Search

Efficient extraction of renewable energy from wind depends on the reliable estimation of wind characteristics and optimization of wind farm installation and operation conditions. There exists uncertainty in the prediction of wind energy tapping potential based on the variability in wind behavior. Thus the estimation of wind power density based on empirical models demand subsequent data processing to ensure accuracy and reliability in energy computations. Present study analyses the reliability of the ANN-based machine learning approach in predicting wind power density for five stations (Chennai, Coimbatore, Madurai, Salem, and Tirunelveli) in the state of Tamil Nadu, India using five different non-linear models. The selected models such as Convolutional Neural Network (CNN), Dense Neural Network (DNN), Recurrent Neural Network (RNN), Bidirectional Long Short Term Memory (LSTM) Network, and linear regression are employed for comparing the data for a period from Jan 1980 to May 2018. Based on the results, it was found that the performance of (1->Conv1D|2->LSTM|1-dense) is better than the other models in estimating wind power density with minimum error values (based on mean absolute error and root mean squared error).

An accurate estimate of wind resource assessment is essential for the identification of potential site for wind farm development. The hourly average wind speed measured at 50 m above ground level over a period of 39 years (1980-2018) from 25 locations in Tamil Nadu, India have been used in this study. The annual and seasonal wind speed trends are analyzed using linear and Mann-Kendall statistical methods. The annual energy yield, and net capacity factor are obtained for the chosen wind turbine with 2 Mega Watt rated power. As per the linear trend analysis, Chennai and Kanchipuram possess a significantly decreasing trend, while Nagercoil, Thoothukudi, and Tirunelveli show an increasing trend. Mann-Kendall trend analysis shows that cities located in the southern peninsula and in the vicinity of the coastal regions have significant potential for wind energy development. Moreover, a majority of the cities show an increasing trend in the autumn season due to the influence of the retreating monsoons which is accompanied with heavy winds. The mean wind follows an oscillating pattern throughout the year at all the locations. Based on the net annual energy output, Nagercoil, Thoothukudi and Nagapattinam are found to be the most suitable locations for wind power deployment in Tamil Nadu, followed by Cuddalore, Kumbakonam, Thanjavur and Tirunelveli.

This paper presents a detailed comparison between viable adaptive intelligent torque control strategies of induction motor, emphasizing advantages and disadvantages. The scope of this paper is to choose an adaptive intelligent controller for induction motor drive proposed for high performance applications. Induction motors are characterized by complex, highly non-linear, time varying dynamics, inaccessibility of some states and output for measurements and hence can be considered as a challenging engineering problem. The advent of torque and flux control techniques have partially solved induction motor control problems, because they are sensitive to drive parameter variations and performance may deteriorate if conventional controllers are used. Intelligent controllers are considered as potential candidates for such an application. In this paper, the performance of the various sensor less intelligent Direct Torque Control (DTC) techniques of Induction motor such as neural network, fuzzy and genetic algorithm based torque controllers are evaluated. Adaptive intelligent techniques are applied to achieve high performance decoupled flux and torque control. This paper contributes: i) Development of Neural network algorithm for state selection in DTC; ii) Development of new algorithm for state selection using Genetic algorithm principle; and iii) Development of Fuzzy based DTC. Simulations have been performed using the trained state selector neural network instead of conventional DTC and Fuzzy controller instead of conventional DTC controller. The results show agreement with those of the conventional DTC.

Welded stainless steel components used in power plants and chemical industries are subjected to mechanical load cycles at elevated temperatures which result in early fatigue failures. The presence of weld makes the component to be liable to failure in view of residual stresses at the weld region or in the neighboring heat affected zone apart from weld defects. Austenitic stainless steels are often welded using Tungsten Inert Gas (TIG) process. In case of single pass welding, there is a reduced weld penetration which results in a low depth-to-width ratio of weld bead). If the number of passes is increased (Multi-Pass TIG welding), it results in weld distortion and subsequent residual stress generation. The activated flux TIG welding, a variant of TIG welding developed by E.O. Paton Institute, is found to reduce the limitation of conventional TIG welding, resulting in a higher depth of penetration using a single pass, reduced weld distortion and higher welding speeds. This paper presents the fatigue crack growth rate characteristics at 823 K temperature in type 316LN stainless steel plates joined by conventional multi-pass TIG (MP-TIG) and Activated TIG (A-TIG) welding process. Fatigue tests were conducted to characterize the crack growth rates of base metal, HAZ and Weld Metal for A-TIG and MP-TIG configurations. Micro structural evaluation of 316LN base metal suggests a primary austenite phase, whereas, A-TIG weld joints show an equiaxed grain distribution along the weld center and complete penetration during welding (Fig. 1). MP-TIG microstructure shows a highly inhomogeneous microstructure, with grain orientation changing along the interface of each pass. This results in tortuous crack growth in case of MP-TIG welded specimens. Scanning electron microscopy studies have helped to better understand the fatigue crack propagation modes during high temperature testing.