本文選題：漿體管道 + 臨界淤積流速��； 參考：《傳感器與微系統(tǒng)》2017年03期

【摘要】：針對(duì)漿體管道臨界淤積流速預(yù)測難度大、精度低等問題,提出了粒子群優(yōu)化—極限學(xué)習(xí)機(jī)(PSOELM)的臨界淤積流速預(yù)測模型。利用PSO算法對(duì)ELM模型參數(shù)輸入權(quán)值和隱元偏置進(jìn)行優(yōu)化,應(yīng)用優(yōu)化得到的ELM模型對(duì)預(yù)測集進(jìn)行預(yù)測。通過實(shí)驗(yàn)仿真得到預(yù)測結(jié)果的最大誤差為5.73%,預(yù)測效果優(yōu)于常規(guī)的ELM模型和反向傳播(BP)神經(jīng)網(wǎng)絡(luò)模型。
[Abstract]:Aiming at the problems of high difficulty and low precision in predicting critical deposition velocity of slurry pipeline, a particle swarm optimization (PSO) model for predicting critical deposition velocity is proposed. The PSO algorithm is used to optimize the input weights and implicit element bias of ELM model parameters, and the predicted set is predicted by the optimized ELM model. The maximum error of the predicted result is 5.73 through experimental simulation, which is superior to the conventional ELM model and the BP neural network model.
【作者單位】： 昆明理工大學(xué)信息工程與自動(dòng)化學(xué)院;云南省礦物管道輸送工程技術(shù)研究中心;
【基金】：國家自然科學(xué)基金資助項(xiàng)目(51169007) 云南省科技計(jì)劃項(xiàng)目(2013DH034) 云南省中青年學(xué)術(shù)和技術(shù)帶頭人后備人才培養(yǎng)計(jì)劃項(xiàng)目(2011CI017)
【分類號(hào)】：U171;TP18
，

本文編號(hào)：1874182