This chapter examines the capability of Support Vector Machine (SVM), Relevance Vector Machine (RVM) and Genetic Programming (GP) for the optimal design of wind turbine. The excellent design has been influenced by various factors, such as profile of the blade, number of blades, power factor and tip speed ratio. The key to design a wind turbine is to Assessing the optimal tip speed ratio (TSR) is the key for designing the wind turbine. This chapter handles the Artificial Intelligence techniques in predicting the optimal TSR and the power factor based on the parameters engaged for NACA 4415 and LS-1 profile types with 3 and 4 blades. The organized machine learning framework is anticipated to be lucrative than the traditional way in foretelling the TSR and power factor. The machine learning models are then compared with the existing Neural Network model and the pros and cons of the various models are inferred from the results.
TopIntroduction
The drastic utilization of renewable energy sources due to booming level of pollution in all sources likewise air, water and soil and global iffy. The energy creators have emphasized on wind energy due to its pros likewise size of the machines, easy and manageable ability, endowment by the bureaucracy, tax benefits, etc. The omnipotence availability, spotless and the free of source make the best decision for encouraging wind power. This supremacy has succored the human race for centuries by mobilizing the ships and catalyzing wind turbines, pump water, etc. The popularly priced readiness and plentiful supply of petroleum, the high cost and uncertainty of wind placed it inconvenience to the economy. The recent hike in oil prices made the people to discern that world’s oil and gas accumulation would be cast off and that led to the development of other energy sources. In India, the wind power was modernized in early 1990’s and indispensably escalated in the last few years. According to the strategy of World Wind Energy Association, 2008, India spotted the position of fifth largest installed wind power capacity in the world. During the year 2009-2010 India competed other dominating countries and accomplished the highest growth rate. India produced 22,644.63 MW through the wind power by the year 2015 which was appreciable.
The range of wind velocity that is utilized by a specific wind turbine for the production of power is called productive wind speed. The power available from wind is proportional to cube of the wind's speed. This depicts that, when the velocity of wind abates, and then there would be a rapid fall in the generation of energy. If we flip that around, the more wind speed may overexert the turbine. Productive wind velocity may vary from 4 m/sec and 35 m/sec. The stipulated least possible speed for most advantageous performance of exorbitant scale wind farms is about 6 m/s. The wind speed makes the propeller like blades around a rotor, which was banded together with the main shaft. It is then whirls a generator to be the source of energy. It was advised to mount the wind turbines on a tower. This is important in order to apprehend the most energy. The wind turbines can do the trick of taking advantages of faster and less wind turbulent at an altitude of 100 feet (30 meters) or more above ground. Wind turbines can be employed for yielding power not only for a building or home but also to a vast dissemination by connecting with the electricity grid.
The outline of the wind turbine blade hinged on different factors, likewise the turbine profile, number of blades, power factor, and tip speed ratio. The optimal Tip Speed Ratio (TSR) is quintessential for devising the wind turbine. This will undeviatingly influence the power generated and, eventually it disturbs the investment made. It was advised to adopt TSR between the range 7 and 8 and during operation it was recommended to 7 for a 3-blade network connected wind turbine. However, the optimal TSR is counted on the profile type adopted and the blade number and could fall out of the boundaries suggested. Tip Speed Ratio can be augured by engaging the classical regression modeling techniques. This chapter employs the recent dominating soft computing techniques like of Support Vector Machine (SVM), Relevance Vector Machine (RVM) and Genetic Programming (GP) for foretelling the Tip Speed Ratio.
The Tip Speed Ratio (TSR) and Power factor (Cp) are the meat and potatoes for the composition of wind turbines. In our literature survey we found certain research in which TSR is calculated using artificial neural network (ANN). ANN, although widely popular has certain drawbacks such as:
- 1.
They are highly specific in nature. One program code is not suitable for multiple purposes.
- 2.
They are termed as “Black Box”, i.e. they are unpredictable and hard to train. There exists no knowledge of what is happening inside the code.
- 3.
They are not statistical in nature, thus reducing the overall reliability.
This chapter will be using Support Vector Machine (SVM), Relevance Vector Machine (RVM) and Genetic Programming (GP) which are prevalent machine learning methods. SVM, RVM and GP are found to be more accurate than ANN in real world research. It also compares the results and use optimum TSR and power factor to optimize the design process of a wind turbine.