基于LM算法的模糊优选神经网络模型

fFuzzy Optimization Neural Network Model Based on Levenberg-Marquardt Algorithm

针对传统模糊优选神经网络模型训练速度慢、训练结果易陷入局部最小解的缺点,提出了基于LM算法的模糊优选神经网络模型且对传递函数进行了改进,并预测分析了黄河内蒙段的三湖河口站和巴彦高勒站冰情。实例结果表明,改进模型训练速度更快、训练预测结果更优。

Aiming at the disadvantages of slow convergence and easily trapping into local minimum of traditional fuzzy optimization neural network model,a new fuzzy optimization neural network model is proposed by using Levenberg-Marquardt algorithm.Moreover,the transfer function of the model is also improved.Meanwhile,taking the Sanhuhekou Station and Bayagaole Station in Inner Mongolia reach of the Yellow River for examples,the new model is applied to forecast the water regimen.The results show that the improved model has faster training speed and better forecasting capability.