Investigation To Computation Design And Optimization For Multilayered Radar Absorbing Coatings

Because of the complexity of designing for Multilayered microwave absorber, it requires special CAD tools. The 1th generation CAD tools are electromagnetic parameters oriented. But some absorbing layers with specific electromagnetic parameters are difficult to realize. The 2th generation CAD tools are confecting parameters oriented that are more complex than before in theory. It take the electromagnetic parameters as intermediate variables. To various methods, electromagnetic parameters are always key parameters.In this thesis we created a simple software package that combined the electromagnetic parameters database and a optimization engine based on genetic Algorithms. Using this tool we can store and manage a large number of electromagnetic parameters towards radar absorbing materials, and can also design multilayered radar absorbing coatings using the materials stored in the database. The material's electromagnetic parameters can be got either from experiment measurement or from theory computation. By the combination of GA and database technology, we solved the problem that some absorbing layers with specific electromagnetic parameters are difficult to realize.The software package is approved to be effective in designing examples. Some tests also proved that the solution towards mutilayered radar absorbing coatings presented in this paper has a good performance in computation. In order to provide high computation speed and not pay too much for creating our software package, we do all the work under open source software system. We also give a simple estimation to the OGA method, find best options of its parameters.Taking the electromagnetic parameters database as a bridge, we can use the OGA method widely and expediently than before in designing work of Multilayered RAM. Although the optimization engine in this thesis can't give a exactly global optimal solution, many local optimal solutions is effectiveenough according to the curves given by the program. So, the operator of max generation of GA need not to be set so large. Enlarging first population is helpful for good result.