Research And Realization Of Crude Oil Stabilizing Heating Furnace Modeling And Parameter Optimization Algorithm

In the process of crude oil stability, since the temperature directly influences the yield of non-condensable gas and light hydrocarbon, temperature controlling in the crude oil heating furnace is extremely important. Therefore, to advance the technique of temperature regulating in the process can significantly increase the production.In this thesis, as well as multidisciplinary methodology, this thesis intensively studied the application of neural network and particle swarm optimization algorithm in the parameter optimization of crude oil stabilizing heating furnace.Through the analysis of the operating status and existing problems of the crude oil stabilizing heating furnace,the factors affecting the stability of the outlet temperature of the heating furnace are mainly in the following aspects: the temperature of the gateway of the furnace, the flow of the crude oil, the entering pressure of crude oil, the flow of fuel gas, the outlet pressure of crude oil, the inlet valve opening of the furnace. BP neural network algorithm was chosen to build the mathematical model of the heating furnace after analyzing different kinds of neural network algorithms’ features, advantages and scope of application, as well as the connection between parameters and the object of study. Then, through theoretical analysis and contrast test, we finally used the particle swarm optimization to optimize parameters. By calculating the minimum of the objective function to seek the optimal parameter combination of process parameters of the original stable furnace.Through the research of this thesis, the optimal parameter combination of crude oil stabilizing heating process and then carried out field comparison of the result obtained by contrastive analysis. Temperature was controlled better after optimization. The method applied in this thesis effectively advanced the function of crude oil stabilizing heating furnace and solved problems like large temperature difference among the four branch pipes, drift coking and big temperature fluctuation in the furnace before optimization. In addition, the new method helps the operator make targeted adjustments to parameters, which greatly increases the capability of regulating the parameters and provides essential theoretical foundation and practical guide for the adjustment of technological parameter of crude oil stabilizing heating furnace.