基于模糊模型挖掘的NOx排放MAP图标定方法研究

Research on Calibration Method of NOx emission MAP Based on Fuzzy Model Mining

针对柴油机原始NOx排放MAP图标定过程中过度依赖于人工经验、时间成本高的问题,本文提出一种基于模糊模型挖掘的NOx排放标定方法。首先,该方法利用所有采集的NOx排放数据计算模糊规则的支持度,通过最大支持度的选取,实现模糊规则的挖掘提取。其次,通过对所有数据的支持度计算,削弱了异常点和噪声的干扰,增加了模型的鲁棒性。为了进一步提高模型的精度,本文设计了基于梯度下降法和输入变量模糊空间增量算法的参数优化过程。整体算法通过多个嵌套循环程序来实现,降低了模型训练难度,缩短了开发周期,适合工程应用。最后,本文将改方法应用于一款柴油机原始NOx排放MAP图的标定,实验结果显示了该方法的准确性、鲁棒性和实用性。

Facing the problems of excessive dependence on manual experience and high time cost in the process of determining the original NOx emission MAP of diesel engine,a NOx emission calibration method based on fuzzy model mining is proposed in this paper.Firstly,this method uses all collected NOx emission data to calculate the support degree of fuzzy rules,and realizes the mining and extraction of fuzzy rules through the selection of the maximum support.Secondly,by calculating the support of all data,the interference of outliers and noise is weakened and the robustness of the model is increased.In order to further improve the accuracy of the model,a parameter optimization process based on gradient descent method and input variable fuzzy space increment algorithm is designed in this paper.The whole algorithm is realized by multiple nested loop programs,which reduces the difficulty of model training and shortens the development period.It is suitable for engineering application.Finally,the modified method is applied to the calibration of the original NOx emission MAP of a diesel engine.The experimental results show the accuracy,robustness and practicability of the method.