| As a vital link of gasoline production, gasoline blending is the popular research focus. The gasoline blending process influences the gasoline products quality and the economic income of the refinery directly. After a comprehensive introduction about the gasoline blending current situation and all the existing octane number prediction models, a novel gasoline blending octane number prediction model is proposed. It shows a better accuracy compared with two existing classic octane prediction models.Gasoline blending is a complicated nonlinear process in consideration of all kinds of nonlinear gasoline properties like the octane number. So the traditional optimization methods can’t meet the requirements of low cost and low quality waste. Then a hybrid optimization method with Gaussian mutation is proposed based on current situation. This optimization algorithm has a better accuracy and convergence rate compared with the particle swarm optimization method and differential evolution and at the same time it could solve the premature convergence problem of the particle swarm optimization. The proposed hybrid optimization method is used to calculate the gasoline blending recipe for the first time. The result shows that it has a better effect compared with particle swarm optimization and differential evolution algorithm and it could reduce the cost of the gasoline blending process and the RON quality giveaway. Meanwhile it could increase the blending success rate.After the above research and discussion, a new system for online gasoline blending process is developed based on Honeywell PKS. It is developed for the batch blending characteristic specially. It has a lot of advanced functions:task management, recipe optimization, recipe management, continuous control, batch control, sequential control, fault management, real-time release via web. This system will make the gasoline blending process more automatic and intelligent. At the same time.it will increase the income of the refineries. |
PDF Full Text Request