Journal of Intelligent Manufacturing
| Hi-index | 0.00 |
In order to simulate a nonlinear system, A BP neural network can be used. First the question is analyzed, we can know what we use to input to the system, the dimensions of the input vectors will be the number of the input layer neurons, The number of the output layer neurons depends on the output parameters, The numbers of the hidden layer neurons depends both on the input layer number and the output layer neuron number, a variety of the data is obtained from the system, it should cover almost all kinds of data, it is used to train the neural network. Before training, The goal and the epochs should be set. After training, the network has the characteristics of the nonlinear system. A group of the testing data is input, we can get the output from the simulated system. We established a BP neural network to simulate a spectrum system, it has 35 input number, 5 hidden layer number, 1 output number to distinguish two kinds. it proved that the system has the accuracy of 100%, so this kind of simulation can be used in the analysis of nonlinear system.