ESTIMATION OF WIND FARM EFFICIENCY BY ANFIS STRATEGY
Годишњак Педагошког факултета у Врању, књига VII, 2016.Dalibor PETKOVIĆ, Ph.DFaculty of Mechanical EngineeringUniversity of NišУДК 620.91- стручни рад -ESTIMATION OF WIND FARM EFFICIENCYBY ANFIS STRATEGYAbstract: A wind power plant which consists of a group of wind turbines ata specific location is also known as wind farm. The engineering planning of a windfarm generally includes critical decision-making, regarding the layout of the turbinesin the wind farm, the number of wind turbines to be installed and the types of windturbines to be installed. Two primary objectives of optimal wind farm planning areto minimize the cost of energy and to maximize the net energy production or to maximize wind farm efficiency. In the design process of a wind farm the aerodynamicinteractions between the single turbines have become a field of major interest. Theupwind turbines in a wind farm will affect the energy potential and inflow conditionsfor the downwind turbines. The flow field behind the first row turbines is characterizedby a significant deficit in wind velocity and increased levels of turbulence intensity.Consequently, the downstream turbines in a wind farm cannot extract as much powerfrom the wind as the first row turbines. Therefore modeling wind farm power production, cost, cost per power unit and efficiency is necessary to find optimal layout of theturbines in the wind farm. In this study, the adaptive neuro-fuzzy inference system(ANFIS) is designed and adapted to estimate wind farm efficiency according to turbines number in wind farm. The simulation results presented in this paper show theeffectiveness of the developed method.Key words: wind turbine, wind farm efficiency, power production, neurofuzzy, ANFIS.IntroductionWind energy is a promising renewable energy resource to help overcomeglobal warming and environmental pollution from the use of fossil fuel. Renewable energy sources are the greatest resource for this purpose. The world’s fastestgrowing renewable energy source is the wind energy [1]. Wind turbines are machines which convert the wind energy to the electricity [2]. Rapid advances in technical aspects and materials lead to an increase in size and output of the producedpower [3]. A problem is in wind turbine sizing and choosing the optimal configuration of the turbine’s parts [4]. Merely, by considering technical aspects, thebest turbine is the most efficient one, which has the highest coefficient of energyor capacity factor [5]. However, taking economic aspects into account can modifythe optimum size and design. Rotor radius, generator capacity and hub height are91
the most influential sizing parameters of the turbines. However, some limitationsare available for their relationship and ratios. It may vary from site to the site andwill be a function of the wind speed distribution at a given site.Wind farms composed of large capacity wind turbines are main electricalenergy sources. The modeling and simulation of the complete model of a windfarm with high number of wind turbines suppose the use of a high-order modeland a long computation time if all the wind turbines are modeled. In order toreduce the model order and computation time, equivalent models was developedto represent the collective response of the wind [6]. These models are based onaggregating wind turbines into an equivalent wind turbine. As wind turbinesbecome larger, wind farm layout design becomes more important. In [7] authorswas enabled simultaneously optimization of the placement and the selection ofturbines for commercial-scale wind farms that are subject to varying wind conditions. Among several wind farm layout design factors, wind turbine arrangementaccording to separation distance is one of the most critical factors for power output, annual energy production, availability, and life time of the wind turbine.Article [8] showed the effect of separation distance between two turbines and itwas found to be crucial for the conceptual design of a wind farm layout. Therefore, during wind farm layout design, separation distance and interaction betweenwind turbines must be carefully considered because they are directly connectedto the initial investment cost, payback period and economic efficiency. In article[9] the characteristics of turbine spacing for optimal wind farm efficiency wereinvestigated using combined numerical models. The results showed that the spacing between the first and the second turbines had the importance to the entirefarm’s efficiency. In [10] authors focused on the site specific optimization ofwind turbines by minimizing cost of electricity production. The study utilizes thecomplete and comprehensive capital cost model for wind turbines plus technicalaerodynamic model based on blade element momentum theory with twenty bladeelements. In article [11] was measured the productive efficiency of a group ofwind farms during the period 2001–2004 using the frontier methods Data Envelopment Analysis (DEA) and Stochastic Frontier Analysis (SFA).The objective of the present study is to analyze relationship between windfarm efficiency, power production, cost, cost per power unit and the number ofwind turbines in wind farm. Aiming at optimizing such systems to ensure optimalfunctioning of the wind farm, new techniques are used today such as the fuzzy logic(FL) [12], artificial neural network (ANN) [13] and neuro-fuzzy [14].Artificial neural networks are adaptable demonstrating apparatuses withproficiencies of taking in the numerical mapping between data and yield variablesof nonlinear frameworks. A standout amongst the most compelling sorts of neuralsystem framework is adaptive neuro-fuzzy inference system (ANFIS) [15].ANFIS indicates great taking in and expectation competencies, which makes it aproficient device to manage experienced vulnerabilities in any framework.ANFIS, as a hybrid intelligent system that enhances the ability to automaticallylearn and adapt, was used by researchers in various engineering systems.92
The key objective of this examination is to create an ANFIS for estimation of the wind farm efficiency. An endeavor is made to recover association between wind farm efficiency in regard to number of wind turbines. The trainingexperimental data will be extracted by an analytical, closed-form wake model[16] which quantifies the aerodynamic interaction between turbines.Wind Farm Efficiency ModelAnalytical wake model named as Jensen’s wake model [16] is chosen inthe study, since energy is considered as saved inside the wake by this model. Thewake grows straightly with downstream separation. In this manner, this model issuitable for the far wake locale. The wake has a sweep, at the turbine which isequivalent to the turbine range 𝑅𝑟 while, 𝑅1 is the radius of the wake in the model.𝑅1 is considered as radius of the downstream wake; the relationship between 𝑅1and 𝑋 is that downstream distance when the wake spreads downstream the radius 𝑅1 ; that increases linearly proportional, 𝑋. The wake extends directly withdownstream separation, as expressed in Jensen's model as demonstrated inFigure 1.Figure 1: Schematic of wake modelFollowing equation [38] was used to determine the wind speed after windturbine rotor as it shown in Figure 1:93
2𝑎𝑢 𝑢0 1 𝑋1 б(((1)21 𝑎)(𝑅𝑟 1 2𝑎))In the equation (1) we have: 𝑢0 is the mean wind rate or which might be clarified as the free streamwind speed and in this study was utilized 𝑢0 12𝑚/𝑠, axial induction factor is denoted by 𝑎 which can be calculated fromthe 𝐶𝑇 , thrust coefficient. This can be determined from the equation:𝐶𝑇 4𝑎(1 𝑎) 𝑋 is the distance downstream of the turbine, while 𝑅1 is related with𝑅𝑟 as represented using following equation:1 𝑎𝑅1 𝑅𝑟 1 2𝑎 б is the entertainment constant and by using the following equation:0.5б 𝑧ln 𝑧0In the above equation, 𝑧 is used to denote the hub height and roughnessof the surface is denoted by 𝑧0 . The value for surface roughness varies from fieldto field. For plain terrains the value for 𝑧0 0.3.The accessible wind force could be computed by utilizing the accompanying comparison:1𝑃𝑎 с𝐴𝑢3 (2)2If the power production from each wind turbine contains the efficiency зof wind turbine then the following equation for the energy or power generatedfrom a wind turbine can be used:1𝑃𝑝 з с𝐴𝑢3 (3)2If that the efficiency of wind turbines is equal to 40%, then the equation will be:40 1𝑃𝑝 с𝐴𝑢3 (4)100 2In the above equation 𝐴 represents cover surface of the turbine bladesduring rotation and it is 𝐴 р402 since the used rotor radius in the study is 𝑅𝑟 40 𝑚, and с 1.2. The following equation will be derived:𝑃𝑝 301𝑢3 𝑊 (5)For the calculation of power into kW we have the following equation:𝑃𝑝 0.3𝑢3 𝑘𝑊 (6)94
Effectiveness is a term which is utilized for the measure of vitality concentrated as a piece of the aggregate vitality accessible. The effectiveness mightbe ascertained from the accompanying comparison: 𝑁𝑡 0.3 𝑢𝑖3𝐸 𝑖 1(7)𝑁𝑡 (0.3 𝑢03 )𝑃𝑡𝐸 (8)𝑁𝑡 (0.3 𝑢03 )where 𝑁𝑡 represents the turbines number in wind farm, 𝑃𝑡 is total delivered power from all wind turbines in wind farm.The expense demonstrate that is utilized as a part of the present study isthe same as that was utilized within prior studies [40, 41]. This model encouragesa decrease of 1/3 in expense for every new wind turbine establishment. In thisway, it might be said that this expense model is a capacity of number of turbines.From the above view the aggregate expense could be explained utilizing theaccompanying comparison:2 12𝐶 𝑁𝑡 [ 𝑒 0.00174 𝑁𝑡 ] (9)3 3where 𝑁𝑡 represents the number of turbines in wind farm.It is useful to analyze also the cost per power unit (CPPU) if only onewind turbine is installed in wind farm and it can be defined as following:𝐶𝐶𝑃𝑃𝑈 (10)𝑃𝑡In this study the used values are as under:𝑋 200 𝑚𝑅𝑟 40 𝑚𝑢0 12 𝑚/𝑠𝑎 0.326795б 0.09437Adaptive neuro-fuzzy inference applicationAn ANFIS model was established in this study to estimate the efficiency,cost, cost per power unit and power production in wind farm in relation to numberof turbines in the wind farm. For the present study equations (1) to (10) were usedto extract training experimental data for the ANFIS network modeling. One 50%of the extracted data were utilized for preparing while the other half is utilized forchecking and acceptance of the model. With a legitimate preparing plan and fineseparated information sets, ANFIS can efficiency, cost, cost per power unit andpower production in wind farm quite accurately since it learns from training data.This measurement-free architecture also makes it immediately available for operation once they are trained.There were three membership functions on each input. In this study bellshaped membership functions were chosen with maximum equal to 1 and minimum equal to 0. Figure 2 shows an ANFIS model with three inputs.95
Figure 2: ANFIS structureIn this work, the first-order Sugeno model with two inputs and fuzzy IFTHEN runs of Takagi and Sugeno's sort was utilized:𝑖𝑓 𝑥 𝑖𝑠 𝐴 𝑎𝑛𝑑 𝑦 𝑖𝑠 𝐶 𝑎𝑛𝑑 𝑧 𝑖𝑠 𝐸 𝑡ℎ𝑒𝑛 𝑓1 𝑝1 𝑥 𝑞1 𝑦 𝑟1 𝑧 𝑡 (11)The first layer has input variables i.e. membership functions (MFs), input1, input 2 and input 3. This layer simply supplies the data qualities to the following layer. At the beginning there are three ANFIS networks which estimatespower production, cost and cost per power unit of wind farm in depend on numberof wind turbines in the wind farm. The outputs of the three ANFIS networksrepresent inputs to the new ANFIS network with three inputs (Figure 2). Theinput 1: x is power production, input 2: y is cost per power unit and input 3: z istotal cost of the wind farm. In the first layer every node is an adaptive node witha node function𝑂 м(𝑥, 𝑦, 𝑧),where м(𝑥, 𝑦, 𝑧)𝑖 are membership functions.Bell-shaped membership functions (3) with maximum 1 and minimum 0was chosen in this study:1𝑓(𝑥; 𝑎, 𝑏, 𝑐) 1 𝑥 𝑐 2𝑏( )𝑎(12)where the bell-shaped function depends on three parameters a, b and c. The parameter b is usually positive. The parameter c located the center of the curve as itis shown in Figure 3.96
Figure 3: Bell-shaped membership function (a 2, b 4, c 6)The second layer (participation layer) checks for the weights of everyMfs. It accepts the info values from the first layer and goes about as Mfs to speakto the fuzzy sets of the separate information variables. Each hub in the secondlayer is non-versatile and this layer reproduces the approaching signs and sendsthe item out like𝑤𝑖 м(𝑥)𝑖 м(𝑥)𝑖 1 (13)Every hub yield speaks to the terminating quality of a standard or weight.The third layer is known as the standard layer. Every hub (every neuron)in this layer performs the precondition matching of the fuzzy tenets, i.e. they register the initiation level of each one lead, the amount of layers being equivalent tothe amount of fuzzy guidelines. Every hub of these layers figures the weightswhich are standardized. The third layer is likewise non-versatile and each hubfigures the proportion of the guideline's terminating quality to the total of all standards' terminating qualities like𝑤𝑖𝑤𝑖 (14)𝑤1 𝑤2𝑖 1,2.The outputs of this layer are called normalized firing strenghts or normalized weights.The fourth layer is called the defuzzification layer and it provides theoutput values resulting from the inference of rules. Every node in the fourth layeris an adaptive node with node function𝑂𝑖4 𝑤𝑖 𝑥𝑓 𝑤𝑖 (𝑝𝑖 𝑥 𝑞𝑖 𝑦 𝑟𝑖 ) (15)where {𝑝𝑖 , 𝑞𝑖 , 𝑟} is the consequent parameters.The fifth layer is known as the yield layer which totals up all the inputshailing from the fourth layer and converts the fuzzy characterization results intoa fresh (paired). The single hub in the fifth layer is not versatile and this hubfigures the general yield as the summation of all approaching indicators97
𝑂𝑖5 𝑤𝑖 𝑥𝑓 𝑖 𝑖 𝑤𝑖 𝑓(16) 𝑖 𝑤𝑖The hybrid learning algorithms were used to identify the parameters inthe ANFIS architectures.ResultsThe proposed ANFIS struc
wind farms during the period 2001–2004 using the frontier methods Data Enve-lopment Analysis (DEA) and Stochastic Frontier Analysis (SFA). The objective of the present study is to analyze relationship between wind farm efficiency, power production, cost, cost per power unit and the number of wind turbines in wind farm.
red wind/red wind xlr h50 t-15m l 35 mm red wind/red wind xlr h80 t-16m l 65 mm red wind/red wind xlr h105 t-17m l 90 mm racing speed xlr h80 t-19m l 74 mm profile rim female valve adapter (option) red wind/red wind xlr h50 t-15f l 37 mm red wind/red wind xlr h80 t-16f l 67 mm red wind/red wind xlr h105 t-17f l 92 mm racing speed .
Wind energy is the kinetic energy of air in motion, also called wind. Total wind energy flowing through an imaginary surface with area A during the time t is: . The London Array Offshore Wind Farm, the largest offshore wind farm in the world with an installed capacity of 630MW, ranks as the world's sixth largest wind farm. .
the first cooperatively owned offshore wind farm became a reality. Today the 40 MW wind farm with twenty modern 2 MW wind turbines developed by the Middelgrunden Wind Turbine Cooperative and Copenhagen Energy Wind is producing electricity for more than 40,000 households in Copenhagen.
Wind farm control can be used to increase wind energy e ciency by maximizing power in wind farms that are already installed. In can also be used to mitigate structural loads to maximize the lifetime of the turbines and better integrate wind energy into the energy market. Currently, turbines in a wind farm are operated to maximize their own .
Jun 27, 2017 · Wind turbines often stand together in a windy area that has been through a robust development process in an interconnected group called a wind project or wind farm, which functions like a wind power plant. These turbines are connected so the electricity can travel from the wind farm to the power grid. Once wind energy is on the
Common concerns about wind power, June 2017 1 Contents Introduction page 2 1 Wind turbines and energy payback times page 5 2 Materials consumption and life cycle impacts of wind power page 11 3 Wind power costs and subsidies page 19 4 Efficiency and capacity factors of wind turbines page 27 5 Intermittency of wind turbines page 33 6 Offshore wind turbines page 41
Off-farm income is included in the discussion of farm income at the household level at the end of this report. 1 For information on state-level farm income, see "U.S. and State Farm Income and Wealth Statistics," available as part of the Farm Income and Wealth Statistics, Farm Income and Costs, Farm Economy Topics, Economic Research
7 COMMERCE (Degree in Finance / Commerce) 10-11 8 ARTS/HUMANITIES (Degrees in Humanities and Social Sciences) 12-13 PROFESIONAL CAREERS AFTER 12th (PCAT) General Degree Courses available after any stream in 12th 9 LAW 14-15 10 BUSINESS MANAGEMENT 16-17 11 HOTEL MANAGEMENT 18-19 12 LIBERAL STUDIES 20 13 MASS COMMUNICATION 21 14 ECONOMICS 22
