基于神经网络的气缸压力识别研究

对传统的内燃机气缸压力识别方法存在的问题 ,提出了应用人工神经网络方法进行气缸压力识别的新方法。以 BP网络构造气缸压力识别模型。通过对网络的训练 ,用实测的缸盖振动信号识别气缸压力。结果表明 ,利用神经网络进行内燃机气缸压力识别 ,识别结果的重复性好 ,精度较高

Research on Cylinder Pressure Identification Based on Neural Networks

To avoid some shortcomings of the traditional methods, a new method for identifying cylinder pressure based on neural network( NN) is presented in this paper. A cylinder pressure identification model has bee n trained with BP neural network. Cylinder pressures are identified by utilizing the measured cylinder head vibration signals with the successfully trained neur al network. This method has taken full consideration of the nonlinear characteri stics of the cylinder head, which was usually neglected. The identified cylinder pressures coincide with the measured cylinder pressures well. And it is of good repeatability and of good resolution to identify cylinder pressure with NN.