比照测试柴油机负荷特性方法对某型废气涡轮增压柴油机进行了全工况热平衡台架试验,并以柴油机各稳定工况点台架试验的转速、加油尺杆位移、排气温度及相应输出扭矩的测试数据为训练、测试样本,通过BP神经网络的反复学习,得到了精度较高且泛化能力较强的柴油机动态输出扭矩预测模型。 According to the test of diesel engine load characteristic method, a full-scale heat balance bench test was carried out for a certain type of exhaust gas turbocharged diesel engine. The rotation speed, the displacement of foot rod, the exhaust temperature and the corresponding output torque Of the test data for the training, testing samples, repeated learning by BP neural network, the higher the accuracy and generalization ability of diesel dynamic output torque prediction model.