基于小波去噪的灰色GM（1，1）与RBFNN在变形预测中应用

摘要

Processing and prediction to two examples based on the combined model of wavelet RBF neural network and the combined model of wavelet gray metabolic GM(1,1). The results indicate that a combination is superior to a single model predicted results. The prediction accuracy of RBFNN significantly affected by noise,Metabolic GM(1,1)prediction model of short-term prediction is high accuracy,and the prediction ac-curacy is unaffected when the noise is small.%应用小波去噪的新陈代谢GM（1，1）模型和小波去噪的RBFNN组合模型对两个实例进行处理和预测，结果表明组合模型预测结果优于单一模型，RBFNN预测模型的预测精度受噪声影响明显，而新陈代谢GM（1，1）预测模型对于短期预测具有较高的精度，且预测精度在噪声较小的情况下不受噪声影响。