【作者】 张进春；

【导师】 吴超；

【作者基本信息】 中南大学 ， 安全管理工程， 2011， 博士

【摘要】 为了实现煤的高碳资源低碳化利用,必须促进能够有效提高煤炭转化效率且环保效益好的以煤气化为核心的新型煤化工的发展。气流床煤气化技术具有煤种适应广、气化效率高、生产负荷大、环保效益好等优势,必将在新型煤化工中发挥着重要作用。气流床煤气化工艺性能稳健优化与控制的研究对于提高煤炭的清洁高效利用,保证气化生产的经济、稳定、安全运行有着重要意义。本文面向气流床煤气化生产过程,以气流床煤气化性能的稳健优化与波动控制为主要研究目的,以典型的气流床工艺——Shell粉煤加压气流床为背景,综合应用实验设计、统计分析、稳健优化和统计控制技术,旨在建立通过调整和优化气流床煤气化工艺参数,使得气流床煤气化各项工艺性能指标在多种随机噪声的影响下依然能够尽最大可能的保持最优与稳定的理论与方法,为气流床煤气化优化生产与稳定生产提供指导。本文研究主要包括：1.选择相互独立的气流床煤气化性能指标CO含量百分比、H2含量百分比、产气率和碳转化率为评价指标。以ChemCAD为仿真平台,采用Gibbs自由能最小化方法,建立了Shell气流床煤气化仿真模型。经与典型的Shell煤气化装置SCGP-1实验数据对比,结果表明所建仿真模型具有较好精度。2.设计了基于Taguchi方法进行可控工艺参数对气化性能的稳健效应分析流程。构建了采用对照进行气流床可控工艺参数的稳健效应估计,并基于半正态概率分布对效应估计进行显著性判别,采用灵敏度分析进行气流床煤气化可控工艺参数对气化性能进行稳健效应分析的方法。针对Shell气化系统进行了实例分析。3.基于DRSM法开展气流床煤气化工艺性能的稳健优化问题研究。构建了气化性能指标与可控工艺参数的双响应面模型,提出了应用改进的损失函数模型将各气化性能指标的均值和方差统一成一个稳健评价指标,设计了气化性能稳健优化的GA随机搜索优化算法。针对煤气化性能指标为多目标的特性,采用AHP1-9标度确定各评价指标的权重,然后采用Desirability函数将各气化性能评价指标集成为一个优化目标,并利用GA随机搜索优化算法进行优化求解。最后针对Shell气化工艺进行了单目标稳健优化与多目标稳健优化,并分别与确定性优化结果加以对比,结果表明,所提出的气流床煤气化性能的单目标稳健优化与多目标稳健优化方法可以实现气流床煤气化性能稳健优化的目的。4.煤气化产品需求具有波动性,而JIT精益生产模式是应对市场需求波动的先进的管理模式。本文在分析了波动需求条件下煤气化生产系统JIT精益生产模式的基础上,针对气流床煤气化工艺性能稳健性,研究了波动需求导致生产强度动态调整的JIT精益生产模式下气化性能稳健优化的数学模型,提出来基于“分解——综合”原则,采用“精英优化”的气化性能的JIT精益稳健优化求解策略,进而提出了基于非线性偏最小二乘的优化方法。最后针对Shell煤气化工艺系统进行了实例分析,结果表明所建立的波动需求条件下的气流床工艺性能JIT精益稳健优化方法是切实可行的。5.以过程能力指数(PCI)作为气化性能稳健性评价的基本指标,进行气流床煤气化工艺性能稳健性的预测评价,以确定所求解最优操作条件是否满足生产工艺要求或下游产品生产需要。由于煤气化性能评价属于多变量过程能力指数评价问题,而目前多元过程能力指数存在计算复杂、难以推广应用等问题,本文提出了一种基于偏差分量模型的线性加权MPCI,并采用单纯型格子点实验设计求解线性加权权重的方法。进而本文提出了基于蒙特卡罗模拟技术,以线性加权MPCI作为气化性能稳健性评价指标进行煤气化性能稳健性预测评价。最后针对Shell煤气化工艺系统,在获得的稳健优化解的基础上,进行了预测评价,结果表明本文所提的稳健性预测评价方法可以对气流床煤气化工艺性能的稳健性进行有效评价。6.针对生产过程中气流床煤气化性能的稳健性,采用统计过程控制(SPC)技术进行监控。构建了基于单变量SPC与多变量MSPC的气化性能“失控”与“失稳”联合监控框架体系。针对休哈特单变量、Totelling T2多变量统计失控监控可能存在漏发报警的不足,采用基于PCA的多变量统计过程控制监控进行气流床煤气化性能监控验证。同时针对传统统计失稳判别准则适用于离散制造系统,而气流床煤气化性能监控要求连续在线监测的特点,开发了基于RBF神经网络的气流床煤气化性能统计失稳智能诊断模型。最后针对Shell煤气化工艺系统,进行了实例仿真验证,结果表明,本文所提的气流床煤气化的统计过程控制可以有效监测气流床气化性能,保证气流床煤气化稳定生产。更多还原

【Abstract】 In order to realize low carbonization using of coal, it must be promoted for the development of the new coal chemical industry based on coal gasification for it’s high carbon conversion and effective environmental protection. Entrained flow coal gasification technology will play an important role in the new coal chemical industry because of it’s advantages such as high adaptability to a wide range of coal types, high gasification efficiency and production load, and effective environmental protection. So it is important to study on the performance of the entrained flow coal gasification robust optimization and control, because it is significance to improve the efficient use of clean coal gasification and ensure coal gasification production economically, stably and safely.This research aims to realize the performance of coal gasification robust optimization for the entrained flow coal gasification production, that is, to optimize the performance and control the fluctuation as well. The typical entrained flow coal gasification technology-Shell coal gasification process was taken as the research background, and many advanced mathematical methods including Design of Experiment (DOE), statistical analysis, robust optimization and Statistical Process Control (SPC) were applied comprehensively in this dissertation. The main research purpose is to establish the theory and method that ascertaining a set of optimum controllable process parameters which can keep the performance of entrained flow coal gasification stable and optimizing as possible under the influence variety of random noise factors such as the fluctuation of moisture, ash or grain size of coal et al. This reassure will provide the instructions for the optimization operation and stable operation of entrained flow gasifier. In summary, the dissertation includes the following contents and contributions:1. The independent performance indexes of entrain flow coal gasification including CO%, H2%, gas yield and carbon conversion were chosen as the evaluation indexes. Shells coal gasification simulation model was founded with the Minimum of Gibbs Energy based on ChemCAD software. The model was verified by the experimental data of typical Shell gasification units SCGP-1. The verification results showed that the simulation model was accuracy.2. The analysis procedure of the robustness effect of the controllable process parameters on the performance of the entrained flow coal gasification based on Taguchi method was designed. The analysis method was founded using contrast to estimate the robustness effects of the controllable process parameters, half normal distribution to test the effects’ significances and the sensitivity to analysis the effects. Then the Shell gasification was taken as the example and analyzed by the above mentioned method.3. DRSM method was used to study on the performance of entrained flow coal gasification robust optimization. Double response surface models, mean-variances RSM models, about the controllable process parameters and the performance indexes of the gasification were founded. An improvement of loss function model was founded and applied to unite mean and variance of each gasification performance indexes into one robust evaluation index. A GA random search optimal algorithm was designed to solve the problem of the coal gasification performance robust optimization. For the indexes of the entrained flow coal gasification performance is multi-objects optimization, so 1-9 scale in Analytic Hierarchy Process (AHP) was used to evaluate the weight of each index, and the Desirability function was applied to integrate the indexes into one optimal object, GA random search optimization algorithm was used to search the optimal solution. Finally, according to the Shell gasification processes, the single objective robust optimal solution and multi-objective robust optimal solution were gained, and that of the certainty optimal solution were also gained to contrast. The contrasting results show that the proposed method about the entrained flow coal gasification performance of single objective robust optimization and multi-objective robust optimization can realize the robust optimization purposes.4. The demands of the coal gasification products have the characters of volatility, and JIT lean production mode is the advanced management to cope with the market demand fluctuation. The JIT lean production mode under the condition of the fluctuation demand for coal gasification production system was analyzed. Then the entrained coal gasification performance dynamic robust optimization while the entrained flow coal gasification implement JIT lean production was study, and the mathematical model was founded. The elitist optimizing strategy based on the principle of "Decomposition and Synthesis" was put forward. And then the optimization method based on nonlinear partial least-square was put forward. Finally, Shell coal gasification process system was analyzed, and the results show that the founded robust optimization method for the performance of the entrained flow coal gasification of the JIT lean production mode is feasible.5. Process Capability Index (PCI) was taken as the evaluation index to forecast evaluate the robustness of the performance of the entrained flow coal gasification, so as to examine the performance of coal gasification according to the optimal robust solution of the controllable operating conditions whether meet the production process requirement or downstream products production needs. Because coal gasification performance evaluation belong to multivariable process capability index (MPCI) evaluation, and currently MPCI is complex and difficultly application, a linear weighted MPCI based on the deviation model was put forward and simplex grid-point design method was adopted to get the weights. Then the method that the robustness performance of entrained flow coal gasification be forecast evaluated based on Monte Carlo simulation technique and with the linear weighted MPCI was proposed. Finally, Shell coal gasification was taken as the example and evaluated for the special optimal process parameters gained former study. The results show that the proposed method can forecast the performance robustness of entrained coal gasification process effectively.6. Statistical process control (SPC) technology was applied to monitor the production process performance of entrained coal gasification. A combined monitoring frame system for the performance of entrained coal gasification based on single variable SPC and multivariate MSPC jointed was constructed, which can inspect the situation of "out of control" and "instability". In allusion to the shortage of current SPC that alarm-omitted may occur, the MSPC based on PCA was adopted to validate the monitoring results. At the meanwhile, the traditional instability criterion in SPC is suitable for the distributed manufacturing system, while the entrained coal gasification belong to process industry, the performance monitoring needs on-line monitoring continuously. So an intelligent diagnosis model based on RBF neural network was developed for the entrain bed coal gasification performance statistics instability monitoring. Finally the application of the proposed monitoring method in Shell coal gasification proves the method can monitor the entrain bed gasification performance effectively and ensure the entrained gasification production stability.更多还原