Dear all,
I have designed a multilevel inverter which produces either 9 level or 13 level output in the form of a staircase. There are four and six angles for the two cases, respectively. I want to calculate the switching angles using Newton Raphson method. I have the m.file code in Matlab, however, the results do not converge because the initial values are taken randomly. In NR method, a judicious choice of the initial values is required to converge quickly to the optimized angles. Can you please show me how these initial values are chosen? Is there research papers which have addressed this issue?
Thank you very much