Multiobjective electromagnetic devices design based on improved EDA with sequential optimization of the probability model

We present an improved estimation of distribution algorithm (EDA) with a new sampling probability model (SPM) to deal with the multiobjective optimization design problems of electromagnetic devices. The new SPM is constructed and updated based on the sequential optimization method. Its sampling efficiency is better than traditional methods and its sample includes more information about the optimization problem. Finally, by investigating a test function and the TEAM Workshop problem 22, we can see that the optimization efficiency of multiobjective EDA can be improved a lot by the new SPM.