In order to overcome the shortcomings of the current methods in the nonlinear analog circuit fault diagnosis based on frequency response, and select a better test stimulus signal, we used Volterra series for modeling, and proposed a method for intelligent optimization of test stimulus by using the frequency basis search together with the genetic algorithm.Firstly, the comprehensive searching technique was adopted to search the best frequency point sets, which make the output frequency point sets of each kernel not coincident with each other.Then, the frequency basis was filtered by taking the sum of the Euclidean distances between the characteristic vectors of the fault status as the objective function.Thereby the optimal test excitation signal was obtained.A nonlinear circuit was used for verification.