基于BP神经网络双燃料发动机放热率计算与试验

Calculation and experiment of combustion heat release in dual fuel engine based on BP neural network

利用BP神经网络理论,建立预测生物制气-柴油双燃料发动机放热规律的BP神经网络模型,经试验验证,预测模型误差在工程允许范围内,可以作为一种新型的预测双燃料发动机放热规律的方法.同时研究了双燃料发动机性能参数对放热规律的影响.结果表明,转速降低,燃烧始点相对提前,放热率峰值对应相位相对提前;转速不变负荷增加时,燃烧始点相对提前,引燃柴油的滞燃期缩短,最大放热率升高,对应的曲轴转角提前,但后燃现象严重.

A predictive model of heat release rate was established by means of BP neural network. After testing, the prediction model was proven to have high precision and could be used as a new means for the heat release calculation of the dual fuel engine. At the same time, the effeets of speed and load on the heat release rate were also studied. The results show that with the decrease of the speed, both the ignition timing and the corresponding phase of the top heat release rate of biogas-diesel dual fuel engine come earlier. But when the speed is fixed and the load increases, the ignition timing advances, the ignition delay period shortens, the maxium heat release rate increases, and its corresponding crank advanees with a serious phenomenon of post-combustion.