Abstract:
The blending effect octane model and its online updating method during the gasoline pipeline blending process are described in this work. Since gasoline's octane number exhibits strong nonlinear attributes during the blending process, a linear model would cause greater prediction error. Although the classical ethyl model is a nonlinear model, it requires a lot of maintenance data during the later modeling process, which results in bigger amounts of work such that it is difficult for the model to adapt quickly the changing of blending condition. This work proposes a blending effect octane model based updating method, which uses recursive least squares method to update the model parameters online so as to ensure the effectiveness of blending effect model for larger range of operating mode. Moreover, the proposed method has simple principle, fast computing ability, and better adaptability to various gasoline pipeline blending.