基于数据修正的改进型灰色-RBF网络预测

Improved Gray-RBF Network Prediction and Based on Revised Data

针对既呈趋势性,又呈波动性的时间序列难以预测的问题,提出了基于失真数据的修正的改进型灰色-RBF网络预测模型及算法.即用改进型灰色模型提取趋势性因素,用神经网络处理波动性因素,另尝试性的提出还须排除异常干扰因素,即查找和修正序列的异常数据.并以南昌铁路车站旅客发送量预测为例,验证了算法的有效性,收到很好的预测效果.

Aiming at the problem that the tendency and fluctuated time-series are difficulte to forecast,this paper put foreward improved gray-RBF network model based on revised data distortion and its algorithms.That is,extracting the trend with the improved gray model,processing the fluctuated factor using radial basis function neural networks,in addition,proposed excluding the unusual interference factors,namely to find out and identify abnormal data of the sequence.and the forecasting of passenger transmission volume in Nanchang Railway Station verified the validity of the model and algorithm,received a good predictive effect.