增压柴油机零维预测燃烧模型建模方法

发布时间：2018-08-08 11:57

【摘要】：基于三韦伯方程建立了增压柴油机燃烧模型,并分别在两个不同尺寸的增压器下负荷特性共142个试验工况点对其校准,得出142个样本模型和各工况点下燃烧模型中的9个燃烧参数.考虑了柴油机转速、循环喷油量、喷油始点、进气压力和进气温度5个运行参数对9个燃烧参数的影响,用同样的数据建立了9个燃烧参数MAP图和神经网络辨识器(NNI),分别建立了基于NNI和MAP的零维预测燃烧模型(NNI-model和MAP-model),进行了校准工况点仿真计算,结果表明:NNI-model和MAP-model均与样本模型具有较高的一致性,NNI-model仿真精度稍差于MAP-model.将NNI-model和MAP-model分别进行了非校准工况仿真计算,结果表明:NNI-model和MAPmodel均与试验结果基本吻合,NNI-model仿真精度仍较高,而MAP-model仿真精度明显变差,与MAP-model相比NNI-model避免了多次线性插值累积误差,在偏离校准工况点时仍保持较高精度,泛化性更好,更适用于增压柴油机零维预测燃烧模型仿真研究.
[Abstract]:The combustion model of turbocharged diesel engine is established based on the three Weber equation, and its load characteristics are calibrated at 142 test conditions at two different size turbochargers. The combustion parameters of 142 sample models and 9 combustion models under different operating conditions are obtained. The effects of five operating parameters, such as diesel engine speed, cycle fuel injection rate, fuel injection point, inlet pressure and inlet temperature, on the nine combustion parameters are considered. Using the same data, nine MAP diagrams of combustion parameters and (NNI), neural network identifiers are established, respectively, the zero-dimension predictive combustion models (NNI-model and MAP-model) based on NNI and MAP are established, and the simulation calculation of calibration operating points is carried out. The results show that both the MAP-model and the NNI-model are consistent with the sample model. The simulation accuracy of the NNI-model is slightly lower than that of the MAP-model. The simulation results of NNI-model and MAP-model show that both the NNI-model and MAPmodel are in good agreement with the experimental results, while the simulation accuracy of MAP-model is obviously worse. Compared with MAP-model, NNI-model avoids multiple linear interpolation errors. When deviating from the calibrated working condition, the accuracy is high and the generalization is better, which is more suitable for the simulation of zero dimensional predictive combustion model of turbocharged diesel engine.
【作者单位】：哈尔滨工程大学动力与能源工程学院;
【基金】：国家科技支撑计划资助项目(2015BAG16B01)
【分类号】：TK421.2

【相似文献】