基于神经网络和模糊控制的可变配气相位系统

提出了基于神经网络和模糊控制的汽车发动机可变配气相位控制方法。采用神经网络构成配气相位产生器，解决了发动机脉谱图的输入与输出非线性映射问题。采用模糊控制器构成可变配气相位控制系统，解决了在无被控对象数学模型情况下，可变配气相位难以有效控制问题。实验表明：基于神经网络的相位产生器，具有良好的非线性逼近能力和泛化能力。系统工作稳定，具有良好的动态性能和稳态性能，高速时汽车恒速燃油消耗降低。

The Variable Valve Timing System Based on Neural Network and Fuzzy Controller

The control method based on neural network and fuzzy controller for engine variable valve timing (VVT) was provided.The optimum valve generator was set up by the radial basis function network in order to establish nonlinear relationship between states of engine and valve timing.The fuzzy controller for VVT was given when math model can be not gained,and it eliminated shortcomings in traditional control method such as low accuracy and unsteadiness.It had been proved by vehicle road test that the valve generator based on neural network had better nonlinear approximation and generalization,the system had better static and dynamic performance,and high reliability.The equal speed fuel consumption of automobile was reduced at high speed.