摘要
为研究某油田长输管道土壤腐蚀性,利用BP神经网络、遗传算法并借鉴相关标准的做法,选取土壤的自然电位、含水量、Cl^-质量分数、SO_4^(2-)质量分数、土壤电阻率、氧化还原电位、pH值作为评价指标,构建土壤腐蚀性GA-BP神经网络评价模型。结果表明:该模型在土壤腐蚀性评价中适用且更准确地反映了土壤的腐蚀特性,有助于掌握长输管道的土壤腐蚀状况。
In order to study the soil erosion of oil pipeline in an oil field,by introducing the BP neural network,genetic algorithm and learning from relevant standard,taking the natural potential of soil,water content,Cl^- mass fraction,SO4^（2-） mass fraction,soil resistivity,oxidation-reduction potential and pH as evaluation indexes,the evaluation model of the soil corrosive GA-BP neural network was built. The results show that the model is applicable to soil corrosion evaluation and reflects the corrosion characteristics of soil more accurately,which can help the management to master the soil corrosion of long pipelines.
作者
任欢
REN Huan(Daqing Oilfield Co., Ltd. Gas Company, Daqing 163000, China)
出处
《管道技术与设备》
CAS
2018年第3期41-44,共4页
Pipeline Technique and Equipment
关键词
BP神经网络
遗传算法
埋地管道
BP neural network
genetic algorithm
buried pipeline
