小波分析与粒子群算法优化的灰色神经网络模型

Gray Neural Network Model Optimized by Wavelet Analysis and Particle Swarm Optimization

针对灰色模型在测量数据波动较大时预测精度下降的缺陷,本文提出了采用小波分析与粒子群算法优化后的BP神经网络对灰色模型进行优化,生成了基于小波分析和粒子群算法优化的灰色神经网络预测模型。并使用青岛市航运中心的沉降观测数据对优化后的灰色模型进行精度验证,证明优化后的模型的预测精度相比于传统的灰色模型与经过小波降噪后的灰色模型,预测精度有了进一步提高。

Aiming at the gray model,the prediction accuracy is degraded when the measured data fluctuates greatly. In this paper,we use the BP neural network optimized by particle swarm optimization and wavelet analysis to optimize the gray model,and generate a gray neural network prediction model. Then we use the settlement data of Qingdao shipping center to verify the precision of the gray neural network model,It can be concluded that the prediction accuracy of gray neural network model is higher compared with the traditional gray model and the gray model after removal of noise by wavelet analysis.