ABSTRACT
In periodic steady-state analysis it is essential to find proper initial conditions for numerical integration. The process of simply integrating from the very beginning can be viewed as a sequence of integrations over one period, where at each integration the initial conditions are taken from the endpoint of the preceding integration. In dedicated algorithms for periodic steady-state analysis the essence is to find the initial conditions in a more efficient way than by taking the result of the preceding integration. The Newton–Raphson (NR) method with respect to initial conditions has been applied to the periodic steady-state problem by T. J. Aprille, F. R. Colon and T. N. Trick. It consists of calculating the zeros of the equation by means of the NR method, where the matrix of derivatives has been exactly calculated. The chapter discusses incremental models for calculating time-domain sensitivities with respect to a chosen initial condition.
