石油含水率測(cè)量動(dòng)態(tài)補(bǔ)償模型研究
Research on Measuring Moisture in Petroleum Based on Dynamic Compensation Metho

摘要:針對(duì)傳統(tǒng)微波透射法測(cè)量石油含水率存在的測(cè)量誤差大等問題，提出了一種基于神經(jīng)網(wǎng)絡(luò)的動(dòng)態(tài)補(bǔ)償方法，確定衰減和相移兩個(gè)參量作為動(dòng)態(tài)補(bǔ)償模型的輸人;針對(duì)傳統(tǒng)BP算法具有收斂速度慢、容易陷人局部極小值等缺點(diǎn)，采用微粒群訓(xùn)練算法對(duì)神經(jīng)網(wǎng)絡(luò)動(dòng)態(tài)補(bǔ)償模型進(jìn)行訓(xùn)練，可使微波透射石油含水率測(cè)量結(jié)果的補(bǔ)償過程具有尋優(yōu)全局性和精確性。實(shí)驗(yàn)結(jié)果表明，利用該技術(shù)對(duì)石油含水率測(cè)量結(jié)果進(jìn)行校正是一種有效的方法，具有一定的應(yīng)用價(jià)值。
關(guān)鍵詞:石油含水率神經(jīng)網(wǎng)絡(luò)微粒群優(yōu)化算法動(dòng)態(tài)補(bǔ)償測(cè)量精度

Abstract; To against the problem of big error in measuring moisture in petroleum by traditional microwave transmission method, the dynamic compensationte chniqueb asedo nn euralne tworkis p roposed.T woo fth ep arameters,i. e. m icrowavea tenuationa ndp hasesh ifta reta kenas thein putso fth ed ynamicc ompensationm odel.C onsideringth es hortcomingso fco nventionalB Pa lgorithm,e .g. c onvergings lowlya nde asily trapping a local minimum value, a learning algorithm using particle swarm optimization《PSO) is adopted to train the neural network。compensation model. This wil enable the compensation process optimal and accurate overal. Experiments show that the use of the technology in calibrating the measurement result of moisture in petroleum is efective and has certain applicable value.
Keywords;P etroleum Moisture Neuraln etwork Particle。optimizationa lgorithm Dynamicc ompensation Measurementa ccuracy