Modeling Of Electromagnetic Compatibility Characteristics For Electronic System Based On Intelligent Algorithms

In order to improve the efficiency of modeling issues of the electronic system electromagnetic compatibility, this paper combine the abstract methods of the physical parameters of electronic system core components and the modeling theory of intelligent algorithm, and propose a modeling method of electronic system electromagnetic compatibility characteristics based on intelligent algorithm. As an example, the method is used to establish a scattering parameters calculating model of various structures microstrip lines.Firstly, a general description method of electronic components electromagnetic properties and physical structure parameters is designed to obtain the correspondence between a variety of physical structures of electronic component and the description parameters, which is the basis of generic model for specific electronic component.Then, in order to get the training data for artificial neural network, an improved Finite Difference Time Domain algorithm is used to developed a visual electromagnetic simulation software, which is used to calculate the electromagnetic compatibility characteristics of different structures of the microstrip lines. However, since the data amount of the electron system electromagnetic compatibility characteristics is very large, it is difficult to establish a model with high precision and high efficiency by conventional modeling methods. Thus, the paper realize a combination modeling method based on mini modeling units and build an efficient model of microstrip line successfully, which can input a large number of structure parameters and calculate a wide range of frequency domain scattering parameters.In addition, the commercial simulation software CST STUDIO SUITE is used to get the scattering parameters of different structure microstrip lines and verify the results of artificial neural models.Finally, the paper establish two models of scattering parameters based on error back propagation neural network and radial basis function neural network and systematically present and analyze the general methods of modeling for electronic system electromagnetic compatibility characteristics based on intelligent algorithm.