Research on PDEs Solution Based on Adaptive Neural Networks
In response to the shortcomings of the current neural network-based PDEs solution method in terms of efficiency and accuracy,a PDEs solution model based on improved BP neural net-work(BPNN)is proposed.Firstly,refer to the adaptive grid method to improve the neural network structure,construct an adaptive neural network,and improve the output accuracy of the model.Second-ly,an Improved Seagull Optimization Algorithm(ISOA)incorporating Levy flight mechanism and Whale Optimization Algorithm(WOA)is proposed to optimize the BPNN,search for the optimal pa-rameters of the BPNN,and improve the performance of the model.Based on the above content,con-struct an intelligent solution model based on ISOA-BPNNPDEs.The results show that the F1 value of the model is 95.74％,with an accuracy of 97.96％and an output error of 0.021,which is superior to the two most advanced PDEs solving models currently available.The above content indicates that the ISOA-BPNNPDEs intelligent solution model constructed in the study can efficiently and accurately solve PDEs,providing a new path for PDEs solution research.
万方数据
维普
