26-07-2014, 04:44 PM
A Neural Approach for Optimum Matching of Wind Turbines to the Potential Wind Site to Obtain Higher Plant Factor
A Neural Approach.docx (Size: 584.53 KB / Downloads: 11)
Abstract-
A Development of techniques for accurately assessing the wind power potential of site is gaining increased importance. This is because of the fact that the planning and establishment of a wind energy system depends upon various factors. Once the details of wind resource is known for potential wind power site, Efficient design of a wind energy system demands optimum matching of wind turbines to the potential wind site to obtain higher plant factor.The model which is designed is useful for planning of wind power stations as it can be applied for accurate assessment of wind power potential at a site. The Weibull model is the most used one and provides good results. However, the accurate determination of the wind speed distribution law constitutes a major problem. Multi Layer Perceptron type artificial neural networks (ANN) are used here for the approximation of the wind speed distribution law. In this paper, site energy characteristic is determined by means of the neural approach and compared with those obtained by the classical method. Distribution law achieved by the neural model provides assessments closer to the discrete distribution than the Weibullmodel[1].
I. INTRODUCTION
Energy is an important input in all sectors of any country’s economy. The standard of living of a given country can be directly related to per capita energy consumption. Energy crisis is due to the two reasons; firstly that population of the world has increased rapidly and secondly the standard of living of human beings has increased. If present trend continues, the conventional sources are depleting and maybe exhausted by the end of the century or beginning of the next century. Nuclear energy requires skilled technicians and poses the safety as regards to radioactive waste disposal. Solar energy and other non-conventional energy sources are the sources; those are to be utilized in future. Solar energy can be a major source of power. Its potential is 178 billion MW which is about 20,000 times world’s demand. But so far it could not be developed in large scale. Wind energy which is an indirect source of solar energy conversion can be utilized to run windmill, which in turn drives a generator to produce electricity. Wind can also be used to
provide mechanical power. Wind energy uses the high wind velocity available in certain parts. California in USA is generating 500MW and 900 wind turbines based on windmills. Wind energy used for pumping the water or power generation. About 0.7 million wind pumps are in operation in different countries. A minimum wind speed of 3m/s is needed. This is considered to have a high efficiency. Coastal, hilly and valley areas are suitable for this process. Potential in India is estimated between 20,000 and 25,000MW. Coastal areas of Gujarat, Maharashtra and Tamil Nadu are considered as favorable. In number of experimental stations has been setup.
V. CONCLUSION
Our aim is to match the different turbines availablecommercially to the particular site. In the present work Feed Forward Artificial neural network is adopted as a tool to match the site with turbine characteristics to 10 different turbines in corresponding to 4 inputs. Output shows that with a (neural network structure 4-17-10)17 hidden neurons its giving better output identification.We conclude that Feed Forward Artificial neural network is also used to match the site to the turbine in order to get the maximum plant factor.
VI. REFERENCES
[1] IEEE Xplore - Site matching of wind turbine generators: a case study Suresh .H.Jangamshetti, Elect. Engg. Dept. Indian Institute of technology kharagpur, India
[2] Wind power analysis and site matching of wind turbine generators In Kingdom of Bahrain: Article by Fawzi A.L. Jowder, Electrical and Electronics Engineering Department, College of Engineering, University of Bahrain,
[3] Simon haykin‘neural network ‘second edition.
[4]"NeuralNetworks:AComprehensiveFoundation",bySimonS.Haykin,(1999), PrenticeHall, Chapter1-15, page1-889.
[5] "ElementsofArtificialNeuralNetworks",by KishanMehrotra,ChilukuriK.Mohan andSanjayRanka,(1996),MITPress,Chapter1-7,page1-339.
[6] “non-conventional energy sources”, by G.D.RAI.(unit2) page 227-285
[7] Artificial neural network - Wikipedia, the free encyclopaedia