基于GA-SVR算法的CO_(2)在地层水中溶解度模型研究

Study on CO_(2)solubility model based on GA-SVR algorithm in formation water

利用高温高压反应釜开展了不同温度、压力条件下CO_(2)在不同矿化度地层水中的溶解度实验,采用支持向量回归(SVR)和遗传算法(GA)建立模型(GA-SVR)来计算CO_(2)在地层水中的溶解度大小,利用实验溶解度数据对GA-SVR模型进行训练和验证。结果表明,GA-SVR模型预测结果的相对误差最小为1.60%,最大为3.73%,平均为2.46%,均方根误差为0.094,与其他3种模型(Chang模型、Furnival模型和PR-HV模型)计算结果相比,GA-SVR模型具有更高的精度。该研究成果为CO_(2)在地层水溶解度的计算提供了一种新方法。

Using high temperature and high pressure reaction kettle carried out under different conditions of temperature and pressure of CO_(2)in different salinity formation water solubility experiment,using support vector regression(SVR)and genetic algorithm(GA)to establish a model(GA-SVR)to calculate the solubility of CO_(2)in the formation water size,solubility data of experiment is used to analyse the GA-SVR model training and validation.The results show that the relative error of GA-SVR model is 1.60%,3.73%at most,2.46%on average,and 0.094 at root mean square.Compared with the calculation results of the other three models(Chang model,Furnival model and PR-HV model),the GA-SVR model has higher accuracy.The research results provide a new method for the calculation of the solubility of CO_(2)in formation water.