A method of predicting demand for aircraft follow-up spare based on discrete particle swarm optimization algorithm and RBF neural network

The traditional method to predict the demand of aircraft follow-up spare has some problems including being short of adapting to the noise data. It leads to local optimum easily and low accuracy of prediction. So a method based on discrete particle swarm optimization and RBF neural network to predict demand for aircraft follow-up spare is put forward. Firstly, the data of the aircraft follow-up spare is reduced by discrete particle swarm optimization algorithm to get the key factors affecting the demand of spare. Then the RBF neural network is built on the key factors to predict the demand of spare. The experimental results show that this method can ensure the rationality of the input parameters and provide a new way of the neural network to predict the demand of the aircraft follow-up spare.