Bird-Strike Prediction Model Application and Research Based on Neural Network
Shu Wen-di · Computer Technology and Development · 2012
Through comprehensive analysis to geographical environment,habitats,birds and their activities,bird informations to an airport,on the basis of collection,standardized,improved data set,it gived the processing method of bird information,proposed a bird-strike prediction model based on neural network,determined the input,output and other parameters,moreover using the BP neural network and RBF neural network methods to establish it.Using the Matlab simulated to the two prediction models,simulation results show that the BP neural network prediction is more accurate and effective than RBF neural network.Finally,through the observed data trained on the airport simulate to the bird beat model,examples verificated the bird-strike prediction model of BP neural network has good prediction results.