基于CEEMD-IBES-ELM的水轮机尾水管压力脉动预测OA

In order to effectively predict the pressure pulsation of the hydraulic turbine draft tube and take corresponding measures to reduce the pressure pulsation,a hybrid prediction model was proposed,which was based on complete ensemble empirical mode decomposition(CEEMD),the improvement of the bald eagle search algorithm(IBES),and the extreme learning machine(ELM)to predict the pressure pulsation signal of the draft tube.Firstly,the non-smooth hydraulic turbine draft tube pressure pulsation signal was split into several relatively stable submodal parts based on the CEEMD decomposi-tion.Secondly,the decomposed submodal data were input into the ELM model for in-depth training and prediction analysis,and the initial weights and thresholds were optimised by IBES.Finally,the prediction result outputs of each sub-modality were superimposed to obtain the final prediction results of the pressure pulsation signals of the hydraulic turbine draft tube.The simulation results show that the proposed CEEMD-IBES-ELM prediction method can reduce the complexity of the prediction process with relatively low reconstruction error.In addition,compared with other models,this model demon-strates significant superiority in terms of prediction accuracy and stability,and has good potential for application.