基于Kalman滤波的神经网络学习算法及其应用

Kalman Filter Learning Algorithm for Feedforward Neural Network and its Applicaion

针对传统神经网络学习算法速度慢、容易陷入局部最优解的缺点 ,将卡尔曼滤波应用于人工神经网络的训练算法中。同时 ,在卡尔曼滤波计算中 ,将奇异值分解应用于卡尔曼滤波的递推公式中 ,提高了协方差阵计算的数值稳定性。最后 ,本文通过将神经网络的卡尔曼滤波算法应用于电力系统短期负荷预测中 ,验证了该方法不仅具有理论意义 。

Conventional algorithms for feedforward neural network always suffer from slow convergent rate and local convergence.Aiming at these shortcomings of algorithms of neural network,a learning algorithm based on Kalman filter is proposed for training a neural network in this thesis.Furthermore,singular value decomposition(SVD)is applied in the calculation of the covariance matrix of the formulas of Kalman Filter.At last,this algorithm for neural network is used in short-term load forecast of power system.The results prove this algorithm.