基于模拟退火-蚁群算法的输油管道碳排放量优化研究

Research on carbon emission optimization of oil pipeline based on simulated anneal⁃ing-ant colony algorithm

为实现输油管道的节能降耗,以碳排放量最小为目标函数,在水力约束和热力约束的条件下,采用模拟退火-蚁群算法的机器学习方法对加热炉、外输泵的组合方式及运行参数进行求解,并在考虑设备运行可靠性的基础上,设计了全年运维方案。结果表明:优化后单日耗油量和耗电量均有所下降,单日碳排放量降低20%左右;对外输泵进行了可靠性分析,依据概率密度函数可知泵的正常检修周期应为1650 h(68 d),在连续运行1120 h(46 d)后应进行预防性保养;环境温度与碳排放量的优化效果呈负相关,全年运维方案中7—9月的碳排放优化效果较差,全年可减少碳排量185.61 t。研究结果可为油气田企业双碳目标的实现提供技术支撑。

In order to achieve energy conservation and consumption reduction of oil pipelines,with the minimum carbon emission as the objective function,under the conditions of hydraulic and thermal constraints,the machine learning method of simulated anneal-ant colony algorithm is used to solve the combination mode and operation parameters of heating furnace and external pump.What's more,on the basis of considering the reliability of equipment operation,the annual operation and maintenance scheme is designed.The results show that the single-day oil consumption and power consumption are reduced after optimization,and the carbon emission is reduced by about 20%per day.The reliability analysis of the external pump has been carried out.According to the probability density function,it is known that the normal maintenance period of pump should be 1650 h(68 d)and preventive maintenance should be carried out after 1120 h(46 d)continuous operation.The environmental temperature is negatively correlated with the optimization effect of carbon emissions.Especially,in the annual operation and maintenance scheme,the optimization effect of carbon emissions is worse from July to September,and the annual carbon emission can be reduced by 185.61 t.The research results can provide technical support for the realization of dual carbon targets in oil and gas field enterprises.