轻烃回收装置数据挖掘及生产优化

以控制制冷量提高液烃收率为目标，将膨胀机进口温度、出口温度及膨胀机进出口温差作为响应值，膨胀机转速、喷嘴压力、膨胀比、膨胀机进出口压差、原料气流量、原料气压力等8个工艺参数则构成影响响应值的主要变量。通过对某轻烃回收装置进行生产数据挖掘，提出了优化该装置制冷系统的生产指导方案。生产试验表明，按该方案生产，可提高液烃收率5％以上。

Data mining and productive optimization for NGL plant

Aiming at increase of NGL recovery ratio by controlling capacity of refrigeration, the scheme chooses eight major variables: dilatomachine rotation speed and the data of import pressure minus output pressure, muzzle pressure, inflated rate, flux and pressure of raw gas, etc. Thus, predicting differences in temperature of dilatomachine. By mining the operating data of a light hydrocarbon recovery plant, a new production plan has been proposed to optimize the refrigeration system, which is evidenced to improve liquid hydrocarbon recovery ratio by 5. 89% .