479Q汽油机电控参数标定研究

Study on Calibration of Electronic Control Parameters for 479Q Gasoline Engine

利用AVL PUMA OPEN汽车动力测试台架对479Q发动机进行稳态电控参数的标定。为了寻求ECU与发动机的最佳配合,利用神经网络建立电控参数预测模型,通过对神经网络模型进行训练预测出稳态工况下发动机连续的ECU控制参数,并进行检验。研究结果表明,该预测模型具有较强的泛化能力,能够准确地预测出发动机电控参数。

The electronic control parameters of 479Q gasoline engine were calibrated with AVL PUMA OPEN automation system under stable conditions. The prediction model of electronic control parameters was built with the neural network to achieve the best matching between ECU and engine, and the continuous ECU control parameters under stable conditions were predicted by the training of model. Then the model was tested for further. The results show the prediction model is able to predict the electronic control parameters accurately because of the stronger generalization ability.