基于神经网络的柴油机故障诊断专家系统

An Expert System for Modern Equipment Fault Diagnosis Based on RBF Neural Network

现代设备大多数集机电液信于一体,对其所出现的故障做出准确而迅速的判断,单纯地依靠传统的故障诊断专家系统是不够的。为此,把神经网络和传统的专家系统相结合,建立神经网络专家系统,达到优势互补的效果。在神经网络诊断中,采用MATLAB的神经网络工具包建立了径向基函数神经网络诊断模型,并以柴油机燃油系统故障诊断为例,验证了该系统的可行性。

Today most equipment are set mechatronics, therefore, it is not enough to make an accurate and prompt judgment of the failure simply relying on traditional expert system. In this paper, neural networks and traditional expert system are combined to make a neural network expert system; it has a complementary character in function. In neural network diagnosis, the neural network toolbox of MATLAB is used to establish a diagnosis model of radial basis function neural network. A diesel fuel system has been used to verify the feasibility of the system.