基于模型的柴油机微粒捕集器碳烟加载量估计研究

发布时间：2018-01-01 23:03

本文关键词：基于模型的柴油机微粒捕集器碳烟加载量估计研究　出处：《吉林大学》2015年硕士论文　论文类型：学位论文

【摘要】：汽车工业的可持续发展面临着能源与环境的双重挑战。日益频发的大气雾霾和日益严厉的排放法规，对柴油机的碳烟排放提出了严格要求，迫切需要通过“机外”后处理系统实现对柴油机碳烟的进一步净化。柴油机微粒捕集器（DPF）是目前公认的降低碳烟排放最有效的技术之一。为了降低由于DPF内碳烟持续加载引起的背压升高对柴油机经济性和动力性的影响，需要定期进行再生。DPF再生面临着再生不完全和不安全（DPF载体烧熔和破裂等）的问题，准确判断DPF再生时机成为解决该问题的关键。而准确判断再生时机实质是需要精确估计DPF的碳烟加载量。传统的碳烟加载量估计方法估计精度有待进一步提高，且实时应用困难。因此，本文基于MATLAB/Simulink平台建立了适用于ECU实时计算的DPF碳烟加载模型，开展了基于模型的DPF碳烟加载量估计研究。针对DPF一维三维建模复杂的特点，本文建立了基于填充床捕集理论和零维集总参数方法的碳烟加载模型。首先，依据填充床捕集理论对DPF载体壁面进行离散分层，基于布朗扩散沉积机理和直接拦截沉积机理建立了DPF壁面加载模型。然后，引入分配系数，实现碳烟加载质量在碳烟层和壁面内的分配，从而实现对不考虑被动再生的碳烟加载过程的建模。最后，结合DPF碳烟加载过程可能发生的宏观多相化学反应（NO2辅助氧化等），依据各反应的起始作用温度，建立了四种模式的DPF零维集总参数质量、能量平衡模型，实现考虑被动再生时DPF加载过程碳烟层和壁面内碳烟质量更新的建模。从而建立了考虑被动再生时DPF碳烟加载估计模型。针对DPF碳烟加载估计模型控制参数，提出了基于线性拟合和最小二乘法的模型参数辨识方法。基于试验数据，实现了对所选稳态加载工况下的DPF捕集参数和化学动力学参数的辨识。通过柴油机台架试验对DPF碳烟加载估计模型进行了验证并对DPF碳烟加载估计模型的仿真结果进行了分析。验证试验包括后处理系统只加装DPF和加装DOC+DPF两组试验。每组试验分别进行洁净DPF压降试验和DPF稳态加载试验。通过比较模型预测的洁净DPF压降、DPF稳态加载压降特性与试验测得的洁净DPF压降、DPF稳态加载压降特性，间接对DPF碳烟加载估计模型进行验证。结果表明，模型对洁净DPF压降和DPF稳态加载压降特性的预测与试验结果有很好的吻合性。基于以上提出的DPF碳烟加载估计模型，运用Simulink仿真软件对其进行了仿真研究。仿真结果表明，基于该模型的DPF碳烟加载量估计，由于考虑了被动再生对碳烟加载量的影响，，碳烟加载量的估计精度得到提高。并且该模型很好的实现了对DPF碳烟加载过程不同时刻的碳烟加载量和压降特性的跟踪。DPF碳烟加载量估计数学模型的建立，为进一步实现基于模型的DPF再生控制策略的开发打下了基础。
[Abstract]:The sustainable development of the automobile industry is facing dual challenges of energy and environment. The increasingly frequent haze and increasingly stringent emission regulations, put forward strict requirements on smoke emission of diesel engine, the urgent need for the adoption of "machine" system to achieve the postprocessing for further purification of diesel soot of diesel particulate filter. Device (DPF) is now recognized as one of the most effective techniques to reduce soot emissions. In order to decrease due to the increase of soot in DPF continuous loading caused by backpressure on the fuel economy and power, the need for regular regeneration.DPF regeneration facing regeneration is incomplete and not safe (DPF vector melting and rupture etc.) problems, accurately determine the DPF regeneration time becomes the key to solve the problem. And accurately determine the regeneration time is need to accurately estimate soot loading DPF soot loading. The traditional estimation method to estimate the precision It needs further improvement and real-time application is difficult. Therefore, based on MATLAB/Simulink platform, a DPF soot loading model suitable for ECU real-time computation is established, and DPF based soot loading estimation is carried out based on the model.
According to the characteristics of DPF one-dimensional complex 3D modeling, this paper established the soot loading model of packed bed trapping theory and zero dimensional lumped parameter method based on. Firstly, according to the packed bed trapping theory of DPF vector discrete wall stratification, Brown diffusion and deposition mechanism of direct interception sedimentation mechanism established DPF wall loading based on the model. Then, the distribution coefficient, realize the distribution of soot loading quality in the soot layer and the wall surface, thus realizing the modeling without considering the soot loading process of passive regeneration. Finally, combining the macro DPF soot loading process may occur by chemical reaction (NO2 assisted oxidation), on the basis of effect of starting temperature of the reaction, set up four kinds of mode DPF and zero dimensional lumped mass and energy balance model, modeling update soot mass considering passive regeneration DPF loading process of soot layer and inside the wall. In order to establish The DPF carbon smoke loading estimation model is considered when passive regeneration is used.
Aiming at the control parameters of DPF soot loading estimation model, a model identification method based on linear fitting and least square method is proposed. Based on the experimental data, the DPF trapping parameters and chemical kinetic parameters of the selected steady-state loading conditions are identified.
The engine bench test of DPF soot loading estimation model was verified on DPF and soot loading estimation model simulation results are analyzed. Experiments including postprocessing system only adding DPF and adding DOC+DPF two group test. Each test were clean DPF drop test and DPF test. Through static loading clean DPF drop the comparison of model predictions, clean DPF DPF drop steady loading pressure drop characteristics with the measured DPF, steady-state load pressure drop characteristics, indirect DPF soot loading estimation model was verified. The results show that the model of clean DPF pressure drop and pressure drop characteristics of DPF steady-state load prediction and the experimental results have a good agreement.
Estimation model based on soot loading DPF proposed above, the use of Simulink simulation software is simulated. The simulation results show that the DPF estimation of soot loading based on the model, considering the influence of passive regeneration of soot loading, the soot loading estimation precision level improved and the model. To achieve a good tracking.DPF soot loading on soot loading and pressure drop characteristics of DPF soot loading process in different time to set up the mathematical model for the further implementation of the estimation model of DPF based on the development of regeneration control strategy to lay the foundation.

【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TK421.5

【参考文献】