基于BP神经网的内燃机燃烧模型

Burning model of internal combustion engine based on neural networks

利用ＢＰ神经网良好的非逻辑归纳特性和高度的非线性特性映射能力，采用组群形式，对建立内燃机燃烧模型进行了尝试．结果表明：基于ＢＰ神经网的燃烧模型建模过程简洁，计算速度快，模拟、预测结果令人满意．

The burning model plays an important role in the steps of modeling internalcombustion engine’s working process. Generally speaking ,to establish a combustion model,three steps are needed:1)according to the actual heat release curve,select model type;2)decidethe model parameters of the known operating model;3 )select the function form between themodel parameters and the operating mode parameters, then determine the exact function. Thesteps are complicated,and what’s more,the selected model type and the function form influencethe prevision directly. BP neural networks are being used.The result shows that the step is clearand simple, the speed of computation is fast,and the most important is that the result issatisfactory.