Prediction for mixed oil trailing based on Gamma-χ² distribution function

The phenomenon of mixed oil tailing is closely related to the accuracy of batch tracking, the reliability of scheduling plan, and the economy of oil-cutting scheme. Therefore, the accurate description on mixed oil tailing is quite significant to control the mixed oil volume and predict the oil quality. In this paper, the variation rule of mixed oil tailing curve was analyzed, and the simulation model for investigating the asymmetry distribution of the mixing interface was established. Then, the concentration distribution characteristic and its gradient change law were obtained on the basis of the data measured at the mixing interface, and the concentration gradient change curves were approximated by Gamma and Gamma-χ² distribution functions, respectively. The distribution functions were simulated by programming and their key parameters were inverted, the concentration distribution curve of mixed oil was established. Based on this, the relationships between the parameters of pipeline systems and the parameters of distribution functions were analyzed under different boundary conditions and operating modes, and the fitting formula was obtained. Additionally, partial data was applied to predict the whole concentration distribution curve, so as to realize the real-time online prediction on the concentration curve. The research results can provide effective guidance for the prediction of mixed oil volume and the operation of mixed oil cutting.