气化用煤配煤模型及动态配煤系统研究

发布时间：2018-02-01 21:25

本文关键词： 煤气化配煤预测配煤优化动态配煤　出处：《西安科技大学》2017年硕士论文　论文类型：学位论文

【摘要】：煤气化技术是目前大型煤化工企业广泛应用的一项重要生产技术,该项技术的长期运行实践表明,煤质的不稳定会影响气化炉装置长期稳定的运行,采用配煤来解决该问题是简便、经济、可行的方法之一,现有的配煤技术因对非线性变化的配煤煤质预测不准确,配比计算时对气化炉的煤质制约因素考虑不够周全,导致配煤煤质仍有波动,无法保证气化炉装置长期稳定的运行,且人工配煤计算已经无法满足新时期煤炭企业的发展需求。论文首先根据对配煤后煤质的变化规律的认识,分别利用多元线性回归、BP神经网络以及遗传算法(GA)优化BP神经网络三种方法建立了基于灰流动温度的配煤预测模型,通过三种模型的拟合效果以及误差分析和预测结果对比,得出:采用GA-BP神经网络构建的灰流动温度预测模型具有一定的可行性和优越性;并通过对目前较为经典的灰粘度预测模型对本论文所研究煤炭煤质的适应性研究,得出易于实现,且能有效预测气化用煤配煤灰粘度特性的基于灰粘度的配煤预测模型,从而达到准确预测配煤煤质的目的。其次根据宁东煤化工基地气化炉的实际入炉煤质关键制约因素,建立以水分、灰分、挥发分、硫份、发热量、灰流动温度、灰粘度为约束条件,以配煤价格,硫份、流动温度为目标函数的多目标配煤优化模型,通过MATLAB优化函数与遗传算法分别对模型求解,并从理论上和实际求解结果上分析对比,得出:遗传算法用于求解多目标配煤优化模型效果更佳,建立基于遗传算法的配煤优化模型,达到计算最优配比的目的。最后,通过对煤化工基地气化配煤生产过程的分析,采用C/S系统架构,Microsoft Visual Studio 2008开发平台,MFC等技术设计并实现了一套适用于气化用煤配煤的动态配煤系统,并将基于GA-BP神经网络的配煤预测模型及基于遗传算法的配煤优化模型应用到系统中。论文对气化用煤配煤模型及动态配煤系统的研究,能够充分考虑气化用煤配煤需要满足的约束条件,及要达到的经济、环保和气化炉稳定运行的目标,根据仓库中现有原料煤,计算出满足不同气化炉需求的最合理配煤比例,并能准确的预测配煤后煤质,降低气化炉因入炉煤质波动造成的故障率,且可以及时更新库存原料煤、气化炉等信息,从而解放劳动力,提高企业的经济效益,降低因煤质硫份过高造成的空气污染率。
[Abstract]:Coal gasification technology is an important production technology widely used in large scale coal chemical enterprises at present. The long-term operation practice of this technology shows that the instability of coal quality will affect the long-term stable operation of gasifier plant. It is one of the simple, economical and feasible methods to solve this problem by coal blending. The existing coal blending technology is not accurate for the nonlinear change of coal blending quality. When calculating the proportion of gasifier coal quality constraints are not fully considered, resulting in coal blending quality still fluctuate, which can not guarantee the long-term stable operation of gasifier equipment. The artificial coal blending calculation has been unable to meet the development needs of coal enterprises in the new period. Firstly, according to the understanding of the coal quality change law after coal blending, the multivariate linear regression is used respectively. BP neural network and genetic algorithm (GA) optimization BP neural network three methods based on ash flow temperature prediction model of coal blending. Through the fitting effect of the three models and the comparison of error analysis and prediction results, it is concluded that the grey flow temperature prediction model based on GA-BP neural network has certain feasibility and superiority; And through the research on the adaptability of the classical grey viscosity prediction model to the coal quality studied in this paper, it is easy to realize. And the coal blending prediction model based on ash viscosity can effectively predict the viscosity characteristics of coal ash for gasification. In order to accurately predict the coal quality of coal blending. Secondly according to the key factors of coal quality in coal gasifier in Ningdong coal chemical base the moisture ash volatile sulfur calorific value and ash flow temperature are established. The multi-objective coal blending optimization model with coal blending price sulfur content and flow temperature as objective function is solved by MATLAB optimization function and genetic algorithm respectively. From the theoretical and practical analysis and comparison of the results, it is concluded that the genetic algorithm for solving multi-objective coal blending optimization model is more effective, and establish a coal blending optimization model based on genetic algorithm. Finally, through the coal chemical base gasification coal blending production process analysis, using the C / S system architecture. A dynamic coal blending system suitable for gasification coal blending is designed and implemented by Microsoft Visual Studio 2008 development platform. The coal blending prediction model based on GA-BP neural network and the coal blending optimization model based on genetic algorithm are applied to the system. The coal blending model and dynamic coal blending system for gasification are studied in this paper. Can fully consider the gasification coal blending needs to meet the constraints, and to achieve the goals of economic, environmental protection and gasifier stable operation, according to the existing raw coal in the warehouse. Calculate the most reasonable proportion of coal to meet the different gasifier demand, and can accurately predict the coal quality after blending, reduce the failure rate caused by coal quality fluctuation in gasifier, and can update the stock of raw coal in time. The information of gasifier can liberate the labor force, improve the economic benefit of the enterprise, and reduce the air pollution rate caused by the excessive sulfur content of coal.
【学位授予单位】：西安科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18;TP311.52;TQ546

【参考文献】