, Grey decision table, normalized Grey decision table, standardized weighted decision matrix and the grey possibility degree of each parameter. The results showed that RH, Tavg, P, Tmax, Tmin, H, n and Vp with the Grey possibility degrees of, respectively, 0.534, 0.551, 0.608, 0.622, 0.635, 0.695, 0.697 and 0.712, were the most important and effective parameters in predicting the dew point temperature. The results obtained through the use of proposed method were also found to be consistent with the results of previous works, except that parameter P was found to be the third most important parameter, after RH and Tavg, in the prediction of dew point temperature. References [1] Y.C.Wang, G.H. Tang. Prediction of sulfuric acid dew point temperature on heat transfer fin surface. Applied Thermal Engineering. Vol. 98, 2016, pp 492–501. Mojtaba Qolipour et al IJECA-ISSN: 2543-3717. June 2019 Page 9 [2] A. Baghban, M. Bahadori, J. Rozyn, M. Lee, A. Abbas, A. Bahadori, A. Rahimali. Estimation of air dew point temperature using computational intelligence schemes. Applied Thermal Engineering. Vol 93, 2016, pp 1043– 1052. [3] T. Zhang, Y. Jiang. Development of temperature and humidity independent control (THIC) air-conditioning systems in China - a review. Renewable and Sustainable Energy Reviews. Vol. 29, 2014, pp 793–803. [4] D.B. Jani, P.K. Manish Mishra. Solid desiccant air conditioning – A state of the art review. Renewable and Sustainable Energy Reviews. Vol 60, 2016, pp 1451–1469. [5] C. Forman, I. Kolawole, R. Muritala, P. Bernd Meyer. Estimating the global waste heat potential. Renewable and Sustainable Energy Reviews. Vol 57, 2016, pp 1568–1579. [6] D. Pandelidis, S. Anisimov. Numerical analysis of the heat and mass transfer processes in selected M-Cycle heat exchangers for the dew point evaporative cooling. Energy Conversion and Management. Vol 90, 2015, pp 62–83. [7] Y. Meng, X. Wen. Characteristics of dew events in an arid artificial oasis cropland and a sub-humid cropland in China. Journal of Arid Land. Vol 8, No. 3, 2016, pp 399-408. [8] Y. Zhuang, W. Zhao. The ecological role of dew in assisting seed germination of the annual desert plant species in a desert environment, northwestern China. Journal of Arid Land. Vol. 8, No. 2, 2016, pp 264-271. [9] X. Wang, Z. Gao, Y. Wang, Z. Wang, S. Jin. Dew measurement and estimation of rain-fed jujube (Zizyphus jujube Mill) in a semi-arid loess hilly region of China. Journal of Arid Land. Vol. 9, No. 4, 2017, pp 547-557. [10] Y. Pan, X. Wang. Effects of shrub species and microhabitats on dew formation in a revegetation-stabilized desert ecosystem in Shapotou, northern China. Journal of Arid Land. Vol. 6, No. 4, 2014, pp 389-399. [11] M. Amirmojahedi , K. Mohammadi, S. Shamshirband, A. Seyed Danesh, A. Mostafaeipour, A. Kamsin. A hybrid computational intelligence method for predicting dew point temperature. Environmental Earth Sciences. Vol. 75, 2016. 415. DOI 10.1007/s12665-015-5135-7. [12] M. Sarkar. A new theoretical formulation of dew point temperatures applicable for comfort air-cooling systems. Energy and Buildings.Vol. 86, 2015, pp 243–256. [13] S. Ebrahimi, M. Jahangiri, H.A. Raiesi, A. Rahimi Ariae. Optimal Planning of On-Grid Hybrid Microgrid for Remote Island Using HOMER Software, Kish in Iran. International Journal of Energetica. Vol. 3, No. 2, 2018, pp 13-21. [14] M. Rezaei, M. Salimi, M. Momeni, A. Mostafaeipour. Investigation of the socio- economic feasibility of installing wind turbines to produce hydrogen: Case Study. International Journal of Hydrogen Energy. Vol. 43, No. 52, 2018, pp 23135-23147. [15] A. Mostafaeipour, A. Sedaghat, M. Qolipour, M. Rezaei, H.R. Arabnia, M. Saidi- Mehrabad, S. Shamshirband, O. Alavi. Localization of solar-hydrogen power plants in the province of Kerman, Iran. Advances in Energy Research. Vol. 5, No. 2, 2017, pp 179- 205. [16] M. Rezaei, A. Mostafaeipour, M. Qolipour, H.R. Arabnia. Hydrogen production using wind energy from sea water: A case study on Southern and Northern coasts of Iran. Energy & Environment. Vol. 29, No. 3, 2018, pp 333- 357. [17] A. Mostafaeipour, M. Zarezade, H. Goudarzi, M. Rezaei-Shouroki, M. Qolipour. Investigating the factors on using the solar water heaters for dry arid regions: A case study. Renewable and Sustainable Energy Reviews. Vol. 78, 2017, pp 157-166. [18] M. Rezaei, A. Mostafaeipour, M. Qolipour, R. Tavakkoli-Moghaddam. Investigation of the optimal location design of a hybrid wind-solar plant: A case study. International Journal of Hydrogen Energy. Vol. 43, No. 1, 2018, pp 100-114. [19] Census of the Islamic Republic of Iran, 2006. Mojtaba Qolipour et al IJECA-ISSN: 2543-3717. June 2019 Page 10 [20] M. Rezaei-Shouroki, A. Mostafaeipour, M. Qolipour. Prioritizing of wind farm locations for hydrogen production: A case study. International Journal of Hydrogen Energy. Vol. 42, No. 15, 2017, pp 9500-9510. [21] W. Ming-dong, K. Bin, L. Xiang-yang, W. Xue-ke. Grey prediction theory and extension strategy-based excitation control for generator. Electrical Power and Energy Systems. Vol. 79, 2016, pp 188-195. [22] S.F. Liu, Y. Yang, Y. Cao, N.M. Xie. A summary on the research of GRA models. Grey Systems: Theory and Application.Vol. 3, No. 1, 2013, pp 7–15. [23] Q. Zhou, V.V. Thai. Fuzzy and grey theories in failure mode and effect analysis for tanker equipment failure prediction. Safety Science.Vol. 83, 2016, pp 74-79. [24] X. Li, W. Hipel, Y. Dang. An improved grey relational analysis approach for panel data clustering. Expert Systems with Applications. http://dx.doi.org/10.1016/ j.eswa.2015.07.066. [25] G. Li, D. Yamaguchi, M. Nagai. A grey based decision making approach to the supplier selection problem. Mathematical and Computer Modeling. Vol 36, 2007, pp 573- 581. [26] O. Gassmann, M. Keupp. The competitive advantage of early and rapidly internationalizing SMEs in the biotechnology industry: A knowledge-based view. Journal of World Business. Vol. 42, 2007, pp 350–366. [27] Meteorological organization of Fars province. [28] J. Lin, K. Thu, T.D. Bui, R.Z. Wang, K.C. Ng, K.J. Chua. Study on dew point evaporative cooling system with counter-flow configuration. Energy Conversion and Management. Vol. 109, 2016, pp 153–165. [29] M. Taqavifard, A. Mehdimalek. Using of Grey approach for priority key indexes to enhance strategic syllabuses effects. Quant. Of Industrial management researches. Vol. 9, No. 22, 2011, pp 135-165. [30] A. Dabaghi, A. Mahdimalek. A new method to assessment & priority organizations goals. Journal of Industry management.Vol. 2, No. 4, 2010, pp 57-74. [31] P. Bhattacharrya. A Grey theory based multiple attribute Approach for R&D project portfolio selection. Fuzzy information and engineering.Vol. 7, 2015, pp 211-225. [32] G. Dong, D. Yamaguchi, M. Nagai. A grey- based decision making approach to the supplier selection problem. Mathematical and Computer Modeling. Vol. 46, 2007, pp 573- 581. [33] W. Cheng pen, L. Fuguo, L. Hongya, Y. Zhanwei, C.H. Bo. Optimization of structural parameters for elliptical cross-section spiral equal-channel extrusion dies based on grey theory. Chinese Journal of Aeronautics. Vol. 26, No.1, 2007, pp 209–216. [34] J. Wei, L. Zhou, F. Wang, D. Wu. Work Safety Situation in Mainland China using Grey Theory. Applied Mathematical Modelling. Vol.39, No. 2, 2014, pp 924-933. http://dx.doi.org/10.1016/