基于径向基函数神经网络的内燃机气缸压力识别

Identification of Internal Combustion Engine Cylinder Pressure Based on Radial Basis Function Neural Network

提出了一种新的利用内燃机缸盖振动信号识别气缸压力的径向基函数 (radial basis function,RBF)神经网络方法。首先 ,给出了该方法的实现原理与步骤 ,并根据内燃机的工作特性 ,对径向基函数神经网络的参数进行了有效的设置 ,建立了完整的内燃机缸盖振动信号与气缸压力之间的非线性映射关系 ;然后 ,对试验数据进行了处理。结果表明 ,这种方法不仅在压力波形而且在特征点的数值上都具有较高的识别精度并有较强的鲁棒性。最后 ,对有关问题进行了讨论。

This paper presents a new approach on identifying cylinder pressure of internal combustion engine from the engine cylinder head vibration signals based on radial basis function (RBF) neural network.The identification principle of this approach and its process are presented at first.And according to the operation characteristic of internal combustion engine,the parameters of RBF neural network are effevtively set and the nonlinear mapping relationship between the vibration signals and cylinder pressure signals is established.Then,the validation of this approach on cylinder pressure identification from vibration signals is demonstrated on experimental data.The results show that the exactness of the waveform and the characteristic values of identified cylinder pressure are high and robust.At last,some relative problems are discussed.