关键词: 粒子群算法, RBF神经网络, 铁矿石消费预测, 全局最优

Abstract: In order to improve the prediction accuracy of iron ore consumption，using a time series forecasting method based on intelligent calculation.First，the particle swarm algorithm was improved，and it was used to optimize the ability of global optimization of key parameters of RBF neural network；finally iron ore consumption prediction model was established.The results showed that: this method hada high prediction accuracy compare with other prediction methods，and providesda new way for iron ore consumption forecast.

Key words: Particle swarm optimization algorithm, RBF neural network, Iron ore consumption prediction, Global optimum