【摘要】：在符號微分法的基礎(chǔ)上結(jié)合差商和自動微分技術(shù)提出一種實用高效的混合求導(dǎo)算法 ,提高了化工過程系統(tǒng)優(yōu)化的效率 .該算法將被求導(dǎo)對象的各個子模塊按照是否適于用符號微分求導(dǎo)分為簡單模塊和復(fù)雜模塊 ,對前者用符號微分求導(dǎo) ,對后者用差商法求導(dǎo) .針對該算法的關(guān)鍵技術(shù)———子模塊導(dǎo)數(shù)累積問題討論了鏈?zhǔn)角髮?dǎo)法和攝動傳遞法 .對某廠精餾塔操作優(yōu)化問題的分析和計算表明 ,引入混合求導(dǎo)算法提高了優(yōu)化的效率
[Abstract]:Based on the symbolic differential method, a practical and efficient hybrid derivation algorithm based on difference quotient and automatic differential technique is proposed, which improves the efficiency of optimization of chemical process system. In this algorithm, each sub-module of the derived object is divided into simple module and complex module according to whether it is suitable to use sign differential derivation. The former is derived by sign differential method, and the latter is derived by difference quotient method. The chain derivation method and perturbation transfer method are discussed for the sub-module derivative accumulation problem, which is the key technology of the algorithm. The analysis and calculation of the operation optimization of a distillation column in a factory show that the hybrid derivation algorithm is introduced to improve the efficiency of optimization.
【作者單位】： 浙江大學(xué)控制科學(xué)與工程系系統(tǒng)工程研究所 浙江大學(xué)控制科學(xué)與工程系系統(tǒng)工程研究所 浙江大學(xué)控制科學(xué)與工程系系統(tǒng)工程研究所
【基金】：國家自然科學(xué)基金項目 (No 2 0 2 760 62 ) 國家重點(diǎn)基礎(chǔ)研究發(fā)展規(guī)劃項目 (No 2 0 0 2CB3 12 2 0 0 )資助~~
【分類號】：TQ015.9

【相似文獻(xiàn)】

相關(guān)期刊論文 前2條

1 邵之江,李翔,錢積新;化工過程系統(tǒng)優(yōu)化中的一種混合求導(dǎo)算法[J];化工學(xué)報;2003年10期

2 ;化學(xué)工程[J];中國石化文摘;2004年07期

，

本文編號：2467856