基于扩展卡尔曼滤波神经网络算法的公路货运周转量预测

介绍了扩展卡尔曼滤波的原理,针对人工神经网络神经元之间权值的调整过程,建立了权值调整的状态空间模型,并采用扩展卡尔曼滤波对该模型的状态变量进行递推估计.文中仿真以全国历年公路货运周转量为例,分别采用BP算法和扩展卡尔曼滤波算法对神经网络进行训练,2种训练方法预测的结果对比表明扩展卡尔曼滤波训练算法具有更好的准确性和更高的运算效率.

Road Freight Turnover Forecasting Based on Extended Kalman Filter Neural Network Algorithm

The principle of the extended Kalman filter was introduced,and a state-space model of weights adjustment for the artificial neural network was established.The extended Kalman filter algorithm was used to estimate the state variable of the model and to train the artificial neural network.The simulation of the algorithm took the road freight turnover as an example,used BP algorithm and the extended Kalman filter algorithm respectively to train the artificial neural network and forecast the road freight turnov...