Study On Modeling And Optimization Of P-xylene Oxidation Process In The Production Of Pure Terephthalic Acid

This paper is based on the project of p-xylene oxidation process with researches on modeling and optimization. In recent years, with the rapid development of PTA production, the technology on p-xylene oxidation is directly related to the level of production of PTA, quality, yield, and raw material losses. In response to the optimization of the production equipment and the adjustment of the load, it is of great significance to establish the model that reflect the p-xylene oxidation accurately, to optimize, and ultimately to guide the actual operation. With the optimization, the material consumption is reduced, the conversion rate of p-xylene to TA is raised, the resources is conserved effectively, then the emissions of the toxic gas that generated during the process are reduced, the atmospheric pollution is also improved, these approaches are greatly response to the energy conservation promoted by the state.The p-xylene oxidation process is very complex, as the process is completed under high temperature and pressure with the catalyst and the process involves the solid-liquid-gas equilibrium. It is very important to establish an accurate model that reflects the actual p-xylene oxidation process, as it is the basis of the follow-up work which is related to the optimization. The main content and innovation of this research can be summarized as follows:First, for optimization of p-xylene oxidation process, some improvements were made and the artificial bee colony algorithm integration of differential evolution algorithm (DE-ABC) was proposed. The algorithm introduces differential evolution strategy to random mutations strategy, the incremental strategy of crossover probability factor and scaling factor is adopted, and the overall strategy of stagnation judgment and treatment is used instead of the local strategy. By contrast of the test function, the results show that the improved algorithm has better performance than the basic bee colony algorithm. The improved algorithm is excellent to solve the diversity inadequate of the population in the early iteration and easily sticking into the local optimum in the latter part of the iteration.Then, the p-xylene oxidation process model is established, and the model is modified according to the existing research, which, as a result, makes the model matching with the actual conditions. The sensitivity analysis is also established. The analysis result shows that the model can accurately reflect the changes of the actual condition.Finally, according to the long-period operation data, to meet the conversion rate of p-xylene to TA, the DE-ABC algorithm is used to the optimization of p-xylene oxidation with the target minimum function of CO and CO2 and the constraints based on the actual situation. After optimization, the model runs stable, and has a better performance, which gives a guide to the plant operation and adjustment.