Abstract: Using aftertreatment devices is expensive for nonroad diesel engines. It is one of the economical and effective methods to simultaneously reduce the original engine emissions and maintain reasonable fuel economy by optimizing combustion. Taking a small agricultural diesel engine developed in China IV stage emission regulations for nonroad mobile machinery as the research object, firstly, a onedimensional thermodynamic simulation model was established through GTPower software according to the existing test data. Secondly, after verifying the accuracy of the model, the response surface method was used to carry out experimental design for the previous seven working conditions of the offroad diesel engine steadystate test cycle, and the regression models of the key performance parameters of the diesel engine were obtained by the stepwise method. Thirdly, the multiobjective optimization of fuel consumption and emissions was conducted using NSGA-Ⅱ. The desired optimal solution and its corresponding calibration parameters were determined from the Pareto optimal solution by TOPSIS under each working condition. Finally, the optimization results were verified by bench test. The results show that it was feasible touse NSGA-Ⅱ for multiobjective optimization of fuel consumption and emissions and screen the optimal solution based on TOPSIS, and the optimization results were accurate. Compared with the original diesel engine, the weighted brake specific fuel consumption, NOX, and soot of the eight working conditions of the optimized steadystate test cycle has been reduced by 2.8%, 27.4%, and 18.2%, respectively. The optimized CO, HC+NOX, and PM emissions of this steadystate test cycle are 2.811 g/(kW·h), 3.342 g/(kW·h), and 0.015 g/(kW·h), respectively, which meets China nonroad stage IV emission standards.

Key words: diesel engine, fuel consumption, emissions, multiobjective optimization

摘要： 非道路用柴油机使用后处理系统的成本较高，通过优化燃烧减少原机排放并保持合理的燃油经济性是经济且行之有效的方法之一。以一款处于非道路第四阶段开发的小型农用柴油机为研究对象，基于已有的试验数据，通过GTPower软件建立其一维仿真模型，在验证模型准确性的基础之上，针对非道路用柴油机稳态测试循环八工况的前7个工况，采用响应曲面法进行试验设计，利用逐步法获得该柴油机关键性能参数的回归模型，基于NSGA-Ⅱ对其进行油耗和排放的多目标优化，并采用TOPSIS筛选出各个工况下理想的最优解及其对应标定参数，最后通过台架试验验证优化结果。研究表明：采用NSGA-Ⅱ进行油耗和排放多目标优化，并基于TOPSIS筛选的最优解是可行的，优化结果较为准确；与原机相比，优化后小型农用柴油机稳态测试循环八工况的加权油耗、NOX以及碳烟排放分别降低2.8%、27.4%以及182%；优化后的整机CO、HC+NOX、PM排放分别为2.811 g/(kW·h)、3.342 g/(kW·h)、0.015 g/(kW·h)，满足中国非道路用柴油机第四阶段排放标准。

关键词: 柴油机, 油耗, 排放, 多目标优化