Apart from weather forecasting, wind speed prediction is required to estimate atmospheric effects on telecommunication systems such as satellites in launch, electromagnetic signal propagation in wireless channel, inland and water vehicle navigation, air traffic control, missile guidance etc. Wind is one of the clean, naturally available, intermittent energy source, which is being harnessed for electric power generation using windmills. Hence wind speed prediction models are a part of power demand forecasting and distribution models. This paper describes the application of neural network back propagation algorithm for the estimation of wind speed at a locality with latitude 12.92 degree N and Longitude 77.5 degree E. The weather data is collected using an Automatic Weather Station (AWS). Data is retrieved into a PC, preprocessed to remove redundancy and to find correlation among parameters. It is then input to the neuronal model for estimating wind speed. The parameters chosen as input are pressure, temperature, sunshine and humidity, as preprocessing has shown these parameters to be most correlated to wind speed.
Keywords: Wind speed estimation, Back propagation, Neural networks, AWS.
[...] A few of these are listed below: Accurate and timely short term forecasts of wind speed and direction are needed for Missile guidance, Air traffic control , navigation control on huge water bodies and in prediction of seasonal and periodic climatic forecasts. While the impact of a modern turbine might be slight, it is not completely eliminated. These solutions require the estimation of wind speeds at those places, typically on a daily and hourly basis i.e. in short Term. LITERATURE REVIEW Many available articles in literature show the usage of neural networks for wind speed prediction. [...]
[...] Apart from the above, literature also include Hybrid intelligent models incorporating many parameters to improve the accuracy of wind speed prediction. Some of them are - numerical weather prediction and mesoscale models generalized equivalent models[23], which includes techniques of ARMA[24], kalman filters[26], bilinear and smooth threshold autoregressive models,[27], Artificial intelligence techniques using multilayered perceptrons[28], radial basis functions[30], logic[31]and a combination of the above two[32]. recurrent neural network and fuzzy FEED FORWARD NEURAL NETWORK WITH BACKPROPOGATION[8] A neural network is a computational structure which resembles a biological neuron. [...]
[...] of test case 3 Name of test Error Tolerance Learning parameter Number of Epochs Number of layers Number of Neurons in each layer Windspeed Back propagation Test Expect ed Predict ed No of Input Patterns Fig Expected v/s predicted outputs for 5 layers Table 4 Test Specifications for 5 output layers and 100 epochs Sl No. of test case Name of test Error Tolerance Learning parameter Number of Epochs Number of layers Number of Neurons in each layer Wind speed 4 Backpropagation Test No of Input Patterns Expected Predicted Fig Expected v/s predicted outputs for 5 layers Table 5 Test Specifications for 5 output layers and 200 epochs Sl No. [...]
[...] Sfetsos and Siriopoulos have used Time series to forecast short term wind speed an d have arrived at a simple linear model for a electric power supply system. They claim that the traditional forecasting schemes are based on forecasting of averaged data based on the information drawn from the historical data of the same interval. In contrast their proposed scheme initially creates predictions based on data of the shorter intervals and then proceeds to average the values to obtain the final forecast. [...]
[...] Wind turbines produce electricity by using the natural power of the wind to drive a generator. For proper and efficient utilization of wind power, the prediction of wind speed is very important. It is needed for site selection, performance prediction, planning of windmills and the selection of an optimal size of the wind machine for a particular site, in an Electrical Power demand forecasting and generation plant. Since power forecasts are a function of wind speed, wind speed forecasts assume importance. [...]
APA Style reference
For your bibliographyOnline reading
with our online readerContent validated
by our reading committee
